Contents

Preface
Introduction

Chapter 1: The Early Years

First Steps (1961-1965)

Part 1      Part 2      Part 3
International Debut (1966-1968)
Part 4

Chapter 2: Rise to the Top

World Class Results (1969-1972)

Part 1      Part 2      Part 3      Part 4
The Road to the World Title (1973-1974)
Part 5      Part 6

Chapter 3: World Champion

First Reign (1975-1977)

Part 1      Part 2      Part 3
Second Reign (1978-1980)
Part 4      Part 5      Part 6      Part 7      Part 8
Third Reign (1981-1983)
Part 9      Part 10      Part 11
First Two Kasparov Matches (1984-1985)
Part 12      Part 13      Part 14

Chapter 4: Later Years

Challenging Kasparov (1986-1987)

Part 1      Part 2
Fruitful Years (1988-1990)
Part 3      Part 4      Part 5      Part 6

Epilogue
Index of Opponents
Endgame Classification
Swipe left for next chapter
 Tibor Karolyi & Nick Aplin

Endgame Virtuoso Anatoly Karpov

SECOND, IMPROVED EDITION

New In Chess 2007

© New In Chess
First edition February 2007
Second edition November 2007
Published by New In Chess, Alkmaar, The Netherlands
www.newinchess.com

All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by
any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission from the
publisher.

Cover design: Steven Boland

Supervisor: Peter Boel
Proofreading: René Olthof
Production: Anton Schermer

ISBN-13: 978-90-5691-202-4
Introduction

To become a World Champion one has to play at the highest level in all aspects of the game. The giants of chess history
had their strong points and their chess was based on a juxtaposition of the elements in different ways. When it comes to
the World Champions, all of them have been capable of playing marvellous endgames and some were exceptional in
their level of application. Mikhail Botvinnik, for example, was outstanding, and yet he thrived especially in
complicated middlegames. Vasily Smyslov and Bobby Fischer were both superb, creating a strong artistic impression
in their endgames.
José Raul Capablanca and Anatoly Karpov have been the two champions who relied most on their exceptional endgame
skills. As Karpov reigned much later than the Cuban, at a time when competitive standards were generally much
higher, it is fair to assume that the Russian played endgames at the highest-ever level.
Changes in the tournament rules of chess have contributed to the preservation of his status. Karpov was the last World
Champion to play the majority of his career games with the rule of adjournment in operation. The last World
Championship final with adjournments was the 1996 Karpov-Kamsky match.
Garry Kasparov also played several matches under these conditions, yet a significant proportion of his reign had no
adjournments. In addition, his style of play was more ferocious, which automatically meant fewer endgames.

Going deeply into the analysis of Karpov’s endgames has highlighted certain aspects of his play to the author of this
book. This collection of games reveals some interesting characteristics. Karpov is always concerned to improve the
scope his pieces before acting and he brings his king into play relatively late. He will fight for the open files and often
manoeuvres his rooks to the seventh rank. It transpired that there was often a fine line between a strong king and a
vulnerable one in the centre. Anatoly had a well-defined sense of timing such strategies with the king. One other
notable characteristic which appears to contradict general rules was that quite a number of times, he placed his pawns
on the same colour as his opponent’s bishop in order to limit their movement. He was not afraid that the bishop would
hunt down the pawns.
No player has ever produced as many magnificent over-the-board endgames as Karpov. Of course such a talent does not
only come from the understanding of a chess genius. It has much to do with excellent concentration, fighting spirit and
a strong determination to grind down your opponents.
In this book we have followed the plan of selecting and analysing some of Karpov’s most breathtaking endgames, and
then drawing some conclusions on his endgame style. As a rule, the analysis starts at the moment that the players went
for simplification. These endgames will bring joy to all who study them, but chess instruction is not the chief purpose
of this book. My main objective is to help young players to learn from Karpov. Ever since Karpov played these games,
the level of analysis has been raised dramatically. Our understanding of endgames has been deepened through the use
of computer programs.
Karpov has analysed many of his games himself. Arguably, even though he is a giant at the chessboard, his published
analysis is not as impressive. In Russian chess culture there have even been some lighthearted comments about the
depth of his analysis. Marin has suggested that certain sections are actually not Karpov’s, the name attached being the
only link. Like many players, Karpov tended to select games for analysis on the basis of their importance for his career
rather than for their artistic value.

The idea to devote an entire book to Karpov’s endgames was born when I had a short conversation with Jan Timman. I
asked him whose playing style was closest to Kasparov’s. The mere mention of Kasparov’s style generated a smile of
admiration from the Dutch grandmaster. Somehow our conversation turned to my best pupil, Peter Leko. The Dutch
grandmaster told me that Karpov had praised Leko’s technique.
The idea stayed with me for quite a while, and once we had finished the second volume of Kasparov’s games, I started
to explore Karpov’s endgames. Later grandmaster Yury Razuvaev, who has worked with Karpov for many years and
probably knows him best, told me that Peter Leko’s endgame play is very similar to Karpov’s. Like all great players,
Peter becomes very shy when it comes to talking about the type of training that helped him become a world-class
player. However, as I know, he never worked systematically on his endgames with anybody other than me. When I
trained Leko as a junior, we went through many of Karpov’s endgames. Today I believe the quality of Peter’s endgame
play speaks for itself – those sessions have not damaged him in any way.
Karpov’s career consists of almost 4 full decades. In the 1960s, as a talented junior, he reached an intermediate summit
by becoming Junior World Champion. At this age he was already capable of playing exceptionally fine endgames. In
the 1970s, he moved on from being a young grandmaster to becoming a World Champion who subsequently dominated
the chess world. In the 1980s, the pattern changed. From being the very best player he had to settle for being the world
number two. Subsequently, his top-ten status signalled a decline. We have included games up until 1990, when Karpov
lost his last match with his successor Garry Kasparov.
Some of Karpov’s endgames have already become classics and excellent commentators have fashioned some beautiful
analysis. These analysed games occupy extra space in the book. There are a dozen positions that motivated much
deeper analysis. Sometimes the analysis became even further extended, but in view of the importance and the beauty of
these lines, it seemed essential to retain them all.
Just when the major part of this book had been completed, out came Kasparov’s study of Karpov in the My Great
Predecessors series. Garry analyses 40 games from Anatoly’s career, but there are few that we have both selected.
Happily, a number of moves that I was proud to discover, were also introduced by Garry in his book.

The ‘I’ in this book refers to Tibor Karolyi, the chief author. I have been a professional player for a dozen years and
have spent many years as a trainer; Peter Leko was my pupil in his junior years, and Zoltan Gyimesi is now. I have
been Zsuzsa Polgar’s training partner in the past, as well as Zsofia Polgar’s trainer for half a year.
The co-author, Nick Aplin, has been an enthusiastic chess amateur for long and is periodically a manager of junior and
senior chess teams travelling from Singapore.

January 2007
Tibor Karolyi
Chapter 1
The Early Years

In the databases the first games of the future World Champion are from 1961.
Karpov learned to play chess at the age of four and at the remarkable age of seven he obtained the level of third category
player. Two years later, in 1960, he became a second and then a first category player, which is already quite a decent
level. In 1965, Anatoly started to play strong junior tournaments and he became the youngest ever Soviet master so far.
At the end of the 1960s he was already winning international tournaments.
Even throughout these early years, his games show a tendency towards the endgame rather than complicated openings
and middlegames.

First Steps (1961-1965)

In the year 1961, when the Soviets sent their first astronaut, Yury Gagarin, into space, Anatoly was ten years old. At this
time, he already showed exceptional qualities in his endgame play. From these first years, until 1965, few games have
been published. Still we can see that Karpov’s endgame arsenal gradually widened and he displayed many weapons
which would later feature in his endgame play as a champion. Even early statistics show that most of his wins were
achieved after long and slow games.
The first game we present to you is against Viacheslav Kalashnikov. This is the only game in the book that is analysed
from start to finish, because an ending is reached quickly after the opening. Young Karpov’s knowledge of opening
theory was limited and the focus in this book will be on the endgame.

GAME 1
Kalashnikov, Viacheslav
Karpov, Anatoly
Zlatoust 1961 (4)

1.d4 Nf6 2.Nf3 e6 3.c4 b6 4.g3 Bb7 5.Bg2 c5?! 6.0-0 Be7?! 7.Nc3 cxd4 8.Nxd4 Bxg2 9.Kxg2 d5?

The ten-year-old Tolya doesn’t mind simplification, so an endgame-like position is soon reached.
10.Qa4+!

This move shows that White is not an amateur player. Without the check White would have no chance to obtain an
advantage.

10...Qd7 11.Qxd7+ Nbxd7

Interestingly, 11...Kxd7 has never been tried. The h8-rook can come into play quickly and, more importantly, the c7-
square is covered. In this type of position the Black king is sometimes a strong piece, which holds the position together;
sometimes it becomes a target. Play may continue:

analysis diagram
12.Rd1! (after 12.cxd5 Nxd5 13.Nxd5 exd5 14.Nf5 g6 Black holds) 12...Bb4 (in case of 12...Rd8 13.cxd5 Nxd5
14.Nxd5 exd5 15.Nf5 the pawn drops; after 12...Nc6 13.cxd5 Nxd4 14.Rxd4 exd5 15.Bg5 Ke6 16.Rad1 Rhd8
17.R1d3! White can trouble Black’s king in the centre) 13.cxd5 exd5 14.Ndb5 Bxc3 15.Nxc3 Ke6 (15...Kc6 loses to
16.Bf4 Nbd7 17.Rac1) 16.Bg5 Nbd7 17.Nb5 Rhc8 (in case of 17...Kf5 White is better after 18.Bxf6 Nxf6 19.Rac1
Kg6 20.Rc6) 18.Rac1 (White can also keep up the pressure with 18.Nd4+ Ke7 19.Nf5+ Kf8 20.Rac1! h6 21.Bf4 Rxc1
22.Rxc1 Kg8 23.Rc7 with a tough ending for Black) 18...Rxc1 (after 18...Ne8 19.Nd4+ Kd6 20.Bf4+ Ke7 21.Nc6+
Ke6 22.Rd3 Black’s king is in trouble) 19.Rxc1 Ne8

analysis diagram

20.Rc6+ (20.Nd4+ Ke5) 20...Kf5 21.Bf4 and Black is struggling to survive.

Back to the game.

12.cxd5 Nxd5!

Navigating his way through contradictory principles Karpov finds a good plan. Perhaps it suited him to work with fewer
pieces. Did he already know or was it just a natural feel of the future champion? With an isolated pawn it is normally
better to keep the minor pieces on and exchange the heavy ones if the opponent has the upper hand.
The point is that with more minor pieces he can prevent White from exerting pressure on the isolated pawn and can lure
his opponent into a pin. If all the minor pieces were off the board, White would double his rooks and exert pressure
with e2-e4, when Black’s rooks defend the d-pawn on the d-file.
According to this principle Black should not exchange, yet Karpov correctly swaps the knights. Principles sometimes
clash, and this time one of the basic principles takes precedence: centralization of the king is often a key factor in
endgames. Naturally, by delaying castling Black can achieve this goal more quickly.
Karpov’s is a remarkable decision. One can only speculate if he could evaluate so clearly, or if he just followed his
instincts.
After 12...exd5, 13.Ndb5 (13.Rd1) 13...Kd8 14.Rd1 a6 15.Nd4 Bc5 16.Bg5 h6 17.Bxf6+ Nxf6 18.Nb3 wins a pawn.
13.Nxd5 exd5

14.Nf5?

This is a mistake. White fixes the isolated pawn structure, but this gives him little to play for. Interestingly this ending
occurred later in Münch-Wunder, Wildflecken 1993. 14.Bd2 Bf6! 15.Bc3 Bxd4 16.Bxd4 f6 17.Rac1 Ke7 was played
and Black held the position without any particular problem.
14.Nb5! was the right move. Black’s pieces are well developed; however, the unfortunate constellation of the a8-rook
and the d5-pawn virtually forces him to sacrifice the d-pawn:
A) 14...Kd8? 15.Rd1 Nf6 16.Nc3 and Black has nothing for the pawn;
B) 14...Bd8? 15.Rd1 Nf6 16.Nc3 Rc8 17.Nxd5 Nxd5 18.Rxd5 Rc2 19.Rd2 Rxd2 20.Bxd2 Bf6 21.Rc1! wins for White;
C) 14...0-0 (this time connecting the rooks is an idea, even if it costs a pawn. Passive defence gives almost no hope of
survival) 15.Nc7 Rac8 16.Nxd5 Rfe8 (in case of 16...Bf6 17.e4 Rfe8 18.Nxf6+ (18.f3 Bd4!) 18...Nxf6 19.f3 Rc2+
20.Rf2 Rec8 21.Be3 Rxf2+ 22.Bxf2 Rc2 23.b4 White will move his pawns from the second rank. He has reasonable
winning chances)
analysis diagram

C1) 17.Ne3 Nf6 18.Rd1 gives White chances, as it prevents the rook invasion. On the other hand, his pieces are tied
up;
C2) In case of 17.e4 Nc5 18.f3 Nd3 19.Rd1 Rc2+ 20.Rd2 Rxd2+ 21.Bxd2 Nxb2 22.Rc1 Bc5 23.Bc3 Na4 24.Ba1 f5!,
Black is in the game after creating a weakness in White’s camp;
C3) With 17.Be3! White finishes his development and obtains an advantage: 17...Rc2 (after 17...Nf6 18.Nxe7+ Rxe7
19.Rac1 Rce8 20.Rfe1 Nd5 21.Bd2 Black has nothing for the pawn) 18.Rfc1! Rxe2 19.Rc7 and White wins.
D) 14...Rc8. Black is obliged to sacrifice a pawn because rooks belong on open files. However, he obtains no full
compensation, viz. 15.Nxa7 Rc2 16.Rd1 (the centralization 16.Kf3 is premature on account of 16...Ne5+, and after
16.e3 Bf6! 17.Rb1 Ke7 Black has great control for the pawn) 16...Rxe2 (16...Bf6 17.Rxd5 Bxb2 18.Bxb2 Rxb2
19.Rad1 loses, or 16...Nf6 17.Rd2 Rc5 18.b4 Rc4 19.a3 and White is a safe pawn up) 17.Rxd5 Bc5 (in case of 17...Nf6
18.Rd2 Re1 19.Nc6 White wins) 18.Rd2 Re5 19.a4 – White rescues his knight and has excellent winning chances due
to his extra pawn.

14...g6 15.Nxe7 Kxe7

Objectively, this ending is not dangerous for Black as the king can help in the centre and is not exposed there.

16.Rd1 Ke6!

The king not only stands safely in the centre but also finds an active role there.

17.Bd2 Rac8

After 17...Nc5! 18.Bc3 Rhc8 19.Rac1 (Black has only one minor piece, but it stops the doubling of rooks on the d-file.
If 19.Rd4 Ne4) 19...Rc7 20.Bd4 Rac8 Black’s play on the c-file gives him an easy endgame.

18.Bc3 Rhd8 19.Rd4 Nc5

Exchanging a pair of rooks with 19...Rc4!? deserved attention, as then Black’s king would have less to fear, for instance
20.Rxc4 (or 20.Rad1 Rxd4 21.Rxd4 Nc5 and White can create little) 20...dxc4 21.Rd1 Rc8 and Black has no problems.

20.f3 Rd6 21.Rad1 Rcd8

22.g4!

White’s aim is to gain space and fix the h7-pawn. Karpov himself would use this theme later in his career to win superb
technical games, even at Linares.

22...f6!

Karpov senses the danger early. He doesn’t let Kalashnikov fix his h-pawn.

23.h4 h6 24.Rf4!

Injecting some power into the equal position.

24...Nd7 25.e4

White is making Black work to earn a draw. After 25.g5 hxg5 26.hxg5 fxg5! (26...f5?! renders Black’s pieces passive:
27.Rh1! Rc6 28.Rh6 Nf8 29.Kg3 and White will open up the position with e2-e4. Black has no active plan) 27.Rg4 Nf6
28.Rxg5 d4 29.Bd2 Kf7 30.Rc1 R8d7 and Black can hold, just keeping his position together, as White cannot really
improve the position of his king, thanks to the doubled rooks.
29...Rc6 is a path to the draw, too, but it is a narrow one: 30.Rxg6 Kf7 (30...Rc2 31.Bg5 Kf5 32.Bxf6 Rd6 33.Rxd4
Rxd4 34.Bxd4 Kxg6 – this position is drawish without the pair of rooks) 31.Rg5 Rc2 32.Rb5 Nd5 (Black has a good
grip in return for the minus pawn. It is hard for White to make any progress) 33.Bg5 Rxe2+ 34.Kf1 Rc2 35.Rxd4 Rh8
36.Kg1 Rhh2 37.Rbxd5 Rcg2+ 38.Kf1 Rf2+ 39.Ke1 Rfg2 and Black has reached the draw.

25...Ne5 26.exd5+ Rxd5

27.Re1!

In the event of 27.Rfd4 Rxd4 28.Rxd4 Rxd4 29.Bxd4 Kd5 Black is safe.

27...a5?!

Black is not just waiting and hoping to get away with a draw. Endings can culminate in a pawn race. It can be useful to
have a pawn closer to the promotion square. Such understanding is remarkable for a boy of this age. The move shows
that the young boy is not intimidated when playing against an adult. This is a good quality for a chess player. On the
other hand the text move turns out to be the cause of his difficulties.
Building a fortress in the centre with 27...g5! was clearly better. After 28.Rfe4 (28.Rf5 Rg8 29.Kg3) 28...R8d7 29.hxg5
hxg5 30.Rh1 (or 30.f4 gxf4 31.Rxf4 Rg7 and Black is safe) 30...Kf7 (or 30...Rc7 31.Rh6 Rg7 – Black holds on by doing
nothing) 31.Rh7+ (31.Kg3 Re7) 31...Kg6 32.Rxd7 Rxd7! (after 32...Nxd7 33.Re7 a5 34.Kf2 Nc5 35.Ke3 Na4 36.Bd4
White is a little better) 33.Kg3 Rd5 it is hard to put pressure on Black.

28.g5!

White opens up the position – credit is due to him for the last couple of moves.

28...hxg5 29.hxg5 f5

Black needed to calculate that the pin is not lethal.

30.Rh4
30...Kd6!

Getting out of the pin in time. This is the only move here and almost anybody would find it, but Tolya had to calculate
and assess it well in advance.

31.Bxe5+

If 31.Rh6 Nd3.

31...Rxe5

The ensuing rook ending is rich in potential and ripe for analysis. Just how many of the intricacies were apparent to
Karpov is a matter of speculation.

32.Rd4+ Kc7!

Going for this in advance shows confidence and experience in endings, in addition to good calculating ability. Black
saves his rook with a small finesse.
After 32...Kc5 33.Rxd8 Rxe1 (in this line Black’s king stands much better on c5) 34.Rg8 Re2+ (the race can be initiated.
Alternatively, 34...Re6 is simple for Black) 35.Kg3 Rxb2 36.Rxg6 Rxa2 37.Rf6 Ra1! 38.g6 (or 38.Rxf5+ Kc4 39.Rf4+
Kd5 40.Kg2 Ra2+ 41.Kh3 Ra1 42.Rg4 Rh1+ 43.Kg2 Rh8 and Black’s pawns are not slower) 38...Rg1+ 39.Kf4 a4
40.Rxf5+ Kb4 41.Rg5 Rxg5 42.Kxg5 a3 Black draws. However, 33.Rxe5+! (here is the difference: White can take the
rook with check) 33...Kxd4 34.Re6 Kc5 35.Rxg6 Rd2+ 36.Kg3 Rxb2 37.Rf6 Rxa2 38.g6 Ra1 39.Kf4 wins for White!

33.Rc4+

Alternatively, 33.Rxe5 Rxd4 34.Re6 Rd6! holds for Black. Going for a pawn race is dangerous, for example 34...Rd2+?
35.Kg3 Rxb2 36.Rxg6 Rxa2 37.Rf6 a4 38.g6 Ra1 39.Kf4 a3 (or 39...Rg1 40.Rf7+ Kc6 41.Kxf5 b5 42.g7 and White is
faster) 40.Rf7+ Kc6 41.Ra7 b5 42.Kxf5 b4 43.g7 Rg1
analysis diagram

44.Kf6. Black loses the race.

33...Rc5?

33...Kd6! was possible. Perhaps Black had had enough of hanging rooks. After 34.Rd1+ (after 34.Rxe5 Kxe5 35.Rc6
Kf4 Black’s king is excellently placed) 34...Rd5 35.Rxd5+ Kxd5 36.Rc2 Re8 White cannot get through to g6.

34.Rxc5+ bxc5 35.Kg3 Rd2 36.Re6 Rd6!

In most cases, removing as many pawns from the board as possible saves the weaker side. Being a pawn down with
little material is not always fatal. Often such positions can be held in rook endings. However, this time White’s g-pawn
would be rather dangerous. After 36...Rxb2? 37.Rxg6 c4 38.Re6! (Black might even take over after 38.Rh6 c3 39.Rh1
(39.Rh4 Rb4 40.Rh1 Kd6 and Black is at least not worse) 39...c2 40.Rc1 Kd6 41.Kf4 Ke6 42.a3 a4) 38...c3 39.Re3 c2
40.Rc3+, Black’s rook is dangerously passive. White will exert huge pressure with g5-g6 and Kf4.

37.Re7+
37...Rd7

The simpler way to play.

Black could have entered a very exciting pawn race by playing 37...Kc6. He survives only with the use of some highly
unusual motifs: 38.Kf4 and now:
A) 38...a4. The a-pawn is no further from its promotion square than the c-pawn, but it is harder to remove the blocking
pawn;
A1) 39.Ke5
A11) 39...c4
A111) 40.Ra7 Rd5+ 41.Kf6 Rd6+ 42.Kf7 Kb5 43.a3 f4! (Black suddenly changes tack and goes after the f-pawn)
44.Rc7 Rd3 45.Kxg6 Rxf3 46.Kg7 Rg3 47.g6 f3 48.Rf7 c3! 49.bxc3 Kc4 50.Rc7+ (50.Kh7 Kxc3 51.g7 Kb2) 50...Kb3
51.c4 Kxa3 52.c5 Kb4 53.c6 Kb5 54.Ra7 f2 and Black escapes to a draw;
A112) 40.a3! Rd2 41.Re6+ Kb5 42.Rxg6 Rxb2 43.Rg8 c3 44.Rc8 – White stops the c-pawn and wins with the g-pawn.
A12) 39...Rd2! 40.Re6+ Kb5 41.Rxg6 Rxb2 42.Rg8 Rxa2 43.g6 Rg2 44.g7 Kc4 45.Kf6 (it looks as if White wins)
45...a3 46.Ra8 Kb3 47.g8=Q+ Rxg8 48.Rxg8
analysis diagram

48...c4!!. Remarkably, Black pushes the pawn which is closer to White’s king and two squares further back than the a-
pawn. Endgames can be magic! 48...a2? would lose to 49.Rb8+ Kc2 50.Ra8 Kb1 51.Ke5!.
49.Rb8+ (49.Ke5 c3 50.Kd4 c2 51.Rb8+ Ka2 52.Rc8 Kb2 is no win; 49.Kxf5 c3 50.Rb8+ Kc2 (or 50...Ka2) 51.Ke4 a2
52.Ra8 Kb1 53.Kd3 c2 is a draw again)
49...Kc2! (Black changes direction. After 49...Ka2 50.Rc8 Kb3 51.Ke5! a2 (51...c3 loses to 52.Kd4 c2 53.Kd3)
52.Rb8+ Kc2 53.Ra8 Kb2 54.Kd4 c3 55.Rxa2+ Kxa2 56.Kxc3 the pawn ending is lost) 50.Ke5 c3 51.Kd4

analysis diagram
51...Kd2! 52.Rh8 (this rook move is rather scary) 52...c2 53.Rh2+ Kd1 54.Kd3 c1=N+! (underpromotion saves the day)
55.Kc3 Ne2+.
A2) The winning line is 39.a3! (removing the a-pawn from the sensitive second rank) 39...Kd5 (39...c4 40.Ra7) 40.Ra7
Rb6 41.Rxa4 Rxb2 42.Ra6 c4 43.Rxg6 c3 44.Rg8 Kc4 45.g6 Rb7 46.Rc8+ Kd3 47.Kxf5 c2 48.f4 and White’s pawns
march on.
B) 38...c4! (pushing the c-pawn without hesitation) 39.Ke5 Rd2 40.Re6+ Kc5 (40...Kb7? loses to 41.Rxg6 Rxb2
42.Rh6 (42.Kd4? Rxa2) 42...c3 43.Rh1 c2 44.Rc1) 41.Rxg6 Rxb2 42.Rg8 Rxa2 43.g6 Rg2 (Black is not slower in the
race) 44.g7 c3 45.Rc8+ Kb4 46.g8=Q Rxg8 47.Rxg8 c2 48.Rc8 Kb3 49.Kxf5 Kb2 50.Ke4 c1=Q 51.Rxc1 Kxc1 52.f4
(52.Kd3 Kb2) 52...a4 53.f5 a3 54.f6 a2 55.f7 and the race finishes with no winner.
Over the board, much of this analysis would probably have been beyond the capability of the young Karpov. But that is
not to say that his post-game analysis could not have revealed some of the details.
Back to the game.

38.Re5

After 38.Rxd7+ Kxd7 39.Kf4 Ke6 White has no squares to penetrate and has to give way: 40.Ke3 (40.b3? Kd5 41.Ke3
c4 42.bxc4+ Kxc4 43.Kf4 Kd4! and shouldering the king wins for Black) 40...Kd5 41.Kd3 (after 41.f4 Kc4 White is
in zugzwang) 41...Ke5 (41...c4+ 42.Kc3; 41...a4 42.b3 a3 43.f4) 42.Kc4 Kf4 43.Kxc5 Kxf3 44.b4 (44.a4 f4 45.b4
axb4 46.Kxb4 Ke2 47.a5 f3 48.a6 f2 49.a7 f1=Q 50.a8=Q Qf4+ 51.Kc5 Qxg5+ 52.Kb6 is a draw) 44...axb4 45.Kxb4
f4 46.a4 Kg4 47.a5 f3 48.a6 f2 49.a7 f1=Q 50.a8=Q. This queen ending is also a draw.

38...Kd6 39.Re8
39...Kc6?!

Karpov clears the way for his rook to invade on the d-file, but this is not the most straightforward line of play.
A) Better was 39...Kd5 40.Ra8 c4 41.Rxa5+ Kd4 42.Ra6 Rb7 43.Rxg6 Rxb2 44.Kf4 c3 (44...Rxa2 loses to 45.Rc6)
45.Rc6

analysis diagram

45...Rb4!! (this sets up an effective battery. Black forces White to take his pawn and then sets up the same battery again.
After 45...c2? 46.g6 Kd3 47.g7 Rb8 48.Kxf5 Black is one tempo short of a draw. Or 45...Kd3 46.g6 Rf2 47.a4! c2
48.a5 Rf1 49.Rxc2 Kxc2 50.a6 Rd1 51.g7 Rd8 52.Kxf5 and wins) 46.Kxf5 Rb5+ (in case of 46...Rc4 47.Rxc4+ Kxc4
48.g6 White promotes with check) 47.Kf4! (after 47.Kf6 Rc5 48.Rxc5 Kxc5 49.g6 c2 50.g7 c1=Q 51.g8=Q Qf4+
Black holds) 47...Rb4! (the battery also works without the f5-pawn. Not 47...Rc5? 48.Rxc5 Kxc5 49.Ke3 as White
reaches the square of the pawn) 48.Kg3 Kd3 (on 48...Rb5? 49.g6 wins) 49.g6 c2 50.g7 Rb8=.
B) 39...c4!? (Black fixes the b-pawn and enables his king to move forward at the same time. This was a very tough
move to find) 40.Rg8 (40.Ra8 Rb7 41.Ra6+ Ke5 (after 41...Kd7 42.Rxg6 Rxb2 43.Rf6 c3 44.Rxf5 Kd6 45.Rf6+ Kd7
46.Rf4 Rb6 Black is also safe) 42.Rxa5+ Kd4 43.Ra6 Rxb2 44.Rxg6 c3 45.Rc6 c2 46.g6 (46.Kf4 Rb4!) 46...Kd3 and
Black holds) 40...Rb7 41.Rxg6+
B1) 41...Ke5
B11) 42.Ra6 Rxb2 43.Rxa5+ Kd4 44.Ra8 (44.Kf4 c3 45.Ra8 c2 46.Rc8 Rb4! – this battery saves Black again – 47.Rxc2
Kd3+ 48.Kxf5 Kxc2 49.g6 Kd3 50.g7 Rb8 51.a4 Ke3 52.a5 Kxf3 53.a6 Ra8 with a draw) 44...c3 45.Rc8 Kd3 46.Kf4
c2 47.g6 Rb1 48.Kxf5 c1=Q 49.Rxc1 Rxc1 50.g7 Rg1 51.Kf6 Ke3 52.a4 (52.Kf7 Kxf3 53.a4 Ke4 Black’s king
reaches the square) 52...Kxf3 53.a5

analysis diagram

53...Rg4! (this is a nice way to activate the rook, particularly as all the other moves lose) 54.a6 Rf4+ 55.Ke7 Rg4
56.Kf7 Rf4+ and Black draws;
B12) 42.Rc6
B121) 42...f4+ 43.Kg4 Rxb2 44.Rxc4 Rg2+ 45.Kh5 Rxa2
analysis diagram

46.Kg6!! – White wins the f4-pawn in a remarkable way and scores the point;
B122) 42...Rxb2 43.Rxc4 Rxa2 44.Rc5+ Ke6 45.f4 Ra3+ 46.Kh4 Rf3 47.Re5+ Kd6 48.Rxf5 – White wins;
B123) 42...Kd5 43.Rc8 (in the event of 43.Ra6 Rxb2 44.g6 c3 45.Rxa5+ Kd4 46.g7 Rb8 Black holds) 43...Rxb2 44.Kf4
Rxa2 45.g6 Rg2 46.Kxf5 Kd4 47.f4 c3 48.Kf6 c2 49.g7 Kd3 50.Kf7 a4 and Black is not worse in the race.
B2) 41...Kd5 42.Rg8 (42.Ra6 Rxb2 43.g6 c3 44.g7 Rb8 45.Rxa5+ Kd4 46.Ra7 c2 (46...Rc8 47.Rc7) 47.Rc7 Kd3
48.Kf4 Rg8 draws) 42...Rxb2 43.Kf4 Rxa2 (but not 43...c3 44.g6 Rg2 45.Kxf5 c2 46.Rc8 Kd4 47.f4 Kd3 48.Kf6 Rg4
49.f5 Rc4 50.Rxc4 Kxc4 51.g7+–) 44.Kxf5 (after 44.g6 Rg2 45.Kxf5 c3 46.g7 c2 47.Rd8+ Kc4 48.g8=Q+ Rxg8
49.Rxg8 Kc3 50.Rc8+ Kb2 51.Ke4 c1=Q the race is undecided) 44...Rf2 45.f4 c3 46.g6 c2 47.Rc8 Kd4 48.g7 Rg2
49.Kf6 Kd3 50.Kf7 a4
51.Rd8+! is equal.
Back to the game.

40.Kf4 Rd2

After 40...Kd5 41.Ra8 Rb7 42.b3 Rb4+ 43.Ke3 White has good chances.

41.Re6+ Kd7!

Karpov’s king keeps an eye on the g-pawn. He seizes the best tactical chance. Not 41...Rd6?? 42.Ke5; or 41...Kd5?
42.Rxg6 Rxb2 43.Ra6 Rxa2 44.g6 c4 45.Kxf5 c3 46.g7 Rg2 47.Rg6 and White wins.

42.Rxg6 Rxb2 43.Kxf5?

White has played the endgame powerfully but now he commits a mistake, probably throwing away the win. It was not
urgent to take the f-pawn – it wastes a tempo in the pawn race. Better was 43.Ra6!, opening the route for the g-pawn. It
also attacks a pawn:
A) 43...c4
A1) 44.Rxa5! Kd6 45.Ra8 wins;
A2) 44.g6 c3 45.Rxa5 Rb4+ (45...Kd6 loses to 46.Ra3! Rb4+ 47.Kg3 Rb1 48.Rxc3) 46.Kg3 Rb1 47.Rc5 Rg1+ 48.Kf4
Rxg6 49.Rxc3 Ra6 50.Kxf5 Rxa2 and now 51.Re3! is the only road to victory.
B) 43...Rxa2 44.g6 (44.Kxf5 c4 45.f4 c3 draws) 44...Rg2 45.Kxf5 c4 46.Rxa5 c3
analysis diagram

Now, less compelling is 47.Rc5? – a natural move, yet second best: 47...c2 and now:
B1) If 48.f4? Kd6 49.Rc4 (49.Rc3 Ke7 50.Re3+ Kf8 draws) 49...Ke7 (one should never forget about zugzwang in
endings:) 50.Rc3! (can Black move now or not?) 50...Rf2 (or also 50...Rd2 51.g7 Kf7 52.Rc7+ Kg8 53.Ke6 (53.Kg5
Rd7 draws as well) 53...c1=Q 54.Rxc1 Kxg7=) 51.g7 (in case of 51.Kg5 Rg2+ 52.Kh6 Kf6 53.Rc6+ Kf5 Black draws)
51...Kf7 52.Rc7+ Kg8 53.Kg6 Rg2+ 54.Kf6 – Black seems to be in huge trouble, but...

analysis diagram

54...Rg6+!! 55.Kxg6 c1=Q draws elegantly.

Nor should one ever forget about stalemate when it comes to playing endings!
B2) 48.Kf6! (White’s king invades) 48...Rf2 49.Kg5 Rg2+ 50.Kh6 Kd6 51.Rc3 and now:

B21) 51...Rh2+ 52.Kg7 Rf2 53.f4 Ke6 54.Kg8 wins;

B22) 51...Ke6?! (this does not spoil the position but makes it harder to defend) 52.Kh7 Kf6 (52...Ke5! 53.f4+ Kf5 still
draws) 53.Rc6+ Ke7 54.f4 Kd7 55.Rc3 Ke6

analysis diagram

56.Kg7!! and Black is in zugzwang! 56...Rd2 (56...Kf5 57.Kf7) 57.Kg8 Rg2 (or 57...Rd8+ 58.Kh7 Rd2 59.g7) 58.g7
Rf2 59.Rc8 Kd7 (59...Rg2 60.f5+) 60.Rc4 Ke7 61.f5 Kd6 62.f6 and Black cannot resist anymore.
If 56.g7 Kf7 57.Rc7+ Kf6 58.g8=N+ Kf5 59.Rc4 Ke6!! (surprisingly the black king has time to approach the c-pawn
and draw. Not 59...Kg4 60.Nf6+ Kf3 61.Nd5 Rd2 62.f5! and Black’s king stands so awkwardly that he cannot save the
game: on 62...Rxd5 63.f6 wins) 60.Nh6 (after 60.Rc5 Black can also win the rook for his c-pawn: 60...Kd6 61.Rc8 Kd5
62.Ne7+ Ke4 and Black holds) 60...Kd5 61.Rc7 Ke4 62.f5 Kd3 draws;
B23) 51...Ke5! (the black king approaches the c2-pawn, which looks so far away) 52.Kh7 (in case of 52.Kg7 Kd4
53.Rc8 Kd3 54.f4 Rf2 Black reaches the draw) 52...Kd4 53.Rc7 Rh2+ (53...Kd3?? 54.g7) 54.Kg8 Kd3 55.g7 Rf2 with
a draw.
Back to the analysis diagram after 46...c3 (line B previous page). Instead of 47.Rc5 White should play 47.f4! (quite
amazingly, here the rook works more efficiently on the first rank, in front of the enemy pawn. Exceptions to the rule
make the game of chess wonderfully intriguing) 47...c2 48.Ra1 Ke7 (48...Rg3 49.Kf6 Ra3 50.Rc1; or 48...Rd2 49.g7)

analysis diagram

49.Re1+. The advantage of having the rook on the first rank is that it can control both the e-file and the pawn from the
e1-square; 49...Kf8 50.Kf6 Rg3 51.Ra1 Ke8 52.g7 and White wins.
Back to the game.
43...c4 44.Ke4

After 44.Rg7+ Kd6 45.g6 c3 46.Rg8 Kd7 Black draws.

44...Rd2

Also a draw is 44...c3 45.Kd3 Rxa2 46.f4 Rf2; or 44...Rxa2 45.Kd4 c3.

45.Ra6 Rxa2 46.Kd4 c3! 47.Kxc3 Ra3+ 48.Kd4 Rxf3 49.Rxa5

49...Ke7!
The rest is evident. Kalashnikov would probably not have tried to win this position against an adult.
What would have happened if Karpov had played 49...Rf1 instead? Surprisingly, the game can end with a most
interesting mutual zugzwang position.
50.Re5 Rf4+ 51.Kd3 Rg4 is the simplest draw. 51...Rf2 also draws after 52.Ke4, if Black finds the only move 52...Rf1!!.
51...Rf1, however, loses to

analysis diagram

52.Ke4! – incredibly, a mutual zugzwang position is reached.

50.Ke5 Kf7 51.Ra6 Kg7 52.Rh6 Rf2 53.Rd6 Rf1 54.Ke4 Rf2 55.Rh6 Rf1 56.Ke5 Rf2 57.Ke6 Rf3 58.Rf6 Re3+ 59.Kf5
Rf3+ 60.Ke6 Re3+ 61.Kf5 Rf3+ 62.Ke5 ½-½

This game, which seemed boring on the surface when it was selected, in reality proved to be an entertaining and
fascinating battle crediting both players. This is a remarkable game from a player who was just starting out on his
career, more than four decades ago. There are mistakes, and yet one can see positive signs and the qualities of the future
endgame maestro.
Zlatoust, where Karpov was born and where this game was played, means ‘gold’. Gold-diggers often have to dig deep.
In this game Karpov’s skill lies right below the surface, it shines through like gold.
 GAME 2
Karpov, Anatoly
Mukhudulin
Zlatoust 1961

1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 d6 6.Be3 e5 7.Nb3 Be7 8.f3 0-0 9.Bc4 a6 10.Qd2 Qc7 11.Bd5 Nb4
12.Rc1 Nfxd5 13.Nxd5 Nxd5 14.Qxd5 Be6 15.Qd2 f5 16.exf5 Rxf5 17.0-0 Raf8 18.c4 Qd8 19.Na5 Bg5 20.Bxg5
Rxg5 21.Kh1 Rh5 22.g3 Qc7 23.Rfd1 Bh3 24.Qd5+ Rf7 25.b4 Qb6 26.c5 Qb5 27.Qc4 Qxc4 28.Rxc4 dxc5 29.Rd8+
Rf8 30.Rxf8+ Kxf8 31.bxc5 Be6 32.Rb4 Bxa2 33.Rxb7 Bd5 34.c6 Bxf3+ 35.Kg1 Bg4 36.c7 e4 37.Nb3 Rb5 38.Rb8+
Ke7 39.Nd4 Rc5 40.Rb6 Bc8 41.Kf2 Rxc7 42.Ke3 Bb7 43.Nf5+ Kf8

44.Nd6

White is two pawns down, but his pieces are active. In particular, his king is better, and moreover there is not that much
material left on the board.

44...Rc3+

The simplification with 44...Rc6 would secure a ‘pawn-up’ position, but White safely reaches a draw with 45.Rxb7
Rxd6 46.Kxe4 and Black can do nothing with his extra material, as his king is isolated.

45.Ke2!?

After 45.Kd4 Rd3+ 46.Ke5 (in case of 46.Kc5 Bd5 47.Rxa6 e3 White’s active king will be cut off, and Black wins)
46...e3 (after 46...Bd5 47.Nxe4 Bxe4 48.Kxe4 Ra3 49.Rb7 White holds) 47.Rxb7 e2 48.Rb8+ Ke7 49.Re8+ Kd7
50.Kf4 Rxd6 51.Rxe2 a5 Black gets his rook behind the pawn and wins.
Or 45.Kf4 Rf3+ 46.Ke5 e3 47.Rxb7 e2 48.Kd4 (48.Kd5 loses to 48...Rd3+ 49.Kc5 Rxd6) and now:
analysis diagram

A) In the event of 48...e1=Q 49.Rb8+ Ke7 50.Re8+ Kxd6 51.Rxe1 Rf8 (with 51...Rf5 Black gets winning chances, but
it is far from over after 52.Re8 a5 53.Ra8) 52.Ra1 Ra8 53.Ra5 Kc6 54.Kc4 Kb6 55.Kb4 it is not clear whether Black
can win;
B) 48...a5!! – interestingly, White has no useful move now. By bringing the a-pawn one rank closer to promotion Black
wins, as all white pieces are tied to the e-pawn: 49.Rb8+ Ke7 50.Re8+ Kxd6 51.Rxe2 a4 52.Ra2 a3 53.Ke4 Rb3
54.Kd4 Ke6 55.Kc4 Rf3 and Black wins.

45...Bc6?
Black misses an interesting winning simplification: 45...Rc6! 46.Rxb7 Rxd6 47.Ke3 Re6 and now:
A) 48.g4?! allows Black’s rook to get behind his a-pawn: 48...Re7! 49.Rb4 Ra7 50.Kxe4 a5 51.Ra4 Ke7 52.Ke5 Ra8
and White is simply forced into zugzwang: 53.g5 Ra6 54.h4 Ra8 55.Kd5 (or 55.h5 Ra6 56.Kd5 h6 57.gxh6 Rxh6–+)
55...Kf7 56.Kc6 Kg6 57.Kb7 Rf8 58.Rxa5 Rf4 and Black wins;
B) 48.Ra7 h5 49.h3 Kg8 (Black can nicely cross the seventh rank and activate his king. In case of 49...g5 50.Rh7 White
holds) 50.Ra8+ Kh7 51.Ra7 Kg6 52.Rc7 Kf6 53.Ra7

analysis diagram

53...g6! (protecting all his pawns) 54.Rh7 (54.Rc7 Ke5 55.Rc5+ Kd6 56.Rc4 Kd5 also loses) 54...Ke5. Now Black just
transfers his king to the queenside and wins: 55.Rd7 Rb6 56.Re7+ Kd5 57.Rd7+ Kc4 58.Kxe4 a5 etc.

46.Rxa6 Ke7

After 46...g6 47.Ra7 Rc2+ (47...h5 48.Rf7+ Kg8 49.Rc7 g5 50.Re7 and White holds) 48.Ke3 Rxh2 49.Nxe4 Rh3
50.Ng5 White can avoid being two pawns down and draws.

47.Nf5+ Kf7

47...Kd8 creates more problems for Black.

analysis diagram

A) Cutting the king off with 48.Ra7 is not effective this time: 48...Bb5+ 49.Kd2 (after 49.Ke1 e3 White is in trouble)
49...Rd3+ 50.Ke1 g6 51.Nh4 Rb3;
B) 48.Ra5! – this is not at all a simple move to find. Would little Tolya have spotted it? We may never know. White
prevents the bishop from becoming strong and the rook has two not very obvious roles to play on a5. 48...g6 (in case of
48...Kd7 49.Ra7+ Ke6 50.Nxg7+ Ke5 51.Ra5+ Bd5 52.Nf5 White holds) 49.Nd4 Bb7 50.Ne6+! (White goes after the
e4-pawn, the tactics work for him) 50...Ke7 51.Nc5 (the rook has made the c5-square available for the knight and
makes another tactic possible) 51...Bd5 52.Nxe4! Bxe4 53.Re5+ and White draws.

48.Nd6+ Ke7 49.Nf5+ Ke6 50.Nd4+

Also possible is 50.Nxg7+ Ke5 51.Ra5+ Bd5 52.Nf5 which saves the game as well.

50...Kd5 51.Nxc6!

Rook endings give good drawing chances.

51...Rxc6
52.Ra7!

Now White defends without trouble.

52...Kd4 53.Rd7+! Ke5 54.Rxg7 Rc2+ 55.Ke3 Rxh2 56.Re7+ Kf5 57.Rxe4 Rh3 58.Rf4+ Kg5 59.Kf3 Rh6 60.Rg4+
Kf5 61.Rg7 Kf6 ½-½

GAME 3
Kirillov, Vladimir
Karpov, Anatoly
Zlatoust 1962

1.c4 e5 2.g3 g6 3.Bg2 Bg7 4.Nc3 Ne7 5.b3 0-0 6.Bb2 Nbc6 7.e3 f5 8.Nge2 Kh8 9.0-0 d6 10.d4 Bd7 11.d5 Nb4
12.a3 Na6 13.Rb1 Qe8 14.Qc2 g5 15.f4 gxf4 16.exf4 e4 17.Nd1 Qg6 18.Bxg7+ Qxg7 19.Qb2 Nc5 20.Qxg7+ Kxg7
21.Ne3 Kf6 22.b4 Na4 23.Rb3
23...a5!

The type of move which would later characterize Karpov as a great player. He intends to open the a-file and fights for
its control. 23...h5 was an alternative.

24.Rc1

Kirillov was an experienced player: he had played Botvinnik in 1931. He had lost to the first Soviet World Champion
without much resistance, yet he played at a reasonable standard in general. With 24.g4 White could have tried to
undermine Black’s pawn centre. After 24...axb4! (after 24...Rg8?! 25.Ng3 axb4 (25...fxg4 26.Nxe4+ Kg6 27.c5 gives
White a big advantage) 26.axb4 Rae8 27.Kf2 White is better) 25.axb4
analysis diagram

Now Black disrupts White’s pawn structure with 25...b5!, e.g.:

A) 26.cxb5 Bxb5 27.Re1 Bxe2 28.Rxe2 Nb6 29.Rd2 Ra1+ 30.Kf2 Rfa8 31.gxf5 R8a2 32.Rbb2 Rxb2 33.Rxb2 Nexd5
and Black has obtained an active position;
B) 26.gxf5 bxc4 (not 26...Nb6 27.c5 Nbxd5 (after 27...dxc5 28.Ng4+ Kf7 29.bxc5 Nbxd5 30.Bxe4 Ra2 31.Ne5+
White is also rather active) 28.cxd6 cxd6 29.Ng3 and Black has problems) 27.Nxc4 Nxd5! 28.Bxe4 Rae8! 29.Bxd5
Rxe2 and with his active play Black would equalize fully.

24...axb4 25.axb4 Nb6 26.Nc3 Ra7 27.c5 Nbc8

27...Na4 would involve much risk in view of the pin: 28.Ra1 Rfa8 (or 28...Nxc5 29.Rxa7 Nxb3 30.Rxb7 and Black is in
trouble) 29.Rba3 dxc5 30.bxc5 Ra6 31.c6 bxc6 32.Rxa4 Rxa4 33.Rxa4 Rxa4 34.Nxa4 cxd5 35.Nc5 Be6 36.Nc2 Nc6
37.Nb3 and the two pawns are no full compensation for the piece.

28.Bf1

After 28.g4 fxg4 29.Bxe4 h5 White is a little better.

28...c6 29.Rd1 Rd8

In the database the move indicated is 29...Ra8 but that would lose by force after 30.dxc6. It is almost certain that the text
was Karpov’s actual move. He was already too accomplished a player to commit a mistake like that.

30.dxc6 Bxc6

31.Ned5+?

An impatient move. Better was 31.Bc4!, maintaining his grip: 31...Ba4 (after first 31...h5 32.Kf2 Ba4 33.Nxa4 Rxa4
34.Rb2 Nc6 35.b5 Black is in trouble) 32.Nxa4 Rxa4 33.Rb2 Nc6 34.b5 (in the event of 34.Kf2 dxc5 35.Rxd8 Nxd8
36.bxc5 Ra5 Black escapes) 34...Na5 35.Be2 (35.Ba2 Ra3 36.Kf2 is testing as well) 35...Ke6 36.g4 and Black is
struggling to create harmony between his pieces.

31...Bxd5 32.Nxd5+ Nxd5 33.Rxd5

33...Ra1!

Another move that reveals signs of the future champion. Already at this very early age he goes for the pin, thus
restricting the movement of his opponent’s pieces.

34.Kf2 Ra2+ 35.Be2 Ke6! 36.Rd4 d5

Suddenly White is in big trouble. Black threatens to mobilize his two central pawns.

37.b5

After 37.Rb1 Ne7 38.Ke1, 38...Rda8 may be the simplest. With 38...Nc6 39.Rdd1 Rc2 Black would also obtain an
overwhelming position.

37...Rc2

Now that Black conquers the c-pawn, he is winning.

38.c6 bxc6 39.bxc6 Rxc6 40.Rb7 Ne7 41.g4

Or 41.Ra4 d4 42.Raa7 Nc8.

41...fxg4 42.Bxg4+ Kf6 43.Ra4 Ng6 44.Raa7 Nxf4 45.Kg3 Nd3 46.Rxh7
46...Ke5!

Another characteristic of the later endgame virtuoso can be seen in the excellent usage of his king. Admittedly, Black’s
position is completely winning and this move is a very natural one.

47.Rhe7+ Kd4 48.h4 Ne5 49.Be2 Rg6+

49...Rc3+! would almost force immediate resignation: on 50.Kf2 (50.Kg2 Rc2) 50...Nd3+ wins.

50.Kh3 Rdg8

Even if the game were played with 29...Ra8 instead of 29...Rd8, which is doubtful, now they transpose to the same
position. An alternative was 50...Nc6 51.Rg7 Rxg7 52.Rxg7 Ke3.

51.Ra4+
51...Nc4?!

Black allows his opponent to separate the passed pawns. With 51...Ke3! he could successfully pursue White’s king:
52.Rxe5 Rg3+ 53.Kh2 Rg2+ 54.Kh1

54...Kf4! 55.Re7 (55.Rxd5 Rg1+ 56.Kh2 R8g2+ 57.Kh3 Rxe2 and White gets checkmated; or 55.Rg5 R8xg5 56.hxg5
Rxe2) 55...Rg1+ 56.Kh2 R8g2+ 57.Kh3 Rxe2 58.Rf7+ Ke5 and Black wins.

52.Bxc4 dxc4 53.Rd7+ Kc3

Moving in the other direction with 53...Ke3 54.Rxc4 Rg3+ 55.Kh2 Rg2+ 56.Kh1 R2g4 would have given Black
excellent winning chances.
54.Rc7 Rg3+ 55.Kh2 Rg2+ 56.Kh1

56...R2g4

All available sources give the bizarre move 56...R2g7, but considering what happens next, it seems more logical that the
text move was actually played. This is also suggested by O’Connell, Levy and Adams in their 1976 book The Complete
Games of World Champion Anatoly Karpov.

57.Raxc4+ Kd3 58.Rc3+ Kd4 59.R7c4+ Ke5 60.Rc5+ Kf4 61.Rh5 Rg1+! 62.Kh2 R8g2+ 63.Kh3 Rd2 0-1

GAME 4
Kolishkin
Karpov, Anatoly
Zlatoust 1962

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Na5 10.Bc2 c5 11.d4 Qc7
12.Nbd2 Nc6 13.dxc5 dxc5 14.Nf1 Rd8 15.Qe2 Be6 16.Bg5 Rd7 17.Ne3 h6 18.Bxf6 Bxf6 19.Red1 Rad8 20.Rxd7
Rxd7 21.a4 c4 22.axb5 axb5 23.Nd2 Na5 24.Rb1 Qd8 25.Ndf1 Be7 26.Rd1 Bc5 27.Rxd7 Qxd7 28.Ng3 g6 29.h4 h5
30.Ngf1
30...b4!

Gaining space.

31.Nd2

Another option is 31.Qd1 Qa7 (the ending after 31...b3 32.Qxd7 Bxd7 33.Bd1 is tenable) 32.cxb4 Bxb4 33.Qd8+ and
White has more breathing space than before, but Black is still better.

31...b3!

This fixes the b2-pawn and also it is nice to have a pawn so close to the promotion square.

32.Bb1 Qb5

The immediate 32...Qa4 looks even more accurate.

33.Nd5 Qa4 34.Qd1

34...Kg7!

This is a typical Karpov move. It improves his king’s position and makes sure that White will not get random chances
against his king with Qh6.

35.Nf3 Bg4!

Karpov pins again, this time restricting the white knight’s movements.

36.Qd2 Qa1

Should one miss a chance to pin?

37.Qc1 f6

White is almost completely paralysed.

38.Nd2
38...Be2!

Improving the bishop and threatening to win its counterpart on b1 at some stage with ...Bd3.

39.Nc7 Kf7 40.Na6 Nb7

40...Bf8!? probably wins more quickly, e.g. 41.Qe1 (or 41.Nb4 Bh6) 41...Bh6 42.Qxe2 Qxb2.

41.Nc7

After 41.Nxc5 Nxc5 42.Qe1 Bd3 43.Qc1 Nxe4 Black also wins.

41...Nd6

White is lost now. Black’s pieces dominate beautifully.

42.Nd5 Bd3 43.Qd1

43...Nxe4

The rest of the game holds little interest. White should have resigned as he loses a piece.

44.Nxe4 Bxe4 45.Ne3 Qxb1 46.Qxb1 Bxb1 47.Nf1 Ke6 48.Nd2 Bc2 49.Kf1 Kf5 50.Ke2 g5 51.hxg5 fxg5 52.Nf3 g4
53.Ne1 h4 54.f3 g3 0-1
 GAME 5
Ziuliarkin
Karpov, Anatoly
Zlatoust 1963

1.d4 Nf6 2.c4 e6 3.Nc3 d5 4.Bg5 Nbd7 5.e3 Be7 6.Nf3 0-0 7.Qc2 b6 8.cxd5 exd5 9.Bb5 Bb7 10.Qf5 c6 11.Bd3 h6
12.Bf4 Bc8 13.Bb1 Bb4 14.Qc2 Ne4 15.0-0 Bxc3 16.bxc3 Ba6 17.Re1 f5 18.Ne5 Nxe5 19.Bxe5 Qh4 20.Rc1 Rf7
21.a4 Bc4 22.f3 Ng5 23.Qf2 Qxf2+ 24.Kxf2 Ne6 25.Ba2 Bxa2 26.Rxa2 c5

27.c4!

White wants to invade along the c-file. The position sharpens up and a single mistake may prove fatal. Both sides have
their pluses and minuses here.
A) With 27.Rb1 White had another way to press for an advantage: 27...Rc8 28.a5 bxa5 (in case of 28...Rc6 29.axb6
Black has problems with his queenside pawns. The e5-bishop is very strong as well) 29.Rb5 a4 30.Ra5 and Black has
an unattractive ending ahead;
B) Forcing a weakening of the b6-pawn was possible at once with 27.a5!?. After 27...Rc8 (27...b5 28.a6 makes the b5-
pawn vulnerable) 28.axb6 axb6 29.Rb1 Rc6 30.Ra8+ (with 30.c4 Nc7 31.Bxc7 White can gain a pawn) 30...Kh7 31.Rb8
cxd4 32.cxd4 Rc2+ 33.Kf1 White’s advantage is more telling with the minor pieces on the board.

27...cxd4 28.cxd5

White keeps an edge with 28.exd4 Rc8 29.Rac2 Rd7 30.Ke3 Kf8 31.cxd5 (31.g4!?; maybe 31.h4 is the most attractive,
as White can play some preparatory moves before taking on d5) 31...Rxc2 32.Rxc2 Rxd5 33.g4 fxg4 34.fxg4 Ra5
35.Rc4 Ke7 36.h4 and Black has to wait passively.

28...Nc5
29.Rd1!

In case of 29.Ra3 dxe3+ 30.Kxe3 Rd8 31.d6 Rfd7 Black gets rid of the d6-pawn.

29...dxe3+ 30.Kxe3 Re7

Although Black’s pawn is on g7, the endgame reminds us of the Grünfeld Defence, where Black often fights against the
passed d-pawn.

31.f4 Rd8
32.Kf3!

This is a strong move. It opens a second front, and sets up g7 as a target. If 32.Kd4 Ne4 33.Kc4 Red7 34.a5 Rc8+
35.Kb4 Rc5 36.axb6 axb6 37.d6 Kf7 38.Ra6 Rc2 Black is safe.
After 32.a5 Ne4 (with 32...Red7 33.axb6 axb6 34.Rb2 (34.Rad2 Kf7 ties White to the d5-pawn) 34...Rxd5 35.Rxd5
Rxd5 36.Rxb6 Nd7 Black also holds) 33.Ra3 (33.axb6 Nc3) 33...Red7 34.axb6 axb6 35.Rad3 Kf7 Black holds.

32...Red7 33.d6

Sacrificing the a-pawn with 33.Rad2 was no fun: 33...Nxa4 34.g4 fxg4+ 35.Kxg4 b5! (vacating the b6-square for the
knight) 36.d6 Nb6 37.Rb1 Nc4 38.Rd3 a6 and Black has decent winning chances.

33...Nb7 34.g4!

White keeps playing excellent moves. He opens up the kingside for his king and his rook.

34...fxg4+ 35.Kxg4 Kf7 36.f5 Rc8

After 36...g6 37.fxg6+ Kxg6 38.Rf2 White invades.

37.Rd4?

White should have gone after the g7-pawn with 37.Rg2!! – no hesitation. After 37...Rc4+ 38.Kh5 Kf8 he has 39.Rg6!,
which is not an easy move to foresee. After 39...Rc5 (39...Nc5 loses to 40.Rdg1 Rf7 41.f6) 40.Re6 Nd8 41.Re7 wins.
The alternative after 37.Rg2 is 37...Rc5 38.Kf4 Rc4+ 39.Kf3 Kf8 and now:
analysis diagram

40.Rg6! (rooks like to be active, and this is an imaginative way to involve the white rook) 40...Rxa4 41.Rdg1 Rf7 42.f6
and Black is in trouble.
This manoeuvre is remindful of a famous game between Efim Geller and Bobby Fischer, played one year earlier. The
great Russian grandmaster found an impressive way to improve his rook against the young American genius, who was
just 19 years old when he got this positional lesson.

Geller-Fischer
Curaçao 1962
22.Ra1!! Rd8 23.Ra4 Bf5 24.Rb4 Bc8 25.Rb6

And Geller went on to win in great style. I wonder if Karpov already knew this positional masterpiece.

37.Rd4? Rc5 38.Kf4

After 38.Re2 Na5 39.Bf4 Nc6 40.Rde4 a6, the game would probably end in a draw.

38...Na5 39.Rg2??

An unfortunate blunder which loses the game. It was probably made in time-trouble. Karpov was a very fast player in
his youth; giving his opponent little time to calculate may have contributed to the error.
Better was 39.Re2 Nc6 40.Rd3 Ra5 41.Bc3 Rxa4+ 42.Re4 Ra2, when White has compensation for the pawn.

39...Rxe5!

Black wins a pawn and creates winning chances.

40.Kxe5 Nc6+ 41.Kd5 Nxd4 42.Kxd4 Rxd6+ 43.Ke5?

This either loses another pawn or results in the exchange of rooks. More resilient was 43.Kc4, but this would ultimately
lose as well: 43...Rc6+ 44.Kb3 Rc5 45.Rd2 a5 46.Rd7+ (46.Rd6 b5) 46...Kf6 47.Rd6+ Kxf5 48.Rxb6 Kg4 49.Rb5 Rg5
and Black wins any pawn race.

43...Rc6! 44.Rd2 Rc5+ 45.Rd5

Throwing in the towel, but White was losing anyway.

45...Rxd5+ 46.Kxd5 a6 47.Kc6 b5 48.axb5 axb5 49.Kxb5 Kf6 50.Kc4 Kxf5 51.Kd3 Kg4 0-1

GAME 6
 Sazontiev, Sergey
Karpov, Anatoly
Vladimir 1964

1.d4 Nf6 2.Nf3 e6 3.Bg5 d5 4.c4 Be7 5.Nc3 0-0 6.e3 Nbd7 7.Bd3 Re8 8.0-0 Nf8 9.Ne5 c6 10.f4 N6d7 11.Bxe7 Qxe7
12.Rf3 f6 13.Ng4 Nb6 14.c5 Nbd7 15.Rg3 Kh8 16.Nf2 e5 17.Qh5 e4 18.Be2 g6 19.Qh6 b6 20.b4 a5 21.b5 Bb7
22.cxb6 Nxb6 23.bxc6 Bxc6 24.h4 Rec8 25.h5 Qg7 26.Qxg7+ Kxg7 27.hxg6 hxg6 28.Rb1

28...Rab8!

Karpov promptly turns his attention to the zone where he is stronger. Black is interested in exchanging pieces on the
queenside, where his superior forces will ensure an advantage. A simple comparison: three boys can beat one boy more
easily than sixteen boys can beat fourteen.
The exchange with 28...Na4? 29.Nxa4 Bxa4 would be too hasty because of the strong invasion 30.Rb7+ Nd7 (30...Bd7
31.Bh5) 31.Bh5! Rc1+ 32.Kh2 Rh8 33.Rxg6+ Kf8 34.g4 and White even takes over.

29.Ba6 Rc7 30.Nfd1 Nfd7!

Defending the b8-rook.

31.Rh3?!

The subtle move 31.Be2! would have kept White in the game. Then, Black has three main options:
A) 31...Kf7 32.Rh3 and the White rook enters the game;
B) 31...Ba4 32.Bh5 and now:
B1) 32...Kh6 33.Bxg6 Bc2 34.Ra1 and White is not worse;
B2) 32...g5 33.fxg5 f5 34.Nxa4 Nxa4 35.Rxb8 Nxb8 36.Rh3 (White creates play of his own on the kingside) 36...Nc3
(after 36...Rc2 37.Be8 Nb6 38.g6 White has some nasty threats) 37.Nxc3 Rxc3 38.Be8 and White may start to press
soon.
C) 31...Rh8 32.Kf2 Na4 33.Rb3 Ndb6 34.Rh3 Rhc8 35.Rh1 and White is doing much better than in the game.

31...Ba4!

He persists on the idea of exchanges, and nicely executes his plan. With 31...Na4 32.Nb5 (after 32.Rc1 Bb5 33.Bxb5
Rxb5 34.g4 Nxc3 35.Nxc3 Rb6 White really misses the second rook on the queenside) 32...Bxb5 33.Bxb5 Nab6 Black
would be better as well.

32.Rh2

White plans to lend support to the queenside in an original way. After 32.Nxa4 Nxa4 33.Rxb8 Nxb8 34.Bb5 Nb6
Black’s rook is about to invade.

32...Bxd1 33.Nxd1 Na4

33...Nc8 was very promising as well: 34.Rxb8 (34.Ra1 Nd6) 34...Nxb8 35.Bxc8 Rxc8 and Black is much better,
probably even winning.

34.Ra1
After 34.Bb5 Nc3 35.Nxc3 Rxc3 36.Rh3 a4 37.g4 a3 Black wins.

34...Nc3!

All according to plan; exchanging in order to penetrate White’s position.

35.g3 Nxd1 36.Rxd1 Rc3

It has taken Black some 11 moves to create his first direct threat.

37.Re1

37...a4!

Exemplary positional play again. This pawn prepares the arrival of the other rook at b2.

38.Rd2 Rb6 39.Bf1 Kf7

Black is improving his king’s position as well. The position is probably winning even without this move, but why not
complete all possible improvements first?

40.Ree2 a3 41.Kf2 Rb4 42.g4

42...Nb6!

Bringing the last non-playing piece into the game.

43.Rc2

Other moves lose as well. White’s situation is already hopeless.

43...Rxc2 44.Rxc2 Rb2 45.Re2 Na4! 46.Ke1 Nc3

As a reward for great positional play Karpov arrives at a picturesque position.

47.Rd2 Nxa2 0-1

GAME 7
Drizgalovich, Mikhail
Karpov, Anatoly
Tula 1965

1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.Nf3 0-0 5.e3 d5 6.cxd5 Nxd5 7.Nxd5 Qxd5 8.Qc2 Nc6 9.Bc4 Qd6 10.0-0 Bg4 11.Be2
Nb4 12.Qd2 c5 13.a3 Nc6 14.dxc5 Qxc5 15.b4 Qb6 16.Bb2 Rfd8 17.Qc2 Bf5 18.Qb3 Na5 19.Qa2 Be6 20.Qb1
Bxb2 21.Qxb2 Nc4 22.Bxc4 Bxc4 23.Rfd1 Bd5 24.Nd4 Qf6 25.Qe2

The position is obviously equal. However, Black shows signs of his intentions to test White’s endgame skill.
25...e5!

Removing the knight from its stronghold. The position is still equal, but this central advance is not pleasant for White.

26.Nf3

After 26.Nb5 Qa6 27.Nc3 White is a fraction worse, but he should be able to hold this endgame easily.

26...a6!!

This is a quite sneaky move. You would almost think that Black just moved the a-pawn with no particular intention.
27.Rac1?!

White already has to choose between several slightly inconvenient endgames. This move does not spoil anything, but
neither does it improve his chances.
A) Trying to double the rooks with 27.Rd2? seems like a good idea, but it actually loses control of this file. After
27...Bxf3 28.Rxd8+ Rxd8 29.Qxf3 Qxf3 30.gxf3 Rd3 White has a gloomy endgame with his passive rook on a1;
B) 27.e4 – this is close to equalizing fully, but after 27...Bc6 28.Qe3 Rd6! (not 28...Qf4 29.Qxf4 exf4 30.Ng5 h6
31.Nh3 Bxe4 32.Nxf4 and the bishop is stronger than the knight, but Black is unlikely to be able to exploit this)
29.Rxd6 Qxd6 30.Rc1 Rd8 31.h3 f6 Black is slightly better;
C) 27.Ne1 Rac8 (with 27...b5 28.e4 Bc4 29.Qe3 Rd6! Black can gain control of the d-file) 28.e4 Bc6 29.Qe3 Rd4 30.f3
Bb5 31.Rac1 Rxc1 32.Rxc1 Qd6 and with the d-file securely in his grip, Black can exert pressure;
D) 27.Nd2! – this is the right move as it prepares to block the bishop’s diagonal: 27...Rac8 and now:
D1) 28.e4 Be6 29.Rac1 (29.Qe3 Rc2 30.Nf3 (30.h3!?) 30...Rxd1+ 31.Rxd1 Bg4 and White is not in big trouble, but he
has a bit of problem here) 29...Bg4!? 30.f3 (after 30.Qxg4 Rxc1 31.Rxc1 Rxd2 32.f3 Kg7 White has to be careful)
30...Qb6+ 31.Kh1 Be6 32.Nf1 Rxc1 33.Rxc1 Qd4 and again this is not fully equal;
D2) 28.Rac1!? Qg5 (in case of 28...Rxc1 29.Rxc1 Qg5 30.f3 White is not really worse) 29.f3 and it is hard to improve
Black’s position.

27...b5 28.Rd2 Bc4

Now the point of the subtle 26...a6 becomes clear. Black has created an outpost on c4 for his bishop. If 28...Bxf3
29.Rxd8+ Rxd8 30.Qxf3 Qxf3 31.gxf3 Rd3 32.Rc5 f6 33.Rc6 Rxa3 34.Rxf6 Kg7 35.Rb6, White holds.

29.Rxd8+?

Mistakenly giving up the d-file. Better was 29.Qe1! e4 (or 29...Kg7 30.Rcd1 – this time White gains control of the d-
file and he is safe) 30.Rxd8+ (30.Nd4 Rd5) 30...Rxd8 31.Nd4 Rd5 (if 31...Qg5 32.a4) and Black can play on the
kingside with his heavy pieces. However, White is better than in the game, e.g. 32.a4! (opening the a-file for his rook)
32...Rg5 (after 32...Qg5 33.axb5 axb5 34.Ra1 Black cannot reinforce his attack on the kingside) 33.axb5 axb5 34.Ra1
and with his active rook White is safe.

29...Rxd8 30.Qb2

With 30.Qe1 White cannot exchange the other rook pair because of 30...Qd6.

30...Qd6!

Making use of the mating threat on the back rank, Black occupies the d-file and defends his e-pawn.

31.h3 f6!

White is prevented from generating play against the e5-pawn.

32.a4

At the end of the game the b-pawn will fall, but waiting absolutely passively was bad as well. Black would still build up
the pressure as he does in the game.

32...Kg7!

This is very typically Karpov. He devotes time to tiny improvements.

33.axb5 axb5 34.Ra1 Rd7

Allowing no counterplay on the seventh rank either.

35.Rc1?

This is unattractive, but it is not easy to find a useful alternative. Still, leaving the open file was not correct. One
possibility was 35.Qc3 when, after ...h5!? and some preparation, Black plays ...Qd3.

35...Qd3 36.Kh2?!
Provoking a tactical skirmish, but it was already hard to find anything better.

36...Ra7 37.Ra1

37...Rd7?!

Karpov’s tactical vision is not as sharp as it would grow later on. With 37...Rxa1! Black had a winning simplification:
38.Qxa1 Qd6! (a nice and slightly surprising double attack) 39.Qa7+ Bf7 40.Qc5 (White has no other way to avoid
losing material) 40...Qxc5 41.bxc5 and now:

analysis diagram
A) Please note that the pawn advance 41...b4 looks winning, but looks can be deceiving: 42.c6 (after 42.Nd2 b3 43.c6
b2 or 42.Ne1 e4 43.Nc2 b3 44.Na3 b2 45.c6 Kf8 Black wins) 42...b3 43.c7 Be6 44.Nd2 b2

analysis diagram

45.Kg1!! (quite incredibly, the white king will reach the b-pawn in time) 45...Kf7 46.Kf1 e4 (in case of 46...Ke7
47.Ke2 Kd7 48.Kd3 Kxc7 49.Kc2 White will survive) 47.Ke1 Ke7 48.c8=Q! (it is anything but easy to see this
surprise move in advance; 48.Nb1 would hold as well, whereas 48.Kd1?? would lose to 48...Bb3+! 49.Ke1 Kd7)
48...Bxc8 49.Kd1 and White can defend.
B) After 41...Bd5! 42.Nd2 e4 (also possible was 42...f5 43.f3 e4 44.Kg3 Kf6 45.fxe4 (45.h4 loses to 45...Ke5)
45...fxe4 46.Kf2 b4 47.Ke2 Ke6 48.Kd1 Kd7 and Black collects the pawn) 43.f3 (43.Kg1 Kf7 44.Kf1 Ke7 45.Ke2
Kd7 and again Black collects the lonely pawn) 43...exf3 44.gxf3 b4 (44...Kf7 may be even simpler) 45.e4 Be6 46.c6
Kf7 47.h4 (47.c7 Ke7) 47...Ke7 48.Kg3 Kd6 Black wins.

38.Ra3 Qd6 39.Kg1 h5!

Black gains space and prepares to open up White’s kingside.

40.Ra1 Qd3 41.Kh2 g5! 42.Rc1 Kh6!

Preventing knight sacrifices and backing up his own pawns.

43.Rc3 Qd6 44.Kg1 g4! 45.hxg4 hxg4 46.Ne1

46...Kg5!

A majestic way to use the king. Karpov improves all his pieces to their optimum potential.

47.Rc1 f5

47...Qd2 was also strong.

48.g3?

White prepares to close the h-file but opens the long diagonal.
The only chance to survive was to improve the rook with 48.Ra1. There follows 48...Rh7! (48...e4 49.g3!) 49.f3 and
White can resist after 49...g3 50.f4+ exf4 51.Nf3+ Kh5 when Black has to be careful with his king, although he is still
winning.
48...Qd5!

Karpov changes the direction of the invasion, as he has noticed an opportunity to go after the enemy king, made
possible by White’s previous move.

49.Ng2 Qe4!

Threatening to put the bishop on the long diagonal, which would be lethal for the white king.

50.Qc2 Qxc2

Black collects the pawn and does not look for checkmates. A professional approach.

51.Rxc2 Rd1+ 52.Kh2 Rb1 53.Rd2 Rxb4 54.Ne1 Rb1 55.Ng2 b4 56.Nh4 b3 57.Ng2 b2 0-1
International Debut (1966-1968)

In 1966, Karpov started to achieve some remarkable results and he had clearly reached a higher level. He won his
international debut tournament in Trinec and shared 4th-5th place in the All-Union schoolboys championship. Often his
games were decided in the endgame. He probably had solid assistance at adjournments, but his wins were obtained
mainly because of his understanding and his sporting qualities. He became the youngest Soviet master ever, which was
an incredible achievement.
In 1967, Karpov played in four significant events and at the start of 1968 he became the European Junior Champion. His
statistics show that he already won many long games and in most cases this means that the decision fell in the endgame.

GAME 8
Ravinsky, Grigory
Karpov, Anatoly
Leningrad 1966 (4)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Qe2 b5 7.Bb3 d6 8.c3 0-0 9.Rd1 Na5 10.Bc2 c5 11.d4 Qc7 12.h3
Re8 13.dxe5 dxe5 14.Nbd2 Rd8 15.Nf1 Rxd1 16.Qxd1 c4 17.Bg5 Be6 18.Qe2 Nb7 19.Ng3?! g6 20.Rd1 Rd8
21.Rxd8+ Bxd8 22.Qe3

The position is equal, and yet there is plenty of reason to play on.

22...Nd7!

The improvement of this knight and the exchange of the dark-squared bishops is a thematic concern in the main Ruy
Lopez.

23.Bxd8 Nxd8 24.Qh6?

This loses some tempi. It provokes ...f6, but that will be a very useful move for Black.
A) Hindering Black’s space-gaining manoeuvres with 24.b4 was better: 24...cxb3 25.axb3 b4 26.Ne2 and the position
is equal;
B) Even 24.a4 Nc5 25.axb5 axb5 26.Nf1 looks equal.

24...f6 25.Qe3 a5!

Starting to push White back on the queenside.

26.h4 b4 27.h5 Kg7!

This king move stops any invasion.

28.Bd1?!

Unfortunately, the bishop will not achieve much on d1. With 28.hxg6 White should be able to improve one of his other
pieces after 28...hxg6. For example, 29.Qd2, 29.Nd2, or even 29.Nf1 looks better than the text.

28...Nf7

Black nicely improves his pieces one by one.

29.hxg6 hxg6 30.Nd2

30...Nc5!

White has to reckon with ...Nd3 and ...Na4 now. Both knight sallies will be unpleasant.

31.b3

White finally stops Black gaining space on the queenside, but it doesn’t solve all his problems.

31...cxb3 32.axb3 Qb6 33.Ngf1?!

The knight does not really help keeping the queenside together from this square. White would have been worse anyway,
but 33.Ne2!? not only puts up tougher resistance, it may well produce a defendable position: 33...Nd6 (after 33...bxc3
34.Qxc3 Black has to work hard to break open the position, though obviously he is better) 34.f3 (here White has a
tough choice between several unsavoury options. If 34.Kf1 bxc3 35.Nxc3 Qb4 is troublesome for him. After 34.c4 a4
35.Nc1 it is hard to tell how much better Black is) 34...bxc3 35.Nxc3 Qb4 36.Na4 and White’s position doesn’t
collapse.

33...Nd6 34.c4 Qc6 35.Bc2?

Other moves were not very attractive either:

A) 35.f3 a4 36.bxa4 Bxc4 37.Nxc4 Nxc4 38.Qe2 Na5 is rather ugly for White;
B) 35.Be2!? (maybe defending the c4-pawn was best, as 35...a4 will allow the white queen some play: 36.bxa4 Nxa4
37.Qa7+ Bf7 and Black is still somewhat better. Alternatively, after 35...Ndxe4 36.Bf3 f5 37.Bxe4 White would hold,
but Black could also attempt to regroup with 35...Ndb7!?.

35...a4!

After this thrust, which clears the way for the b-pawn, there is probably no longer any way for White to save the game.

36.bxa4 Bxc4 37.Ng3

After 37.Nxc4 Nxc4 38.Qe2 Na3 39.Bd1 Nxa4 White is lost.

37...Bf7!

The bishop protects the king against checks from a7. White is simply lost now.

38.a5 Qb5 39.f4 Qxa5

Black takes the pawn, but the queen can return to the centre in good time.
40.fxe5 Qa1+ 41.Kf2 Qxe5 42.Nf3

42...Qb2!

The pin is the simplest method.

43.Qxc5 b3 44.Nd4

If 44.Ne1, 44...Nc4 wins.

44...bxc2 0-1

GAME 9
Karpov, Anatoly
Tsamriuk, Alexander
Leningrad 1967 (1)

1.e4 c5 2.Nc3 Nc6 3.g3 g6 4.Bg2 Bg7 5.Nge2 d6 6.0-0 Nf6 7.d3 0-0 8.h3 Bd7 9.Be3 Rb8 10.Qd2 b5 11.Nd1 Qa5
12.c3 Rfc8 13.Bh6 Bh8 14.g4 Ne8 15.f4 b4 16.f5 bxc3 17.bxc3 Ne5 18.Nf4 Qd8 19.Ne3 Nc7 20.Qf2 Qe8 21.Rad1
Rb6 22.h4 Rcb8 23.Rd2 Rb1 24.d4 cxd4 25.cxd4 Rxf1+ 26.Bxf1 Nc6 27.Ned5 Nxd5 28.Nxd5 Nb4 29.Nf4 Nc6
30.h5 Rb4 31.Ne2 a5 32.Rd3 Rc4 33.hxg6 fxg6 34.fxg6 hxg6 35.Rf3 Bg7 36.Bxg7 Kxg7 37.g5 Bg4 38.Rf4 Bxe2
39.Bxe2

In a position where both kings are slightly exposed and both sides have their pluses, Black becomes over-optimistic.
39...Rxd4??

There were two main alternatives.

Already better than the text was 39...Rc2!, when Black ‘kills two birds with one stone’ – i.e. the rook. It pins the bishop
and simultaneously occupies the second rank. Both effects would be representative of Karpov’s style in endgames.
A) Now, if 40.Kf1 e5 41.dxe5 Nxe5! (on 41...dxe5, 42.Rf6 Nd4 43.Qh4 wins) 42.Qa7+ (with the knight on e5 there is
no attack: 42.Qh2 Rxa2; or after 42.Rf6 Qe7, Black is somewhat better) 42...Qd7 and White will have to struggle in the
endgame;
B) White’s best reply would be 40.Rf3!, fuelling the attack by transferring to the h-file:
B1) After 40...e5 41.Rh3 White wins;
B2) 40...Rd2 41.Qh2! creates a very scary threat (incidentally, 41.Rh3? loses to 41...Nxd4, and 41.Qh4 Rxe2 42.Qh6+
Kg8 43.Rh3 (43.Rf8+ Qxf8 44.Qxg6+ is a perpetual) 43...Kf7 44.Rf3+ Kg8 45.Rh3 is a repetition): 41...Nxd4
42.Qh6+ Kg8 43.Bc4+ (this looks like a win) 43...d5! 44.Bxd5+ e6 45.Rf6 (45.Rh3 Ne2+ 46.Kf1 Ng3+ 47.Kg1 is a
draw)
analysis diagram

45...Ne2+! (this saves Black) 46.Kf2 Nf4+ 47.Ke3 Rd3+ 48.Kxf4 Qb8+ 49.e5 Qb4+ 50.Be4 Qd2+ 51.Kg4 Qd1+ and
Black has a perpetual;
B3) 40...Rxa2 41.Rh3 (41.Qh2 Nxd4 42.Qh6+ Kg8 43.Bc4+ d5 44.Bxd5+ e6 45.Rh3 Nf3+! 46.Kh1 Rh2+ leads to a
draw) 41...Qg8 42.Qh4 Rxe2 43.Rf3 Qe8 (43...Qh8 44.Rf7+) 44.Qh6+ Kg8 45.Rh3 Kf7 46.Rf3+ Kg8 and White again
has a choice between drawing by repetition or by a perpetual: 47.Rh3 Kf7 48.Rf3+.
The second possibility is 39...Rc1+!!, which sets White a tough task:
A) 40.Kg2?
analysis diagram

A1) In case of 40...e5 41.dxe5 Nxe5 (after 41...dxe5 42.Rf6 Rc2 43.Qh4 Ne7 44.Rf2 the position is balanced) 42.Qa7+
Nf7 (after 42...Qd7 43.Qxa5 Rc2 44.Qa6 White is a pawn up but also tied up – the position is probably equal) 43.Qxa5
Rc2 44.Kf1 Qe5 Black is safe;
A2) 40...Rc2! – now the rook kills three birds with one stone; it also attacks the king. 41.Kf1! (41.Rf3 Rxa2 42.Rh3
Qg8 43.Qh4 loses to 43...Nxd4) 41...e5 42.dxe5 Nxe5 and Black is somewhat better.;
B) 40.Bf1! – the bishop steps into a pin in order to clear the second rank: 40...e5 41.dxe5 Nxe5 (41...dxe5 42.Rf6 Nd4
43.Qh2 is dangerous) 42.Qd2 Qc6 with a balanced position.

40.Bb5!

If the bishop isn’t pinned, it can create a pin of its own. Black actually loses the knight now. The fact that Karpov
noticed this tactical possibility underlines what we already know. He had grown into such a great endgame player not
just by understanding positional finesses, but also through his exceptional tactical vision in endgames. His alertness
contributed significantly to his success.

40...Rd1+

After 40...Rb4 41.Bxc6 Qxc6 42.Rf7+ White mates in two.

41.Kg2 Rc1

Defending the knight, but in vain.

42.Qb2+ e5 43.Qxc1 exf4 44.Qxc6

White collects the fruit of the pin – the bruised knight.

44...Qe5 45.Qd7+ Kh8 46.Qe8+ 1-0

 GAME 10
Popov, Nikolay
Karpov, Anatoly
Leningrad 1967 (6)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.0-0 f6 6.d4 exd4 7.Nxd4 c5 8.Nb3 Qxd1 9.Rxd1 Bd6 10.Na5 b6 11.Nc4
Be7 12.Bf4 Be6 13.Ne3 Bd8 14.Nc3 h5 15.Ncd5 Rc8 16.h4 c6 17.Nc3 c4 18.a4 Ne7 19.Bd6 0-0 20.a5 b5 21.Bc5
Re8 22.Ne2 Ng6 23.g3 Bg4? 24.Nxg4 hxg4 25.Rd7! Rxe4 26.Nc3 Re5 27.Bb6 b4 28.Na4 Be7

29.Rad1?!

White should have gone after the a-pawn with 29.Ra7!. This gives rise to a double-edged position; however, White’s
passed pawn seems stronger than Black’s activity after 29...Re8 30.Rxa6 Re2 and now:
A) 31.Bc7 Rxc2 (31...c5? loses to 32.Rb6 Rxc2 33.a6) 32.Rxc6 c3 33.bxc3 b3 34.Rb1 Ba3 35.a6 b2 36.Nxb2 and White
is somewhat better;
B) 31.c3 Ne5 32.Kf1 Rc2 33.Bd4 bxc3 (or 33...Bf8 34.Kg2 c5 35.Bxe5 Rxe5 36.Rc6 Ree2 37.Rf1 and the a-pawn is
very dangerous) 34.bxc3 Nf3 35.Rxc6 Ba3 36.Rd1 Rd2 37.Ra1 Ra2 and Black at least draws;
C) 31.Rc1 White keeps the pawn on c2, denying the knight entrance:
C1) 31...Ne5 32.Kf1 Rd2 33.Be3 Rdd8 34.Nb6 and White’s distant passed pawn is about to cause problems for Black;
C2) 31...Bd6 32.Kf1 Rd2 33.Ra7.

29...Re2?

This is an unfortunate moment to occupy the second rank. 29...Bf8 was better.
30.c3?

There was a much better move. Maybe Karpov missed something in the line that starts with 30.Kf1!! Rxc2 (after
30...Re5 31.Ra7 Re8 32.Rxa6 Rd5 33.Rxd5 cxd5 34.Bc7 Rc8 35.Ra7 White’s a-pawn is decisive) 31.h5! Ne5 (31...c3
loses to 32.hxg6 cxb2 33.Rxe7) 32.Rxe7 Nd3 33.Re2 b3 34.Nc5 Nxb2 35.Rxc2 Nxd1 36.Rxc4 and White wins.

30...Bf8 31.cxb4 Ne5 32.Kf1

Here and on the next move, White should still have taken the a-pawn with 32.Ra7 Nd3 33.Rxa6.

32...Rc2 33.Rd8? Rxd8 34.Rxd8

34...Kf7!

Karpov steps away from the pin. His rook is now very strong indeed.

35.Rd1?!

White goes down without much resistance.

A) After 35.Bc5 Bxc5 36.bxc5 Nd3 37.Ra8 Rxf2+ 38.Kg1 Rf3 39.Kg2

analysis diagram

39...Nxb2!! would have won.

B) 35.Ra8 offers more practical chances, as Black might not find the best way to pursue the king.
With 35...Rc1+ (35...Bxb4! 36.Be3 (36.Rxa6 Rc1+ and Black mates) 36...Bxa5 37.Rxa6 Bd2! weakens the white king as
well) 36.Ke2 Bxb4 37.Ke3 Ke6 38.Kf4 (38.Rxa6 Kf5 catches White’s king) 38...Re1 White’s king would be in big
trouble.

35...Nd3 36.Bd4

After 36.b3 Bxb4 37.bxc4 Ne5 38.Nc5 Bxc5 39.Bxc5 Nxc4 Black’s extra pawn should be decisive.

36...Bxb4

Black is now winning, Karpov’s rook on the second rank is too powerful and he has an extra pawn to boot.

37.Nb6 c5 38.Bc3 Bxc3 39.bxc3 Rxf2+ 40.Kg1 Rf3 41.Kg2 Ne5 42.Na4 Rd3 43.Rc1 Rd2+ 44.Kf1 Ra2 45.Nxc5
Rxa5 46.Ne4 Ke6 47.Ke2 Nd3 48.Rb1 Re5 0-1

GAME 11
Lukin, Andrey
 Karpov, Anatoly
Moscow USSR Junior Ch sf 1967 (3)

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Bg5 h6 5.Bh4 c5 6.d5 d6 7.f3 0-0 8.e4 Re8 9.Bd3 Bxc3+ 10.bxc3 Nbd7 11.dxe6 fxe6
12.f4 Qc7 13.Nf3 b5 14.Bxf6 Nxf6 15.e5 dxe5 16.fxe5 Nd7 17.0-0 Bb7 18.Bc2 bxc4 19.Qe2 Bd5 20.Rae1 Qa5
21.Ng5 Qxc3 22.Bh7+ Kh8 23.Bb1 Re7 24.Qh5 Qd4+ 25.Kh1 Bxg2+ 26.Kxg2 Qd2+ 27.Kh1 Qxg5 28.Qf3 Rb8
29.Qe4 g6 30.Rg1 Qf5 31.Qe3 Rxb1 32.Rxb1 Kg7 33.Rbe1 Rf7 34.Qe2 Qd3 35.Rg3 Qd5+ 36.Qg2 Qxg2+ 37.Kxg2
Rf5 38.Ra3 Nxe5 39.Rxa7+ Kf6 40.Rf1 Nd3 41.Ra4 Nb2 42.Ra5 Nd3 43.Ra4

Black has three pawns for the exchange and a strong knight.

43...Ke5!

In endgames it is important to use the king well. Karpov’s king becomes a dominating force. With 43...c3 44.Rc4 Ke5!
45.Rxc3 Kd4 Black could also have won.

44.Rxc4 Kd5 45.Rc2 Rg5+ 46.Kf3 Kd4 47.h4 Re5 48.Rd1 c4 49.a4 Rf5+ 50.Kg3 Ra5 51.Ra2 c3 52.Rf1 Nc5 53.Rf4+
Kd3 54.Ra3 Rxa4!

This is a cute move, although 54...e5 55.Rf3+ Kd4 wins as well.

55.Rf3+ Kd2 56.Rf2+

After 56.Raxc3 Ne4+ 57.Kh3 Nxc3 wins.

56...Kd1 57.Rf1+ Kc2

58.Rc1+

Alternatively, 58.Rxa4 Nxa4

A) If 59.Rf6 Kb2 (the most instructive, though 59...Nc5 60.Rxg6 Kb1 61.Kf2 c2 is simpler) 60.Rxe6 c2 61.Re1
(61.Re2 loses to 61...Nc3 62.Rh2 Kb1 63.Rxc2 Kxc2 64.Kf4 Kd3 65.Ke5 Ne4) 61...c1=Q 62.Rxc1 Kxc1 63.Kf4 Nc5
64.Ke5 g5 65.hxg5 hxg5 66.Kf6 Ne6 Black wins;
B) 59.Kf4 Kb2 60.Ke5 c2 61.Kxe6 Nc3 62.Rf2 (in case of 62.Kf6 Nd1 Black gets a queen) 62...Kb1 63.Rxc2 Kxc2
64.Kf6 g5 65.hxg5

analysis diagram
65...Ne4+ 66.Kf5 Nxg5 67.Kg6 Nf7 and Black wins.

58...Kb2 0-1

Vujakovic was a strong Yugoslav junior player who stopped playing chess in 1969 for a period of 17 years. This was
their second game from a national team match. Karpov had won the first.

GAME 12
Vujakovic, Branko
Karpov, Anatoly
Sochi USSR-Yugoslavia 1968 (2)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.d4 exd4 6.0-0 Be7 7.e5 Ne4 8.Nxd4 Nxd4 9.Qxd4 Nc5 10.Nc3 0-0 11.Be3
Nxa4 12.Qxa4 d5 13.exd6 Bxd6 14.Bf4 Bxf4 15.Qxf4 Be6 16.Rfe1 Qd7 17.Rad1 Qc6 18.Rd3

18...Rae8

Black is behind in development, so it is natural that he wants to catch up.

19.Rde3 f6

Black is very cautious as he has no pieces in the vicinity of his king. The text ensures that the g7-pawn can be defended
along the seventh rank.

20.h4

White is ambitious. He wants to gain benefit from his better development. With 20.Qe4 he could have simplified to a
draw. After 20...Qxe4 21.Nxe4 Kf7 22.f3 the position is equal.

20...Bd7!
Karpov nicely neutralizes White’s pluses.

21.Qd4 Rxe3 22.Rxe3 Be6

23.h5?!

White is still trying to put pressure on the black position, but this is not justified as the bishop’s slight superiority over
the knight gives Black an equal game. The more pieces come off the board, the more one should avoid placing pawns
on the colour of the opponent’s bishop. In this game it just becomes a headache to defend this pawn.

23...h6 24.Rg3

Maybe White should have considered restructuring his pawns with 24.a3.
24...Rf7!

This is another slight improvement of the position.

25.Qd8+ Kh7

Now he plays for a win.

26.Qd3+ f5 27.Rg6

The rook moves too far away from its camp. White’s sense of danger lets him down.

27...Qd7! 28.Qxd7 Bxd7 29.Nd5?

Still moving forward and this is a bad mistake. After 29.Rg3 f4 30.Rd3 Bg4 31.Rd5 f3 Black is slightly better.

29...f4!

Karpov feels fine – after defending for some time, now it’s time to counter. His first blow hits White right in the face.

30.g3?

A) 30.g4 would have been bad for another reason: 30...Bc6! 31.Nc3 Rd7 and White is lost;
B) 30.a3 Kg8!! (this prepares ...Bf5, which would not be possible immediately on account of 31.Nxf4. Now we can see
how unfortunate the game continuation was, as it deprives the white rook of the g3-square) 31.g4 (31.Nxc7 Bf5 wins)
31...fxg3 32.Rxg3 Be6 and Black is clearly better.

30...fxg3?!

Karpov probably saw that the text move wins and missed a very nice and simpler win.
Instead of 30...fxg3, 30...Bf5!! traps the rook. 31.Nxf4 – or does it? 31...Rd7 – yes it does, Black has to resign.

31.Rxg3 Rf5 32.Nxc7

After 32.c4 Rxh5 33.Nxc7 Rc5 Black wins.

32...Rc5 33.Rc3

33.Na8 would be equally hopeless: after 33...Rxc2 34.Rb3 Bc6 it may take a little time, but Black will surely win.

33...Rxc3 34.bxc3 Bg4 35.Kg2?!

This protects the b7-pawn! But White is lost anyway.

35...Bxh5 36.Ne6?

If 36.Nd5 Bf7 37.c4 Kg6 (bringing in the king is more convincing than winning the pawn with 37...b5 38.Nc7 Bxc4
39.Nxa6 Bxa2, although Black should win this without any problem as well) 38.Kf3 Kf5 Black wins.

36...Bf7 37.Nc5 Bxa2 0-1

GAME 13
Karpov, Anatoly
Sukhanov, V.
Moscow University Ch. 1968/69 (7)

1.e4 c5 2.Nc3 Nc6 3.g3 g6 4.Bg2 Bg7 5.d3 Nf6 6.Be3 d6 7.h3 Nd4 8.Nce2 Qb6 9.c3 Nxe2 10.Qxe2 Bd7 11.Nf3
Qa6 12.Nh4 Bc6 13.Qc2 d5 14.e5 Nd7 15.f4 Bb5 16.Bxd5 Bxd3 17.Qg2 0-0-0 18.Rd1 c4 19.b3 e6 20.Bxc4 Bxc4
21.bxc4 Qxc4 22.Qe2 Qxe2+ 23.Kxe2 f6 24.exf6 Bxf6 25.Nf3 e5 26.fxe5 Nxe5
27.Bxa7

White has to take this pawn if he wants to play on.

27...Rde8

Black’s position is good enough to draw, as the white king has no shelter and his pawns are separated. However, he has
to find counterplay to compensate for the pawn deficit.

28.Nxe5

White can’t consolidate the extra pawn with 28.Bd4 because after 28...Nxf3+ 29.Kxf3 Rhf8 (sometimes the margin is
small between a strong king in the centre and a vulnerable one caught in the crossfire) 30.Kg4?? – this would make the
king a tragic figure after 30...h5+ 31.Kf3 Bxd4+ 32.Kg2 Re2+.

28...Bxe5

Going after the a-pawn with 28...Rxe5+ was good enough: 29.Kd3 Ra5 30.Bd4 Bxd4 31.cxd4 Rxa2 (after 31...Ra3+
32.Kc4 Rxg3 Black is also safe) 32.Ra1 Rxa1 33.Rxa1 Kc7 and Black can draw fairly easily.

29.Kd3 Bxg3 30.Rhf1

The rook naturally moves to an open file.

30...Re6?

This is a careless move. It is hard to understand why Black played it. It would be very naive to think that White would
allow him to trap the bishop. Black could have reached comfortable equality by contesting the f-file with 30...Rhf8!.
After 31.Bc5 Rxf1 32.Rxf1 Re5 33.Bd4 Rf5 (after 33...Rh5 34.Rf3 Bd6 Black has no problems either) White would
soon have to settle for the draw.

31.Bd4 Rd8 32.Rf7 h5

Better was 32...Rd7!, neutralizing the rook on the seventh rank: 33.Rdf1 g5 (after 33...Ree7 34.Rxe7 Rxe7 35.Bc5 Rd7+
Black has nothing to worry about either) 34.Rxd7 Kxd7 35.Rf7+ Re7 36.Rf5 Bf4 and Black cannot lose.

33.Rb1 b6

In case of 33...Ra6 34.Rbxb7 Rxa2 35.Kc4 it would be threatening for Black to see the opponent’s king positioned so
strongly.

34.a4 Bc7

35.Kc4

The king has become a strong piece.

35...Rc6+

Chasing away the white rook from the seventh rank at once with 35...Rd7 was possible:
A) 36.Rbf1 g5!? (Black can consider giving up his two kingside pawns for the white h-pawn. When pawns are on one
and the same side, many positions can be defended with relative ease despite a material deficit. Karpov would later
hold a rook endgame with 3 pawns against 4 against Kortchnoi) 37.R1f5 Rxf7 38.Rxf7 g4 39.hxg4 hxg4 40.Rg7 g3 and
Black is safe even if the g-pawn falls;
B) 36.Rf8+ Rd8 37.Rbf1 Rxf8 38.Rxf8+ Kb7 39.Kb5
analysis diagram

39...g5! (39...Re1 40.Rf7 Rb1+ 41.Kc4 Re1 42.Kd5 is unpleasant) 40.Rg8 g4 41.hxg4 hxg4 42.Rxg4 Re1 (with 42...Rh6
43.Rg5 Rh1 Black also holds) 43.Rg1 Rxg1 (simplest) 44.Bxg1 Bd8 and Black draws.

36.Kb5!

Karpov’s pressure is getting more and more annoying.

36...Kb7 37.Re1 Rd5+

It was a little late for 37...g5 now: 38.Ree7 Rd5+ 39.Kb4 Kc8 40.Re8+ Rd8 (or 40...Kb7 41.Be5 Rdc5 42.Kb3 and
Black is tied up) 41.Rxd8+ Bxd8 42.Rg7. As White’s king is active now, Black cannot easily reach a drawn endgame
by exchanging the kingside pawns: 42...g4 (42...Rd6 43.Kb5) 43.hxg4 hxg4 44.Rxg4 Kb7 45.Kb5 (45.Rg7+ Rc7)
45...Re6 46.Rg7+ Bc7 (notably, the endgame after 46...Re7 47.Rxe7+ Bxe7 48.Bxb6 is winning for White even with the
wrongly-coloured bishop and the rook pawn!) 47.Bf2! (47.c4? Re1!) 47...Rf6 48.Bg3 Rf5+ 49.Kb4 Rc5 50.a5 and
White wins.

38.Kb4 Rf5 39.Rxf5 gxf5 40.Kb5 Rd6 41.Kc4!

After 41.Re5, 41...Rxd4! 42.cxd4 Bxe5 43.dxe5 f4 44.Kc4 Kc6 45.Kd4 f3 46.Ke3 Kd5 draws.

41...Rg6

Passively defending the pawn with 41...Kc6 was an option, but it is hard, if possible at all, to defend: 42.Re5 Rf6 43.Re8
Rh6 (or 43...Rf7 44.Rh8 f4 45.Rh6+ Kb7 46.Rxh5 f3 47.Bf2 and White has good winning chances) 44.Rf8 f4 45.h4 Bd6
46.Rf5 Bc7 47.Kd3 (also after 47.Bf6 Kd7 48.Kd5 White is better) 47...Kd7 48.Rf7+ Kc6 49.Bf6
analysis diagram

... and Black is very passive.

42.Re7

Karpov pins the bishop on the queenside, but he is actually considering kingside operations.

42...Kc6 43.Rh7 f4 44.Rxh5

Collecting the h-pawn. Now Black pays a high price for failing to destroy White’s h-pawn when he had the opportunity.

44...f3 45.Kd3 Rg2 46.Rf5 Ra2 47.Rxf3 Rxa4

48.h4

White’s h-pawn is the decisive factor. Look how strong his bishop is on d4.

48...Ra5 49.Rf6+

Sometimes general principles do not work. In case of 49.Rh3?! the rook is not so effective after 49...Rh5 50.Ke4 Bd8
and Black can resist for longer than in the game.

49...Kb7 50.Rh6 Rf5

51.Ke4!

With the help of the king, the h-pawn is unstoppable.

51...Rf1 52.Kd5 Rf5+ 53.Ke6 Rf3 54.h5 Kc6

Interestingly, Black sets up a battery for White, but this is the only way for him to seek counterplay.

55.Rf6 Rh3 56.h6 Bd8 57.Rg6 b5 58.Kf7+ Kd5 59.Kg7 b4 60.cxb4!

The easiest road to victory.

60...Kxd4 61.Rd6+ Ke5 62.Rxd8

This is now an elementary win for White.

62...Rg3+

One of the motivations for publishing this book was to help juniors learn from Karpov’s endgame masterpieces. If
White had no b-pawn, this position would be a draw and this check would be Black’s only move.

63.Kh8 Kf6 64.h7 Kf7 65.Rd1 Rb3 66.Rf1+ Ke7 67.Rf4 1-0

GAME 14
Karpov, Anatoly
Romanishin, Oleg
Riga USSR Team Ch. 1968 (9)

1.d4 d5 2.Nc3 Nf6 3.Bg5 c6 4.Nf3 Nbd7 5.e3 g6 6.Bd3 Bg7 7.0-0 0-0 8.Re1 Re8 9.h3 Qb6 10.Rb1 e5 11.Be2 Ne4
12.Bh4 exd4 13.Nxe4 dxe4 14.Nxd4 Ne5 15.c4 c5 16.Nb5 Be6 17.Qa4 Qc6 18.Rec1 h6 19.Rc2 g5 20.Bg3 Red8
21.b4 cxb4 22.Qxb4 Nd3 23.Qa3 a6 24.Nc7 Rac8 25.Qa5 b5 26.Nxe6 fxe6 27.Bg4 Ne5 28.Bxe5 Bxe5 29.Rbc1 Rd6
30.g3 Qb6 31.Qe1 bxc4 32.Rxc4 Rxc4 33.Rxc4 Qb7 34.Qc1 Rb6 35.Rc8+ Kg7 36.Qc5 Bd6 37.Qc3+ e5

38.Kg2!

This is a typical Karpov move as it prevents any black checks. It is also unpleasant because Black now has to work out
the nature of any concealed white threats.
Interestingly, Garry Kasparov himself appears to have learned from Karpov simply by playing him so often. In a famous
rapid game with Kramnik, the 13th World Champion employed a similar move.

Kasparov-Kramnik
Frankfurt rapid 1999 (8)
33.Kb2!

Preventing all counterchances with ...Bxb3 before launching his attack.

33...Qd8 34.Qf4 Qc7 35.Rd5 exd5 36.Nd4 1-0

Back to Karpov-Romanishin.

38.Kg2! Rb1?!

After 38...Qd5 39.Rd8 Qc5 40.Qd2 Black is struggling.

39.a4 Qf7

39...Rb6 40.Qc4 was also hopeless.

40.Rd8 1-0
Chapter 2
Rise to the Top

At the end of the 1960s, Karpov became a contender for the Junior World Championship, which he was to win in 1969.
1970 was the year that Bobby Fischer started his magical three-year run, which must have had a strong effect on
Karpov’s career. Starting in 1969, he worked hard with his trainer Semen Furman and became the greatest young
promise and a ‘superstar’ in the Soviet Union. Karpov started to compete in the semi-finals and the extremely strong
finals of the USSR championships. His shared win with Stein at the Alekhine Memorial in Moscow, 1971, ahead of
four World Champions, was a tremendous feat. This was the first time that Karpov showed the level of a World
Champion.
In 1973 Karpov set out on his quest for the world crown, qualifying in the Leningrad Interzonal, beating Lev
Polugaevsky in the Candidates’ quarterfinal, Boris Spassky in the semi-final and, finally, Viktor Kortchnoi in the final,
thus becoming the challenger of Fischer.

World Class Results (1969-1972)

The year 1969 was a very successful one for Karpov, who won two individual tournaments and qualified for the Junior
World Championship. Game 19, against Eugenio Torre, was crucial for this qualification. His concept against
Rashkovsky from the RSFSR Championship in Kuibyshev (Game 23) is also very deep and very long.
Rafael Vaganian was Karpov’s closest rival as a junior. Vaganian has very original ideas over the board as a player and
was predestined for great successes. But Karpov’s superiority is reflected in their lifetime record so far: 6:2 with 14
draws.

GAME 15
Vaganian, Rafael
Karpov, Anatoly
Leningrad Qualifier1969 (1)

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.e3 Bb7 5.Bd3 Be7 6.0-0 c5 7.Nbd2 Nc6 8.a3 cxd4 9.exd4 d5 10.cxd5 Qxd5 11.Bc4
Qd6 12.Nb3 0-0 13.Qe2 h6 14.Rd1 Rfd8 15.Be3 Rac8 16.Rac1 Nd5 17.Bd2 a5 18.Rc2 Qb8 19.Nc1 Nf4 20.Bxf4
Qxf4 21.Rcd2 Bf6 22.Nd3 Qc7 23.Nde5 Nxd4 24.Rxd4 Rxd4 25.Nxd4 Qxe5 26.Nb5 Qxb2 27.Nd6 Qxe2 28.Bxe2
Rc7 29.Nxb7 Rxb7 30.a4 Rc7 31.Bb5 Rc2 32.h3 g6 33.Rd6 Rc1+ 34.Bf1 Ra1 35.Rxb6 Rxa4 36.g3 Bd4 37.Rb8+ Kg7
38.Be2 Ra2 39.Kf1 a4 40.Rb4 e5 41.Rb7 a3 42.Rxf7+ Kh8 43.Rf8+ Kg7 44.Rf7+ Kxf7 45.Bc4+ Kf6 46.Bxa2 g5

Fixing the pawn on f2 and keeping the white king in the defence with 46...e4 was a winning plan, as White is
defenceless against the king march to b2.

47.f3 h5 48.Ke2

After 48.g4 h4 49.Ke2

analysis diagram

49...e4! opens the passage for the invasion. In opposite-coloured bishop endings you often have to play aggressively.
50.fxe4 Ke5! (White must go one way with the king, then the Black king invades in the other direction. Just like taking
a penalty in soccer) 51.Kf3 (on 51.Kd3 Bb2 52.Bb3 Kf4 53.Ba2 Kg3 54.Ke2 Kxh3 55.Kf3 Kh2 wins) 51...Bb6
52.Bb3 Kd4 53.Ba2 Bc7.

48...e4!!

Black has to clear the way for his king. The white e-pawn will have no value.

49.fxe4
After 49.g4 hxg4 (49...h4 wins as well) 50.hxg4 e3 Black has the two passed pawns necessary for the win: 51.Kd3 Ke5
52.Bb3 Kd6 53.Ba2 Kc5 54.Bg8 Kb4 55.Ba2 Be5 56.Bf7 e2 57.Kxe2 Kc3 and Black breaks through.

49...g4!

Black fixes the pawn on g3, which becomes the key factor.

50.hxg4 hxg4 51.Bb3

51...Be5!

Black ties the king to the defence of the g3-pawn.

52.Kf2 Ke7 53.Ke3

White doesn’t wait till the black king wins the bishop, but gives up the pawn. Sadly for him it’s all in vain.

53...Bxg3 54.Kd4 Be5+ 0-1

GAME 16
Steinberg, Mikhail
Karpov, Anatoly
Leningrad Qualifier 1969 (8)

1.c4 e6 2.g3 Nf6 3.Bg2 c5 4.Nf3 Nc6 5.0-0 d5 6.cxd5 Nxd5 7.d4 Be7 8.dxc5 Bxc5 9.Ng5 Be7 10.Ne4 0-0 11.Nbc3
Nxc3 12.Nxc3 Qb6 13.Rb1 Rd8 14.Qa4 Bd7 15.Be3 Qb4 16.Qxb4 Bxb4 17.Ne4 Be8 18.a3 Be7 19.Rfd1 b6 20.Bf4
Rxd1+ 21.Rxd1 Rd8 22.Rxd8 Nxd8 23.Bc7
23...Bc6!

Karpov was leading the event and so he was content with a draw. Nevertheless this is a very lively move.

24.b4 b5 25.Bb8

Steinberg had to play for a win, which is why he did not force a draw with 25.Bxd8 Bxd8 26.Nf6+ Bxf6 27.Bxc6 Bb2
28.a4.

25...f5 26.Bd6?!

Better was 26.Nd6 Bxg2 27.Kxg2 Nc6 which, however, would leave White with very little to play for.

26...Bxe4 27.Bxe7
27...Nc6!!

Probably at this point Karpov started to think about actually winning. This is a very good quality of his. Not only does
he make very strong moves, but he also controls the fight and can quickly adapt to new situations.
The principle for Black here is to build up his position rather than take the opponent’s bishop and develop his king. This
often constitutes just a slight difference, but out of such small gains one can sometimes build enough to win. 27...Bxg2
would have secured a draw after 28.Kxg2 Nc6 29.Bc5 Kf7 (29...a6 30.f4) 30.f4 and both sides have no route by which
to penetrate.

28.Bxe4?

This looks like a small inaccuracy, as he takes voluntarily instead of building up his own position. White should try to
create a fortress as well. After 28.Bc5! a6 29.f4! the black knight is kept away from c4. The game should end in a draw.

28...fxe4

Postponing the exchange had a price; Black has to accept the doubled pawns.

29.Bc5?!

White should have given up a pawn with 29.Bd6! after which the position would be still equal: 29...Nd4 (in case of
29...a6 30.e3 White can keep the knight away from c4, and after 29...Kf7 30.Kf1 White is also safe) 30.Be5! Nxe2+
31.Kf1 Nc1 32.Bb2 Nd3 33.Bd4 a6 34.Ke2 and White is not worse despite the pawn deficit.

29...a6 30.Kg2

A) 30.Kf1 at once would have been better;

B) White cannot control the key e5-square with 30.f4? because of 30...exf3 31.exf3 Ne5 32.Kf2 Nc4 33.f4 Nxa3
34.Kf3 Nc4 and he is a pawn down;
C) 30.Bd6! still holds: 30...Nd4 31.Be5 Nxe2+ 32.Kf1 Nc1 33.Bb2 Nd3 34.Bd4 followed by Ke2 and White is in the
game.

30...Ne5!

Now Black’s domination becomes obvious. He has accomplished a great deal by visualizing the idea of putting the
knight on c4.

31.Be3 Kf7 32.Bc1 Nc4 33.Kf1

There are two main alternatives, both of which favour Black.

A) 33.f4 g6 34.g4 e5 35.e3 exf4 36.exf4 Ke6 37.h3 Kd5 38.Kf2 h5 39.Ke2 hxg4 40.hxg4 Kd4 and White is in
zugzwang;
B) 33.f3 Kf6 34.fxe4 Ke5 35.Kf3 Kd4 36.Kf4 (or 36.h4 Kc3 37.g4 Kb3 38.Kf4 Nxa3 39.Bd2 e5+) 36...Kc3 37.h4
Kc2 38.Be3 e5+ 39.Kf3 Nxe3 40.Kxe3 Kb3 and Black wins.

33...Kf6 34.Ke1 Kf5 35.h3 h5 36.Kd1 g5 37.f3

There is not much hope either after 37.Kc2 e3 38.Bxe3 (38.f3 h4 and Black gets the h3-pawn, or 38.fxe3 Ke4)
38...Nxe3+ 39.fxe3 Ke4 40.Kd2 e5 41.g4 (41.Kd1 Kxe3 42.Ke1 e4 or 41.h4 g4 42.Ke1 Kxe3) 41...h4 42.Kd1 Kxe3
43.Ke1 e4 44.Kd1 (44.Kf1 loses after 44...Kf4 45.Kf2 e3+ 46.Kg2 Ke4 47.Kf1 Kd4 48.Kg2 Kc3 wins) 44...Kf4
45.Kd2 Kg3 46.Ke3 Kxh3 47.Kxe4 Kg2 and Black wins.

37...e3 38.g4+

After 38.Kc2 h4 Black gets the h3-pawn and wins.

38...hxg4 39.hxg4+ Ke5 40.Kc2

40...Kd4

According to the database 40...Kd5 – a horrible mistake – was played. This was probably an error when the game was
entered into the computer. At this age Karpov was such a good player, he would not have done something like this.
With 41.Kd3 White would at least be back in the game.

41.Kb3

White resigned, as 41...e5 followed by ...e4 will win very easily.

GAME 17
Karpov, Anatoly
Evrosimovski
Moscow USSR-Yugoslavia U-18 1969 (3)

1.c4 f5 2.Nc3 Nf6 3.g3 e5 4.Bg2 Nc6 5.Nf3 Be7 6.d4 e4 7.Ng5 0-0 8.0-0 Qe8 9.c5 h6 10.Nh3 d5 11.Rb1 Nd8 12.b4
c6 13.Bf4 g5 14.Be5 Qh5 15.f4 g4 16.Nf2 Qg6 17.Qb3 h5 18.Bxf6 Bxf6 19.e3 h4 20.Rfc1 Kg7 21.b5 Rh8 22.a4 Qh5
23.Ne2 Ne6 24.Nh1 Rh7 25.Kf2 Bd7 26.Qa2 Qf7 27.Rg1 h3 28.Bf1 Kh6 29.b6 axb6 30.Rxb6 Ra7 31.Qb3 Bc8
32.Qb4!

White protects the a5-square and also prepares to astonish his opponent.

32...Bd8
33.a5!!

Karpov builds a fortress with the aid of an exchange sacrifice. Fortresses are easy to visualize, once you conceive of
them. With all the pawns on one side they tend to occur much more often.

33...Bxb6 34.axb6! Ra2 35.Ke1

Black vainly tried to win for another 36 moves. Karpov’s fortress withstood the pressure and the game was drawn.

GAME 18
Karasev, Vladimir
Karpov, Anatoly
Leningrad Armed Forces tt 1969 (7)

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 0-0 5.Ne2 d5 6.a3 Be7 7.cxd5 Nxd5 8.Qc2 b6 9.Nxd5 exd5 10.Nc3 Bb7 11.Bd3 h6
12.0-0 Nd7 13.b4 a6 14.Qb3 Nf6 15.f3 Rb8 16.Bb1 c5 17.Rd1 Qd6 18.bxc5 bxc5 19.dxc5 Qxc5 20.Ne2 Qc4!?
21.Nd4 Qxb3 22.Nxb3
22...Bc6!

A subtle move. The bishop will exert pressure from the edge of the board. This is an unusual realignment.

23.Nd4 Ba4 24.Rd2 Nd7!

Improving the knight as well, and now we can see the justification of the exchange of queens: Black hopes to gain
access to the c4-square. Without b-pawns this idea has even greater merit.

25.Bc2

It is hard to choose between 25.Ba2 and the text. Both are unpleasant to play. After 25...Nb6 26.Bb2 g6 27.Rc1 Rfc8
Black would be somewhat better.

25...Bxc2 26.Rxc2 Bf6 27.Bb2

After 27.Rc6!? Rb6 (in case of 27...a5 28.Bd2 a4 29.Bb4 White’s pieces are active enough) 28.Rxb6 Nxb6 29.Rb1 Nc4
30.Kf2 White should not lose.

27...Rfe8 28.Kf2
28...Ne5!

The knight is getting closer to the soft spot c4.

29.Rd1 Rec8

Black exchanges in order to have better prospects to invade. In the event of 29...Nc4 30.Bc1 Rb1 31.Rd3, White is just a
fraction worse.

30.Rxc8+ Rxc8 31.Ke2 Nc4

It has taken Black 11 moves and some effort to plant the knight on this square. It will grow very, very tall here.

32.Bc1 Nb6 33.Bb2

33.Kd3 Na4 ties White up.

33...Na4 34.Ba1 g6 35.Kd2 Be7 36.Nc2 Nb6

This shows that he is not thinking too rigidly. The knight leaves the fertile c4-square – but only temporarily.
37.Bd4?

Karasev, who was a well-known master in the Soviet Union, commits a serious though not decisive mistake. He gives
up the b-file free of charge.
A) Better was 37.Kd3! a5 (37...Nc4 38.Nb4) 38.Rb1 Nc4 39.Rb5 and White is not at all worse;
B) Also after 37.Rb1! Nc4+ 38.Kd3 Bxa3 39.Nxa3 Nxa3 40.Rb6 Nc2 (after 40...Ra8 41.Rd6 White is safe) 41.Bf6 a5
42.Rb5 a4 43.Kd2 a3 44.Ra5 Black is so tied up that he cannot win.

37...Nc4+ 38.Kd3 Rb8!

Karpov occupies the only open file, this is another device he likes to employ in endgames.
39.a4?

White misses the surprising pawn sacrifice 39.Nb4!, which would have allowed him to escape because White’s king is
much more active than Black’s. There is sometimes a fine line between a strong, centralized king and a vulnerable one:
39...Bxb4 40.axb4 Rxb4 41.e4 (after 41.Ra1 Rb3+ 42.Kc2 Ra3 43.Rxa3 Nxa3+ 44.Kb3 Nb5 45.Kb4 White is not
losing either):
A) Now, taking the pawn with 41...dxe4+ turns the White king into a hero: 42.Kc3! a5 (after 42...Ra4 43.Kb3 Black
drops the knight, which is also trapped after 42...Nb2 43.Kxb4 Nxd1 44.fxe4) 43.Bc5 Ra4 44.Kb3 winning a piece;
B) Or 41...Rb3+ 42.Kc2 Rb5 43.Kc3 Na5 44.Bf6 and White is absolutely safe.

39...f5

Stopping e3-e4 and preparing a passage for his king towards the centre.

40.Ke2 Kf7!

It took a while to bring up the king, but other aspects were more important up till now. Sooner or later, though, it must
happen.

41.Ba1 Ke6 42.h3 h5

This stops g2-g4 and removes his last pawn from a vulnerable black square.

43.Bd4?

Waiting with 43.Nd4+ was better. After 43...Kd6 44.Nc2 Black cannot easily swap the bishops.

43...Bd6 44.Rd3 Rb7

Again Karpov occupies himself with positional improvements.

45.Rd1
45...Be5!

There are no more improvements to be made, so Black exchanges in order to invade.

46.h4 Bxd4!

And again in a way that makes White’s position fall apart.

47.Nxd4+ Ke5 48.f4+ Kd6

Black has reached a winning position. Not 48...Ke4?? 49.Nb5 with mate next move. One has to acknowledge the limits
of an active king.

49.Ra1 Rb2+ 50.Kf3 Rd2 51.a5

White cannot really occupy the open file with 51.Rb1 because 51...Rxd4 wins. After 51.Re1, 51...Ra2 wins as well.

51...Rd3 52.Nc2 Kc5

Black still has time to collect material later. His domination grows unbearable.

53.Ra4 Rc3 54.Ra2 d4 55.Kf2 d3 56.Nd4 d2 57.Ra1 Nxe3 0-1

In an interview in the late 1970s, Karpov called the next game the most difficult of his life until that moment.

GAME 19
Torre, Eugenio
Karpov, Anatoly
Stockholm Wch-jr prel-B, 1969 (3)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.d4 exd4 7.e5 Ne4 8.b4 Nc3 9.Nxc3 dxc3 10.a3 0-0 11.Qd5 b5
12.Bb3 a5 13.Be3 Bb7 14.Rad1 axb4 15.Qxd7 Qc8 16.e6 fxe6 17.axb4 Kh8 18.Bxe6 Qxd7 19.Rxd7 Bc8 20.Rxc7
Bxe6 21.Rxc6 Bc4 22.Rb1 Ba2 23.Rd1 Bxb4 24.h3 Rfc8 25.Rxc8+ Rxc8 26.Ra1 Bc4 27.Ne5 Bd5 28.Rb1 Ba3
29.Rxb5 Be4 30.Bd4 h6 31.Rb3 Bb2 32.Nd3 Ba1 33.Ra3 Bxd3 34.Rxa1 Bxc2 35.Ra7 Rg8 36.h4 Kh7 37.Bxc3 h5
38.f3 Bf5 39.Kf2 Kg6 40.Ke3 Be6 41.Ke4 Bc8 42.Rc7 Re8+ 43.Be5 Bf5+ 44.Kf4

Torre’s own words paint a poignant picture of what unfolds.

‘During the endgame I felt I was on the verge of scoring the full point, but at the same time I was not sure if the game
was winning or saveable for Black. I tried to play what I thought then were the best moves but somehow towards the
end I sensed that my advantage was disappearing.’

44...Be6

Karpov is actually in deep trouble. He is already one pawn behind and he is about to lose a second one. However, there
is a problem for White – if the rooks are exchanged the opposite-coloured bishop endgame is a draw. That fact alone
gives Karpov chances to survive. 44...Rg8 would not save the pawn in view of the instant reply 45.Bxg7.

45.Rxg7+ Kh6 46.g4?

There is no definitive evaluation of the position after this move, but in a few years computer programs will be able to
dissect the position. The present evaluation can be questioned, but what is clear is that the g4-pawn is so much in
trouble that Eugenio could not win the game. His rook can never really avoid being shackled to the g-pawn.
Very promising was 46.Ra7 and now:
A) After 46...Rg8 47.Ra6 Rg6 White has a lovely win with 48.Bf6 Bc4 49.Bg5+ and White safely covers the g-file – he
should be winning;
B) 46...Rf8+ 47.Ke4 Rg8 48.Bf4+ Kg6 49.Bg5 and White wins after carrying through g2-g4.

46...hxg4 47.fxg4 Rf8+!

Karpov diverts the king from the defence of the g4-pawn.

48.Kg3

If 48.Ke3 Rg8 49.g5+ Kh5 50.Rh7+ Kg4 51.Rh6 Bf5 52.Bf6 Bg6 White’s rook is trapped, saving Black.

48...Rf7!

Karpov seizes his chance.

49.Rg5

This is the only way to keep the rooks on the board. If 49.Rxf7? Bxf7 Black easily holds the position, while 49.Rg8?
Rf3+ 50.Kxf3 Bxg8 is a draw as well.

49...Ra7 50.Bd4!

Improving the scope of the bishop is a good preparation for the advance of the king. 50.Rh5+ looked promising;
however, it was premature to free the rook: 50...Kg6 51.Bf4 Ra3+ 52.Kf2 and now:
A) 52...Ra4 53.Rg5+ Kf7 54.Kf3 Ra3+ 55.Be3 Kf6 56.Ke4 Kf7 57.Bd4 (57.h5 Ra8! 58.Kf4 Rg8 59.Rxg8 Kxg8 60.g5
Kh7 and Black draws) and White has improved his position: 57...Rh3 58.Bf2 (58.h5? Bxg4) 58...Ra3 59.h5 Ra8 60.Kf4
(on 60.Bd4 there again follows 60...Rg8 61.Rxg8 Kxg8 and now 62.g5 Bg4! and Black holds) 60...Rg8 61.Rxg8 Kxg8
62.g5 Kh7 and again, Black saves the day;
B) 52...Ra2+ 53.Ke3 (not 53.Kf3 when the lovely shot 53...Bxg4+! saves Black) 53...Ra3+ 54.Ke4 Ra4+ 55.Ke5
analysis diagram

Black now has a nice tactical simplification, but first let’s look at some winning motifs for White in this position.
B1) If 55...Bc8? 56.Rh6+ Kg7 57.g5! (interestingly, most of the time in this endgame, the g-pawn has to be pushed;
57.h5? Bxg4 58.Rg6+ Kf7 59.Rxg4 Ra5+ 60.Kd6 Rxh5 is a draw) 57...Ra5+ (if instead 57...Bg4 58.Be3 and White will
transfer the bishop to d4, with real winning chances) 58.Kd6 Bf5 59.Be5+ Kg8 60.Rh8+ Kf7 61.h5 (this looks
dangerous for Black. White plans to put the bishop on f6 and give a check from the seventh rank on the queenside. If
Black places his rook on the seventh rank, then the white rook goes to a6 and he removes the bishop from f6 to prepare
g5-g6+) 61...Be4 (after 61...Ra6+ 62.Kd5 Ra5+ 63.Kd4 Ra4+ 64.Kc3 Ra5 65.Bf6 Ra3+ 66.Kb4 Rd3 67.Rb8 Rd1
68.Rb7+ Rd7 69.Rb6 Rd1 70.Bb2 White wins) 62.Bf6 Rd5+ (62...Bd3 63.Rd8) 63.Kc7 Ra5 64.Rb8 Bf3 65.h6 Be4
analysis diagram

66.h7! Bxh7 67.Kb6 wins;

B2) 55...Bxg4! (Black can reach a drawn position with the help of some cute tactics) 56.Rg5+ Kh6 57.Rxg4+ Kh5
58.Rg8 (58.Rg1 Ra5+) 58...Ra5+! (this check is important – 58...Kxh4 loses to 59.Kf5) 59.Ke4 Kxh4 and Black holds.

50...Ra3+ 51.Kf4 Ra4 52.Ke5 Bb3

Karpov wants to retain the option of attacking the g4-pawn. White’s king might win a tempo by attacking the bishop,
but the c2-square is far enough to be safe.

53.Rh5+ Kg6

54.Rg5+?!

Torre abandons the idea of freeing his rook. The most direct was 54.Rh8! which almost certainly wins after 54...Bd1
(54...Ra5+ 55.Kf4 Ra4 56.h5+ Kf7 57.Ke3 Bd1 58.g5, or 57...Bc2 58.Rc8 Bb1 59.g5 may well be winning for White)
55.g5 (after 55.h5+ Kg7 56.Rd8 Bxg4 57.Kf4 Rxd4+ Black survives) 55...Ra5+. Now the question is if the rook can
cause enough trouble to escape with a draw. 56.Kf4 and now:
A) 56...Bc2 57.h5+ Kf7 58.Rh6 – White carries out g6 and wins;
B) 56...Ra4 57.Rh6+ Kf7 58.g6+ and White wins;
C) 56...Rf5+ 57.Ke3 Rf3+ 58.Ke4 Rg3 59.Bf6 Rh3 60.Be5 Kf7 61.Rh7+ Kg8 62.Rh6 Bc2+ 63.Kf4 Kf7 64.h5 and
White wins.

54...Kh6 55.Kf6 Ra6+ 56.Kf5

White can’t hide on the seventh rank: 56.Ke7 Ra4 57.Bg7+ Kh7 58.Bf6 Ra7+.

56...Ra5+ 57.Be5 Bc2+ 58.Kf6

The king is exposed to checks on the fifth and sixth ranks. It was simpler to search for shelter on his own half of the
board with 58.Kf4!? Ra4+ 59.Ke3
A) 59...Ra7 60.Rg8 and now:
A1) 60...Rf7 61.Rh8+ Rh7 (or 61...Bh7 62.Re8 Bb1 63.Re6+ Kh7 64.h5, and g5 and g6 soon follows) 62.Rd8 Re7
63.Rd6+ Kh7 64.Kf4 Rf7+ 65.Kg5 and White wins as he threatens both Rh6 and Rd8. If Black parries, h4-h5 and Kh4
decides;
A2) 60...Bd1

analysis diagram

61.Bd4! – the bishop covers the d-file and the fourth rank. It turns out that the bishop on d1 is short of squares on the
diagonal: 61...Re7+ 62.Kd2 Bf3 63.g5+ Kh5 64.g6 Kxh4 (64...Rd7 loses to 65.g7) 65.g7 Bd5 66.Rd8 and Black loses
the rook.
B) 59...Re4+ 60.Kd2 Bb1 61.Kc3 Re3+ 62.Kd4 Re4+ 63.Kd5 Bc2 64.Bg7+ Kh7 65.Bd4 Bd3 66.Rg7+ Kh6 67.Rg8
Rf4 68.Rh8+ Kg6 69.g5 Rf5+ 70.Kd6 Ra5 71.Bf6 and White probably wins with h4-h5 and a rook check from the
queenside.

58...Ra6+ 59.Ke7 Ra7+ 60.Kd6 Ra6+ 61.Kc5 Ra4 62.Rg8

62...Bh7!

This is a very strong move that keeps the young Filipino busy.

63.Rg5

The alternative was 63.g5+ Kh5 64.Rh8 and now:

A) After 64...Kg6 White does not seem to have a convincing winning method – maybe it does not even exist.
One try is 65.Bd4 (or 65.Bg3 Rg4 and the rook keeps harassing the bishop) 65...Ra5+ 66.Kb4 (in the event of 66.Kd6
Ra6+ 67.Ke5 Ra4 the rook keeps causing problems) 66...Ra2 67.Rf8 Kh5 68.Be3 Rh2 69.Bf2 Bg6 (69...Kg4 70.Rf6)
70.Rf6 (after 70.Kc5 Kg4 71.Rf6 Bf5 Black eases his game) 70...Rh1 71.Kc5 Rd1 72.Rf4 Rd3 and Black has blocked
the white pawns well.;
B) Black can choose 64...Kxh4 as an interesting way of forcing matters: 65.g6 Kg5 66.gxh7 (66.g7 Kg6)
analysis diagram

66...Kg6 – this position brings to mind the fourth game of the Kramnik-Kasparov World Championship match in 2000.
White had an h-pawn and an extra knight, but still Black was able to hold. That time the knight was constricted to
defensive duties, this time the rook is: 67.Kd5 (White could stop the black rook reaching the seventh rank with 67.Kb6,
but only temporarily: 67...Re4 68.Bc3 Re7 wins the h-pawn after all) 67...Ra7 68.Ke6 Ra6+ (Black cannot take the
pawn and simplify to a rook and bishop versus rook ending with 68...Rxh7? because of 69.Rg8+ and White wins)
69.Bd6 Kg7 – remarkably, with the rook on a6 this is a draw, whereas White would win with the rook on b6, viz.
69...Rb6? 70.Rd8 Kxh7 71.Kf6 and White wins) 70.Rd8 Kxh7 71.Kf6 Ra2 – now White cannot win: 72.Rd7+

analysis diagram
72...Kh8!! (the only move that draws. If the rook had moved to a1 on move 71, then 72...Kg8 would draw here as well)
73.Bc5! (73.Kg6 Rg2+ 74.Kh6 Kg8 and the black king leaves the danger corner in time) 73...Ra6+! – and here we see
the difference between the rook going to the a- or the b-file on move 69.
Back to the game.

63...Bc2 64.Bc3

After 64.Bg7+ Kh7 65.Bc3 Bg6 (not 65...Bd1? 66.Rg7+ Kh6 67.g5+ Kh5 68.Rh7+ Kg6 69.Rh6+ Kf5 70.g6 Bb3
71.Bb4) 66.Bb4 (66.h5 Bd3) 66...Ra7 67.Kd4 Rd7+ Black can put up some resistance.

64...Bd1 65.Bb4 Be2 66.Rh5+ Kg6

With another patient waiting move Karpov restricts White’s pieces even more.

67.Re5

Torre tries to improve his rook, but soon it has to return to the g-file.
With 67.Rd5!? White could have attempted to bring the rook back in a devious way. It places Black in a zugzwang-like
situation:
A) 67...Bf3? (the bishop indeed cannot move) 68.h5+ Kh7 69.Rd7+ Kg8 70.h6 Be4 71.Rg7+ Kh8 72.Bc3 and the
battery wins;
B) 67...Ra1? 68.h5+ Kh7 69.Rd7+ Kg8 70.Rd8+ Kf7 71.h6 Kg6 72.Bd2 Ra7 73.g5 Bh5 74.Rd6+ Kh7 75.Bc3 and
White wins;
C) 67...Ra7 (Black has to defend the seventh rank) 68.Rd6+ Kf7 (in case of 68...Kh7 69.g5 Ra6 70.Rd2 Re6 71.Bc3
Kg6 72.Rd8 White is close to obtaining a winning position) 69.g5 (throughout this endgame, pushing this pawn gives
the best practical chances) 69...Ra6 70.Rd8 (Black’s position remains very difficult) 70...Kg6 71.Bc3 Kf5 72.Rh8 Bf3
73.Bf6 and White should win as he can now push his h-pawn closer to promotion.
67...Bf3 68.h5+?!

So finally he pushes one of the pawns, but this is an unfortunate moment. Now White no longer has a winning position.
He could still have created enormous obstacles for Black, but then Torre would have had to come up with spectacular
moves. Maybe he had missed something, or he felt that there was no point trying to go on squeezing and wait for a
mistake. With hindsight it is easy to say that pushing the other pawn was better.
A) After 68.Rf5 Be2 69.h5+ Kg7 70.Rf4 Ra7 71.Kd6 Kh7 72.Bc3 Kh6 73.Bd2 Kh7 Black is still alive;
B) 68.Kb5 Ra7 69.Bc3 Kh7 and Black is living very dangerously, but the game is not over yet.

68...Kg7
69.Re6?!

This probably doesn’t spoil the win, as it may not even exist, but certainly he could have done better than this.
After 69.Kb5 Ra7 (69...Ra8 loses to 70.Bc3) 70.Rg5+ Kh7 71.Bc3 Be2+ White is tied to the defence of the g4-pawn.
Amazingly there was a tremendous alternative to these rook and king moves.
69.Bd2!!. A study-like move. White sacrifices material in order to create the best chances of victory. We love endgames
anyway, but this kind of magic moves enhance this feeling. White allows Black to take his precious g4-pawn with
either of the two pieces.
The alternatives for Black are:
A) Waiting passively with 69...Ra7 is insufficient: 70.h6+ Kh7 71.g5 and White will invade on the eighth rank:
71...Bh5 72.Rd5 Bg6 (72...Rb7 73.Rd8) 73.Rd8 Bf5 74.Kd6 Kg6 75.Rg8+ Kh7 76.Rf8 and White wins;
B) 69...Bxg4? looks like the logical move. However, it allows a fabulous win: 70.h6+ and now:
B1) 70...Kf7 71.Rg5 Be2 72.Rg7+ Kf8

analysis diagram

B11) With 73.Rd7! White has an ‘ordinary’ win at his disposal: 73...Rc4+ 74.Kd5 Rc2 (74...Rh4 75.h7; or 74...Kg8
75.Bg5 Bd3 76.h7+) 75.Bf4 Bc4+ 76.Kd4 Bg8 77.Bd6+ Ke8 78.Re7+ Kd8 79.Rg7 and White wins;
B12) 73.Bb4!! (what a wonderful battery!) finishes Black off: 73...Bd3 (now after 73...Ra1 74.Kd4+ Ke8 75.h7 wins,
and after 73...Rxb4 74.Kxb4 Bd3 it makes all the difference that Black’s king is not in the corner. White’s king
marches to f6 and then h6-h7 wins) 74.Rg3 wins the exchange or the bishop outright, as well as the game.
B2) 70...Kg6 71.Rg5+ Kh7 72.Rg7+ Kh8 73.Bc3 Ra3
analysis diagram

White now has a lot of tempting moves that do not work:

B21) 74.Kb4 Rxc3 75.Kxc3 is a draw. If the white pawn was on h5 and the king on h6, White would win;
B22) 74.Bf6 Ra6 75.Bd4 (in case of 75.Rf7+ Kg8 76.h7+ Kxf7 77.h8=Q Rxf6 Black holds, e.g. 78.Qh7+ Kf8!)
75...Rxh6 76.Rxg4+ Kh7 77.Rg7+ Kh8 78.Be5

analysis diagram

78...Rc6+!! and Black escapes in a magical way with the aid of the stalemate motif;
B23) 74.Be5 Ra5+ 75.Kd4 Rxe5 76.Kxe5
analysis diagram

76...Bf5!!– this is the only move that brings salvation for Black;
B24) 74.Bd4!! (saving the bishop) 74...Ra5+ 75.Kb4 and finally White wins.
C) Best is 69...Rxg4!. Only this capture of the pawn saves Black.

analysis diagram

70.h6+ Kh8 (70...Kg8 71.Re8+) 71.Re8+ (White can also play on with 71.Re7 Be4 72.Bc3+ Kg8, but Black can hold)
71...Rg8 72.Re7 (threatening to win with Bc3+ and h6-h7) 72...Rc8+ (it looks as if Black has a few checks and that’s
all. But life is full of painful surprises) 73.Kb4 Rb8+! (this is the way to hold the position by force. 73...Bd5 74.Bc3+?!
Rxc3!! (the only move. The defensive plan is remarkable, even though it has been known for a long time) 75.Kxc3 and
now 75...Kg8 loses, but 75...Ba2 draws) 74.Ka3 Ra8+ 75.Kb2 Rb8+ 76.Kc2 Rc8+ and remarkably, White cannot run
away from the checks without allowing the rook to take the bishop.
So White cannot make it without stepping onto the d-file: 73.Kd4.

analysis diagram

C1) 73...Bh5 74.Bf4!;

C2) With 73...Bd1 Black would make his own life extremely hard: 74.Bf4! (saving the bishop) 74...Bc2 75.Be5+ Kg8
76.Rg7+ Kf8 77.Bd6+ Ke8 78.Bc7! (a marvellous way to stalemate the black rook. Now the white king plans to walk
to b7 and capture it) 78...Kf8 79.Kc5.
analysis diagram

This wins after 79...Bd3 (79...Re8 80.Bd6+; after 79...Ra8 White wins the rook with 80.Bd6+ Ke8 81.Rg8+) 80.Kb6
Be4 (after 80...Ra8 81.Bd6+ Black loses the bishop, and in case of 80...Rxc7 81.Kxc7 the black king does not reach the
corner, so he loses) 81.Bd6+ and this time the bishop goes.
It all looks beautiful, or is it too good to be true? Indeed there is a flaw – on move 79 or 80 Black could play ...Rxc7
followed by ...Kg8! and he draws again.
The alternative to 79.Kc5 is 79.Kd5 Bb1 and now:
C21) 80.Kd6 Re8;
C22) 80.Bd6+ Ke8 81.Ke6 Ba2+ 82.Kf6 Bb1 (82...Rc6 83.h7) and Black has not yet lost;
C23) 80.Kc6 Ra8! 81.Bd6+ Ke8 and although Black is living very dangerously, there is no win.
C3) But instead of 73...Bd1, 73...Rd8+ is clearly simpler.

analysis diagram

74.Ke3 Rxd2! 75.Kxd2 Bg4 and Black draws, as White’s king is too far removed from the action and the bishop can be
transferred to the b1-h7 diagonal. Please note, if White could play 76.Kf4 now he would win!
Back to the game.
69...Kf7 70.Rg6 Be2!

Karpov defends this position extremely patiently. This move ties all White’s pieces up. It would have been premature to
play for the rook exchange with 70...Ra8? because 71.Bc3! wins a vital tempo: 71...Be2 (71...Rg8 72.Rf6+) 72.Rg7+
and White wins.

71.Rg5

Or did White have a more useful move?

71...Ra8!
It would not have helped White because of this superb defensive resource. Karpov played patiently when he had to, but
acted when the position required this. The rook is not only protecting the eighth rank, it also creates a threat.

72.Kd5

White cannot do anything against Black’s threat. If 72.h6 Rg8, or 72.Kd6 Rg8 73.Rf5+ Kg7 74.Bc3+ Kh6 75.Bd2+
Kh7; or 72.Bd2 Rg8 73.Rf5+ Ke6 and Black easily holds in all of these lines.

72...Rg8!

The rook exchange secures the draw.

73.Rf5+ Kg7 74.Bc3+ Kh6 75.Rf6+

After 75.Bd2+ Kh7 76.Rf7+ Rg7 77.Rxg7+ Kxg7 it is also a draw.

75...Kg5

Not 75...Kh7?? 76.Rf7+.

76.h6 Rd8+

Catastrophic would be 76...Kxg4?? which would spoil the fruits of many hours of intrepid defence: 77.h7 Rh8
(77...Rd8+ 78.Rd6) 78.Rf4+ Kxf4 79.Bxh8 Bd3 80.Be5+ and White wins.

77.Rd6 ½-½

Torre gave up trying to win and offered a draw, which of course was accepted. Karpov showed superb resilience, Of
course, the assistance of Furman at this World Championship certainly helped, especially in this game, which was
adjourned twice. Let’s have a quick look at the moves if Torre had tried to continue!
If 77.Ke5 Re8+ 78.Kd6, 78...Bd3 is the simplest as Black doesn’t have to suffer in the rook versus rook and bishop
endgame. 78...Bxg4 79.h7 Bf5 80.Rf7 Bxh7 is of course a draw, too, but in practice many players lose.

GAME 20
Adorjan, Andras
Karpov, Anatoly
Budapest Hungary-RSFSR, 1969 (2)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.0-0 f6 6.d4 Bg4 7.c3 Bd6 8.Be3 Qe7 9.Nbd2 0-0-0 10.Qc2 h5 11.h3 Bxf3
12.Nxf3 g5 13.dxe5 Bxe5 14.Nxe5 fxe5 15.Rad1 Rxd1 16.Rxd1 g4 17.Qd3 Nh6 18.Bxh6 Rxh6 19.Qe3 Rd6 20.Rxd6
cxd6 21.hxg4 hxg4 22.Qh6 Qf7 23.b3 Kc7 24.Qg5 Qd7 25.Qg6
Black’s play is slightly restricted because the g4-pawn requires attention.

25...d5!

Karpov is not just waiting and hoping that he can keep his position together. Instead, he responds actively. White was
going to go after the g4-pawn with Kh2 and Kg3.

26.Qf6

26...Qd6!

Black sacrifices a pawn to create a passed pawn. Remember, passed pawns are particularly dangerous in queen endings
and usually they have great value there. Why? Because in a knight or bishop ending the piece can be sacrificed for the
passed pawn. This is sometimes also possible in a queen ending, but it is much rarer.

27.Qg7+ Kb6 28.exd5

The Hungarian player, who had finished second behind Karpov in the Junior World Championships, goes for a better try
than 28.Qxg4. Then, after 28...d4 29.cxd4 exd4 30.Qd1 d3 31.Kf1 Qd4 32.f3 Black has time to create another passed
pawn with 32...c5!. When both sides have passed pawns, extra pawns are a less important factor than speed, e.g. 33.g4
(in case of 33.Qd2 Kc6 34.g4 b5 35.g5 c4 Black is not worse) 33...Kc7 34.g5 b5 35.g6 c4 36.bxc4 bxc4 37.Qa4 d2
38.Ke2 Qd3+ and Black is safe.
With 32...Qc3! the queen can try forcing matters as well. This line is also convincing, but it requires more precision:
33.g4 Kc5 34.e5 Kd4! (activating the king is important here) 35.e6 Qc2! 36.Qa1+ (on 36.Ke1??, 36...Ke3
checkmates) 36...Ke3 37.Qe5+ and the game ends in a draw.

28...cxd5 29.Qxg4 d4 30.cxd4 exd4

31.Qe2?!

Allowing Black’s pawn to advance one rank further than was wise. With 31.Kf1! the king should have tried to block the
pawn at once. Then, after 31...Qe5 (31...d3?! 32.Ke1! d2+ (in case of 32...Qf6 33.Qb4+ Ka7 34.Qc5+ Ka8 35.Kd2
Black is in trouble) 33.Kd1 the d-pawn falls: 33...Qc5 34.Qg6+ Ka7 35.Qc2 is no fun for Black) 32.Qe2 (or also
32.Qg6+ Kc5 33.Qd3 Kb4 and here, unlike in the game, White has chances to do something with his f- and g-pawns)
32...Qd5 (32...Qc5 33.Qd3) 33.Qd3 Ka5, it is hard to say whether White would have gained an advantage by blocking
the pawn earlier, but he would have had some chances for sure.

31...d3 32.Qd2
32...Qd4

Black’s d-pawn is strong, being so close to the promotion square. This compensates for the pawn deficit. This must have
been the sort of position Karpov had envisaged when he sacrificed the pawn.

33.g3 ½-½

Adorjan just offers a draw with a move that indicates he is not trying to win. With 33.Kf1 White could still try to get in
front of the pawn with his king and push his kingside pawns. After 33...Kc6 34.f4 (in the event of 34.f3 Kd5 35.g4 b5,
Black should hold) 34...Kd5 35.g4 Ke4 36.f5 Qg7 37.Qe1+ Kf4 38.Qf2+ Ke4 Black is safe.
 GAME 21
Karpov, Anatoly
Krogius, Nikolay
Kuibyshev RSFSR Championship 1970 (1)

1.e4 e6 2.d4 d5 3.Nd2 c5 4.Ngf3 Nc6 5.exd5 exd5 6.Bb5 Bd6 7.0-0 Ne7 8.dxc5 Bxc5 9.Nb3 Bb6 10.Re1 0-0 11.Be3
Bg4 12.Bxb6 Qxb6 13.Bxc6 Nxc6 14.Qxd5 Nb4 15.Qe4 Bxf3 16.gxf3 a5 17.a3 Nc6 18.Qe3 Qb5 19.a4 Qh5 20.Qe4
Qg5+ 21.Kh1 Qf6 22.Nc5 Rad8 23.c3 b6 24.Nd3 h6 25.f4 Rd7 26.Re3 Rfd8 27.Rg1 Ne7 28.Ne5 Rd1 29.Ree1 Rxe1
30.Rxe1 Nf5 31.Ng4 Qg6 32.Ne5 Qh5 33.Qf3 Qh4 34.Nc4 Qf6

35.h3!

Karpov makes room for his king and marks time, essentially putting the ball back in his opponent’s court.

35...Nh4

Black doesn’t want to wait and do nothing.

In case of 35...Kh7 36.Kh2 Kg8! (36...Rd7 37.Re8) 37.Re5 Nh4 38.Qe4 Ng6 39.Re8+ (39.Rb5 Qh4) 39...Rxe8
40.Qxe8+ Kh7 41.Qe3 Nxf4! White keeps an extra pawn, but his damaged pawn structure on the kingside makes it
difficult to exploit this.

36.Qe4 Ng6?

Now Black gets pushed back. He should have played 36...Nf5 and attempted to exchange knights:
A) 37.Qe5 Qc6+;
B) 37.Kh2 allows the knight exchange with 37...Nd6! 38.Nxd6 Qxd6 39.Qe5 Qd2 40.Re2 Qd1 and it is not easy for
White to convert his extra pawn;
C) With 37.Ne5! White could avoid the swap: 37...Kh7 38.Kh2 Rd6 39.Ng4 Qg6 40.Qc2 Rd5 41.Re5 and White keeps
his winning chances alive.
37.f5! Nh4?

The knight attacks the f5-pawn, but it gets stuck on the h-file. Even after 37...Nf8!? 38.Kh2 Rd7 39.Re3 White’s
pressure would still be there.

38.Ne3 Kh8

39.Re2!

The typical Karpovian small improvement.

39...Qg5 40.Kh2

Yet another one.

40...h5?

A mistake, as the pawn occupies a square that the queen could have used.
Better was 40...Kg8! 41.Qg4 (in case of 41.f4 Qh5 White’s advantage is not convincing; even less so after 42.Rf2 Rd6)
41...Re8! (after 41...Rd3 42.Qxg5 hxg5 43.Nc4 White would be much better) 42.Rc2 Re5 and Black is alive.

41.f4! Qh6

On 41...Qf6 White wins with 42.Nd5 Qd6 43.c4.

42.Kg3!

Reminiscent of the famous Alekhine-Capablanca game (AVRO 1938), where the white king threatened to catch the
knight.

42...g5
This saves the knight, but White can now dissolve his doubled pawns.

43.fxg6 Nxg6 44.Nf5 Qf8 45.Qf3

Collecting a second pawn.

45...Kg8 46.Qxh5 Rd3+

47.Kg4!

White has to make cautious moves, but these are not difficult to find. It shows Karpov’s confidence.

47...Qa8 48.Nd4! Rd1 49.Nf3

It looks as though Karpov is just defending. In reality he is slowly but surely building up an attack.

49...Rd5

With 49...Rf1 Kg3 he maintains the advantage.

50.Ng5

White suddenly catches the black king.

50...Qc8+ 51.Kg3 Rf5 52.Qh7+ Kf8 53.Qh6+ Kg8

54.Ne6! 1-0

GAME 22
Karpov, Anatoly
Sergievsky, Vladimir
Kuibyshev RSFSR Championship 1970 (3)

1.e4 d6 2.d4 Nf6 3.Nc3 Nbd7 4.Nge2 b5 5.e5 dxe5 6.dxe5 Nxe5 7.Qxd8+ Kxd8 8.Nxb5 a6 9.Nbd4 Bb7 10.Bf4
Ng6 11.0-0-0 Kc8 12.Bd2 e6 13.Nb3 Ng4 14.Be1 Bd6 15.h3 Nf6 16.Na5 Rd8 17.c4 Nf4 18.Nxb7 Kxb7 19.Kc2
Be5 20.Nc3 Rxd1 21.Nxd1 Rd8 22.g3 Ng6 23.Bg2+ Kb6 24.Nc3 Bd4 25.Ne2
25...Be5?

Black preserves the bishop, but allows himself to be pushed further back.
A) After 25...c5 26.Nxd4 cxd4 (or 26...Rxd4 27.Bc3 Rxc4 28.Rd1 and White is almost winning; he can trap the rook
next move) 27.b4 Ne5 28.c5+ Kc7 29.Bd2 White has an edge because of the two bishops;
B) 25...e5 and now:
B1) If 26.b4 c5 (after 26...c6, 27.c5+ Kc7 28.Nxd4 exd4 29.b5 is strong) 27.a3 (in case of 27.Nxd4 exd4 28.a3 Ne5
Black is active enough) 27...Re8 28.Nc3 e4 Black is in the game;
B2) 26.Nxd4 exd4 27.Bd2 Re8 (better than 27...c5 28.Re1 Re8 (in case of 28...a5 29.Kb3! Rd7 30.Ka4 White’s king
will cause trouble) 29.Rxe8 Nxe8 30.f4 Nd6 31.Kd3 and White’s two bishops give him the better endgame) 28.b4 Re2
29.c5+ Ka7 30.Rf1 Ne5 and Black’s pieces are active enough to create a balanced position.

26.f4!

Karpov is not only good at ‘massaging’ the position in endgames, he can play sharply too.

26...Bd6 27.Nc3!

Threatening to push Black further back with 28.Na4.

27...c6

28.c5+!!

To play endgames at this exceptional level, one has to handle endgame tactics superbly as well.

28...Bxc5

28...Kxc5 leads to a forced checkmate: 29.Bf2+ Kb4 30.a3+ Ka5 31.b4+ Bxb4 32.axb4+ Kxb4 33.Rb1+ Kc4 34.Bf1+
and Black can only postpone the mate for one more move.

29.Na4+ Kb5

30.Kb3!

Suddenly Black’s king is caught.

30...Rd3+ 31.Bc3 Rxc3+

This allows Black to avoid immediate resignation, but it cannot avert the loss.

32.Nxc3+ Ka5 33.Na4 Bf2 34.Rc1 Ne7 35.Rc2 Be3 36.Bxc6 Nh5 37.Be8 Nxg3 38.Kc4 Nd5 39.a3 Nb6+ 40.Kd3
Nxa4 41.Kxe3 Nf5+ 42.Kd3 1-0

GAME 23
Karpov, Anatoly
Rashkovsky, Nukhim
Kuibyshev RSFSR Championship 1970 (8)

1.c4 Nf6 2.g3 g6 3.Bg2 Bg7 4.d4 0-0 5.Nc3 d6 6.Nf3 Nc6 7.0-0 Bf5 8.d5 Na5 9.Nd4 Bd7 10.b3 c5 11.dxc6 bxc6
12.Bb2 c5 13.Nc2 Rb8 14.Ne3 Nc6 15.Qd2 Nb4 16.h3 Bc6 17.Ncd5 Nbxd5 18.cxd5 Bb5 19.Bc3 Ne8 20.Ba5 Qd7
21.Rac1 Ba6 22.Rfe1 Rb7 23.Bc3 Bxc3 24.Qxc3 Ng7 25.Nc2 Rb6 26.e4 e5 27.dxe6 fxe6 28.Rcd1 Qf7 29.f4 Nh5
30.Rd2 Qg7 31.Qe3 e5 32.Rf2 exf4 33.gxf4 Qh6 34.f5 Qxe3 35.Nxe3 Nf6 36.Rd1 Kg7 37.Rfd2 Ne8
38.fxg6!?

This time Anatoly finds no time for the improving move 38.Kh2.

38...hxg6

With 38...Kxg6 39.Nf5 c4 Black could try to exchange as many pawns as possible, and now:

A) Not 40.Nxd6?? c3! – oops!;

B) Better is 40.bxc4 Bxc4 41.Nxd6 Nxd6 42.Rxd6+ Rxd6 43.Rxd6+ Rf6, but Black probably holds after 44.Rd4 Bxa2
45.Ra4 Bb1 46.Rxa7 Re6;
C) 40.Ne7+!! – the knight goes after the bishop! According to Ashot Nadanian, this is reminiscent of Capablanca’s
famous pawn winning manoeuvre against Yates in New York 1924. 40...Kg7 41.Nd5 Rc6 (41...Rb8 42.Bf1) 42.Nb4
Rb6 43.Nxa6 Rxa6 44.Bf1. Moreover, Black still has a weak d-pawn and the white bishop is superior to the black
knight.

39.e5!

Another typical Karpov idea: invading the seventh rank. This temporary pawn sacrifice prepares the passage.

39...dxe5 40.Rd7+ Rf7 41.Rxf7+ Kxf7 42.Rd7+ Ke6 43.Rxa7 Nf6 44.Rc7 Bd3 45.Rxc5

White has won a pawn, but the game is not over by far. Black’s king is better placed and his pieces are more active.

45...Ra6

After 45...e4 46.Rg5! (46.Kf2 Ra6 47.a4 Rb6 48.Rc3 is a little passive) 46...Ra6 47.a4 Rb6 48.Rxg6 Kf7 (48...Rxb3
loses to 49.Ng4) 49.Rg5 Rxb3 50.Kf2 and the white king will soon reach f4, assuming greater power.
46.Rc6+?!!

After the exchange of rooks, White’s queenside pawns will have freer passage, but Black’s king also becomes more
active in the centre. The fact that he makes an extremely hard and responsible decision just before sealing a move
shows Karpov’s self-confidence. His deep concept will prevail and win the game 24 moves later.
Objectively however, this was not best. After 46.a4 Rb6 47.Rc3 e4 48.a5!? (48.Kf2 transposes to the position analysed
on move 45) 48...Rb5 49.b4 Rxb4 (on 49...Ke5 50.a6 Rxb4 White wins with 51.Ra3) 50.Rc6+ Ke5 51.Rxf6 Rb1+
52.Rf1 Bxf1 53.Bxf1 Ra1 White is, of course, clearly better, though it is not easy to support the a-pawn.

46...Rxc6 47.Bxc6 Kd6 48.Bg2 Kc5 49.a3 Kd4 50.Kf2 e4

51.a4!

Karpov sealed this move. Particularly if it is not obvious, such a situation puts the pressure on the opponent, who has to
analyse positions with pawns on several different squares. The text seems to violate the principle of avoiding the colour
of the opponent’s bishop. Sometimes principles clash in chess. Karpov lays the emphasis on his passed pawn.

51...Kc3 52.a5!

Diverting the black pieces from the kingside. In case of 52.Bf1 Nd7 53.a5 Nc5 Black would hold.

52...Nd7?

Black wants to stop the a-pawn with the knight and collect the b-pawn with the king. But this loses.
A) After 52...Kxb3 53.Bf1 Ne8 54.a6 Nc7 55.a7 Ka4 (55...Bxf1 loses: 56.Kxf1 Ka4 57.Nd5 Na8 58.h4 Ka5 59.Ne7
Kb6 60.Nxg6 Kxa7 61.h5 Nc7 62.h6) 56.Nd5 Na8 57.Bg2 Kb5 58.Nf4 Bc2 59.Nxg6 Kb6 60.Ke3 Kxa7 61.Bxe4
Bxe4 62.Kxe4 White’s h-pawn is not to be stopped;
B) Putting the pawn on a dark square and keeping the bishop on d3 does not help either: 52...g5 53.Bf1 and now:
B1) 53...Bxf1 54.Nxf1 Kxb3 55.a6 Nd5 56.a7 Nc7 57.Ng3 Kc4 58.Nxe4 Kb5 (58...Kd5 loses after 59.Nxg5 Kc6
60.h4 Kb7 61.h5 Ne8 62.h6 Nf6 63.Kf3 Kxa7 64.Kf4) 59.Nxg5 Kb6 60.Ne6 Na8 61.h4 Kxa7 62.h5 and White wins;
B2) 53...Kxb3 54.a6 Ne8 55.a7 Nc7 56.Nd5 Na8 57.Bxd3 exd3 58.Ke3 Kc4 59.Nf6 Kc3 (or 59...Kb5 60.Ne4 Kb6
61.Nxg5 Nc7 62.h4 Nd5+ 63.Kxd3 Kxa7 64.Ke4 Nf6+ 65.Kf5 Nh5 66.Kg6 Ng3) 60.Ne4+ Kc2 61.Nd2 Kc3
62.Nf3 Kc4 63.Ne5+ Kc5 64.Kxd3 Kb6 65.Nf7 Kxa7 66.Nxg5 and White wins.
C) 52...Kb4! – it is a bit of a surprise that Rashkovsky made a fatal mistake right after the adjourned position. Maybe
he had not anticipated Karpov’s sealed move in the first place. Maybe, as a young and inexperienced player, he had
misjudged the difficulty of eliminating White’s h-pawn. 53.Ng4
analysis diagram

C1) Interestingly, now the most natural move 53...Nd5? runs into trouble after 54.Ne5! (threatening to take on d3.
After 54.a6 Bxa6 55.Bxe4 Nf4 56.h4 g5! 57.hxg5 Nh3+ 58.Ke3 Nxg5 Black would draw) 54...Nc7 (54...e3+ 55.Kg3;
54...Nf4 55.Nxd3) 55.Ke3 Kxa5 (after 55...Kxb3 56.Bxe4 Bxe4 57.Kxe4 Ka4 58.Nxg6 White wins) 56.Bxe4 Bxe4
57.Kxe4 Ne6 (57...Kb4 58.Nxg6) 58.h4 Nf8 59.Kf4 Kb4 60.Kg5 White wins as the black king is too far away;
C2) The simplest is 53...Nh5! as after 54.Ke3 (54.Ne5 Kxa5) 54...Kxa5 55.Bxe4 Bxe4 56.Kxe4 g5 Black holds.

53.h4!

Getting closer to promotion and fixing the g6-pawn on the bishop’s colour.

53...Kxb3 54.Nd5 Nc5 55.Ke3 Ka4 56.Nf4 Kxa5

57.Nxg6

White has only one pawn left, but that is sufficient. A rook pawn is usually to the advantage of the defender, but not
when the latter must sacrifice a piece for it. Both for the knight and the bishop it is harder to sacrifice, and Black does
not manage it here.

57...Nd7

In case of 57...Kb6 58.Bh3 Bc4 (58...Bb5 loses to 59.h5 Be8 60.Bf5) 59.h5 Bg8 60.h6 Kc7 61.Ne7 Bh7 62.Bf5, White
has a decisive advantage.

58.Bxe4 Bc4

If 58...Bxe4 59.Kxe4 Nf6+, 60.Kf5 Nh5 61.Nf4 Ng7+ 62.Kg6 wins.

59.Kd4 Be2 60.Bf5 Nf6 61.Ke5 Ne8

Or 61...Nh5 62.Nf4, or 61...Ng4+ 62.Kf4 Nf6 (62...Nh6 63.Be6) 63.Kg5 Nh7+ 64.Kh6 Nf6 65.Nf4 and White wins.

62.Nf4 Bd1 63.Bg6 Nc7 64.Bh5 Bxh5

The main alternative is 64...Bb3 (64...Bc2 65.Bf3 Ne8 66.h5 wins for White) 65.Bf3 Bf7 66.Kf6 Be8 67.Ng6 when
White has prepared the passage of the pawn perfectly.

65.Nxh5

Black cannot to stop the h-pawn.

65...Kb6 66.Nf4 Ne8 67.h5 Kc7 68.h6 Nd6 69.h7 Kd7 70.Kf6 1-0

In Caracas Karpov made his debut on the international scene. In 1971, he became the youngest grandmaster in the world
at his first shot.
 GAME 24
Karpov, Anatoly
Parma, Bruno
Caracas, 1970 (3)

1.c4 Nf6 2.Nc3 e6 3.d4 Bb4 4.Qc2 0-0 5.Nf3 c5 6.dxc5 Na6 7.Bd2 Nxc5 8.e3 b6 9.Be2 Bb7 10.0-0 d6 11.Rfd1 a6
12.b3 e5 13.a3 Bxc3 14.Bxc3 Qe7 15.Ne1 Rac8 16.Rac1 Nfe4 17.b4 Nxc3 18.Qxc3 Ne6 19.Qd3 Rfd8 20.Bf3 Bxf3
21.Nxf3 g6 22.Nd2 Nc7 23.Ne4 Ne8 24.Qd5 Kg7 25.h3 Nf6 26.Nxf6 Kxf6 27.Qe4 Kg7

28.Rd5 Qc7?!

Black decides to wait and see how White is going try to crack his position.
The freeing move 28...b5!? was possible. It probably leads to an ending where Black is a pawn down, but which can be
defended:
A) 29.Rdd1 f5 30.Qd5 Qf7 31.Qxf7+ Kxf7 32.cxb5 Rxc1 33.Rxc1 axb5 34.Rc7+ Kg8 (Black’s king is passive, but he
holds the material balance. Also, 34...Ke6 35.Rxh7 Ra8 36.Rb7 Rxa3 37.Rxb5 Rb3 gives good drawing chances)
35.Rb7 Ra8 36.Rxb5 Rxa3 37.Rb8+ Kf7 38.Rb7+ Kg8 39.b5 Rb3 and Black probably holds;
B) 29.c5 f5 30.Qd3. In his analysis Borislav Milic stops here as White appears to win. Matters are not that clear,
however. After 30...dxc5! 31.Rcxc5 Rxd5 (in case of 31...Rxc5 32.Rxd8 Rc7 33.Rd6 Black has an unpleasant position,
though it is hard to tell whether White can win, e.g. 33...Ra7 34.Qd5 e4 and Black must suffer) 32.Rxd5 Qe6 33.Rd6
Qc4 34.Rd7+ Kf6 35.Qd6+ Qe6 and White is better, but his advantage is not winning, e.g. 36.f4 (or 36.h4 h5 37.f3
Qxd6 38.Rxd6+ Kf7 39.Rxa6 Rc3 and Black has decent drawing chances) 36...Qxd6 37.Rxd6+ Kf7 38.Rxa6 Rc3
39.Kf2 Rc2+ (also, 39...exf4 40.exf4 h5 41.Ra5 h4 gives Black better chances of drawing than White of winning)
40.Kf1 e4 41.Ra5 Rc3. Even if White can do something with his extra pawn, it won’t be easy.
29.f4! Re8

Considering the prolonged agony that Black will now suffer, Parma should have tried 29...exf4. A more open type of
position would have given him better chances to defend, as White’s king has little shelter. But it would not give Black a
joyful position:
A) If 30.Qxf4 Kg8:
A1) After 31.a4 Qc6 32.Rcd1 Qxa4 33.c5 bxc5 34.bxc5 Qxf4 35.exf4 Rd7 Black holds;
A2) 31.Qf6 b5 32.c5 dxc5 33.Rdxc5 Qd7 34.Qxa6 Rxc5 35.Rxc5 Qe7 and Black has real drawing chances because of
White’s exposed king;
A3) 31.e4 Qe7 32.e5 (32.a4 Rc6) 32...dxe5 33.Rxe5 Qd7 and White has a small advantage.
B) 30.Qd4+ Kg8:
B1) 31.exf4?! – Karpov stops his analysis here, claiming the position is plus/ minus. It seems that he just put into his
analysis what he had considered during the game. There was in fact a surprising rejoinder. 31...b5! would save Black:
32.c5 dxc5 33.Rcxc5 Rxd5 34.Qxd5 Qb6 and Black escapes;
B2) 31.Qxf4 b5 32.c5 Qe7 33.c6 Qe8 34.c7 Rd7:
B21) 35.Qd4 Qe7 (35...Rdxc7 36.Rxc7 Rxc7 37.Rxd6 Rc8 38.Rxa6 Rc4 39.Qd2 Qe5 gives Black an active position,
probably good enough for a draw) 36.Rc6 Rdxc7 37.Rcxd6 Rc1+ 38.Kh2 Qc7 and Black is able to hold;
B22) After 35.Rc6 Rdxc7 36.Rxa6 Re7 37.Raxd6 Rxe3 38.Qf6 Re7 39.Kh2 Rb8 White is somewhat better.

30.fxe5 dxe5

30...Rxe5 was not clearly better. It does not lose, but neither does it allow Black to force a draw: 31.Qd4 Rd8 32.c5 bxc5
33.bxc5 f6 34.c6! (after 34.Rxd6 Rxd6 35.Qxd6 Qxd6 36.cxd6 Rd5 Black has drawing chances) 34...Rxd5 35.Qxd5
and White can press for more.
31.c5!

This pawn ties Black’s heavy pieces to the task of preventing its advance.

31...Re6 32.Qd3!

White makes sure that he remains in control of the d-file.

32...bxc5

32...Rc6?? loses to 33.Rd7 Qb8 34.Rf1.

33.bxc5

Not 33.Rdxc5?! Qd8! and Black has almost equalized.

33...Qc6 34.Rb1 Qc7 35.Rf1 Rf8

36.Kh1

Typically, Karpov improves his king first, making sure Black will have no checks.

36...Qc6 37.Rb1 Qc7 38.e4 Rb8 39.Rf1 Rb7 40.Qc3 Rb5 41.a4

Karpov gradually pushes his pawns further.

41...Rb8 42.Rc1

In case of 42.a5 Rb5 43.Qd3 Re7, Black survives.

42...Rc8

After 42...a5 43.Kh2 Rb4 (with 43...Rc6 44.Rcd1 Rb7 45.R1d3 Black’s position remains unpleasant) 44.c6 Rb6 45.Rc5
Rb4 46.Rb5 f6 (or 46...h6) 47.Qc5, White can still press very hard.

43.Rb1 Kg8 44.Rbd1 Qe7 45.Rf1 Rc7

46.a5

Karpov has made three pawn moves in the preceding part of the game, which has improved his position considerably.

46...Rec6 47.Rc1 f6

This opens the seventh rank near the king. 47...Re6 would have led to an almost losing position: 48.Rb1 Rec6 49.Rxe5
Qf6 50.Re8+ Kg7 51.e5 Qf5 52.e6+ f6 53.Re1 Rxc5 54.Qb4 Qf2 55.Rd8 Kh6 56.e7 Rc2 57.Rg1 Rc1 58.Rdd1 Rxd1
59.Rxd1 Qe2 60.Qf4+ Kg7 61.e8=N+ and White wins.

48.Qd2 Kf7

Not 48...Rxc5?? 49.Rcxc5 Rxc5 50.Rd8+.

49.Kh2!

Karpov makes his small improving king move even now. It not only improves his position, but it puts the ball in Black’s
court and keeps him guessing.

49...Ke8?

This loses at once. Of course if one has had to defend for so long, sooner or later a mistake will occur. The king is more
vulnerable on e8 than it was on the kingside.
One gets the impression that Black is losing anyway. After 49...Kg7 50.Rd6 (or 50.Qd3 Qe6 51.Rc3 (not 51.Rd6? Rxd6
52.cxd6 Rxc1 53.d7 Rc8 and Black stops the passed pawn) 51...Kf7 52.Rb3 and Black is in trouble) 50...Qf7 51.Rc3
(also after 51.Rf1 Rxc5 52.Rdxf6 Qxf6 53.Rxf6 Kxf6 54.Qd6+ Kf7 55.Qxa6 White is better) 51...f5 52.exf5 Qxf5
53.Rf3 Qc8 54.Qg5 White wins.
50.Rd6!!

This opens up the position decisively.

50...Rd7

Even now Black cannot remove the c-pawn with 50...Rxc5; after 51.Rxc5 Rxc5 52.Rxa6 his king will be caught.

51.Rd1 Rcxd6

51...Rdxd6 loses prettily after 52.cxd6 Qd7 53.Qd5 Kf8 54.Qxc6.

52.cxd6 Qe6 53.Qd3 Qa2 54.Qxa6

White wins a pawn and the d6-pawn is still choking Black. The rest is simple.

54...Qc2 55.Qa8+ Kf7 56.Qd5+ Kg7 57.Rd2 Qc3 58.Ra2 h5 59.Rd2 h4 60.Rd1 Qc2 61.a6 Qa4 62.Qd3 g5 63.Rb1 f5
64.Rb7 g4 65.hxg4 fxg4 66.Qe2 1-0

GAME 25
Karpov, Anatoly
O’Kelly de Galway, Alberic
Caracas 1970 (7)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nb8 10.d4 Nbd7 11.Nbd2 Bb7
12.Bc2 Re8 13.Nf1 Bf8 14.Ng3 g6 15.a4 Bg7 16.Bd3 d5 17.Bg5 dxe4 18.Bxe4 Bxe4 19.Nxe4 exd4 20.Nxd4 c5
21.Bxf6 Nxf6 22.Nxc5 Rxe1+ 23.Qxe1 b4 24.Rc1 bxc3 25.bxc3 Qd5 26.Ndb3 Bf8 27.Qd1 Qxd1+ 28.Rxd1 Rc8
29.Nxa6 Rxc3 30.Rb1 Bd6 31.Rd1 Bf8 32.Rb1 Nd7
The Belgian grandmaster has refused to repeat moves and he tries to trap the a6-knight. Karpov is a pawn up; if he can
find a way to solve the problem of the a6-knight he has good prospects of winning.

33.a5!

White takes the opportunity to push the pawn. Time and again Karpov recognizes the tactical potential within an
endgame.

33...Rc6?

The Belgian grandmaster chooses the right square, but the wrong piece. He should have aimed to move his knight to c6.
Some commentators have missed it, but Black still had a way to lock up the a6-knight with 33...Ne5! 34.Nb8 (34.Kf1
Nc6) 34...Bd6 35.Nd4 Rc8 36.Na6 Rd8 and White is tied up. Black has a good enough grip on the queenside with his
better pieces.

34.Rd1!

If your opponent attacks your piece, one way of handling this problem is to attack one of his.

34...Ne5

After 34...Rxa6 35.Rxd7 Bb4 36.Rd5 White wins slowly with the help of the knight and rook. He simply pushes the a-
pawn.

35.Rd5!

He keeps attacking the knight.

35...Nc4
36.Nb8!

This now clears the way for the advance of the a-pawn with tempo.
A) Incorrect would have been 36.Rd4? because of 36...Nxa5! and Black gets the pawn back;
B) Getting closer to his pieces with 36.Nac5? would not work either because of 36...Nxa5!.

36...Rc8 37.Nd7 Be7 38.Nb6

It is simpler to win with one minor piece on the board.

38...Rc6 39.Nxc4 Rxc4 40.a6 Ra4

After 40...Rc8 41.Rd7 Kf8 42.a7 Ke8 43.Rb7 Bd6 44.g3 White wins by putting the knight on c6 and then the rook on
b8.

41.Ra5 1-0

In 1970, Karpov played for the first time in a USSR Championship final. He was on plus one when he encountered
Mikenas in round 18. Karpov was to finish 5th-7th on plus three.

GAME 26
Karpov, Anatoly
Mikenas, Vladas
Riga USSR Championship 1970 (18)

1.e4 Nf6 2.e5 Nd5 3.d4 d6 4.Nf3 Bg4 5.Be2 Nc6 6.exd6 exd6 7.c4 Nf6 8.0-0 Be7 9.h3 Bh5 10.d5 Bxf3 11.Bxf3 Ne5
12.Be2 0-0 13.Nc3 Ned7 14.Be3 a5 15.Qc2 Nc5 16.a3 Nfd7 17.b4 axb4 18.axb4 Rxa1 19.Rxa1 Na6 20.Qd2 Nxb4
21.Rb1 Na6 22.Rxb7 Nac5 23.Ra7 Qb8 24.Nb5 Bd8 25.Qa2 Qb6 26.Ra8 c6 27.Nxd6 cxd5 28.Nc8 Qb7 29.Bf3 Ne4
30.cxd5 Nc3 31.Qa6 Qxa6 32.Rxa6 Bf6
White is a pawn up, which ensures him of an advantage. Nevertheless Karpov finds an uncompromising way of
demonstrating a forced win.

33.Ra8!

White is going to harass Black on his back rank.

33...Ne5 34.Bc5!

This forces the win.

34...Rd8 35.Bb6 Nxf3+ 36.gxf3 Rf8

In the event of 36...Re8, 37.d6 h5 38.d7 Rf8 39.Ne7+ is the final touch.

37.Bc5 Rd8 38.Be7 Re8 39.d6

The preceding couple of moves demonstrate Karpov’s impressive handling of his pieces.

39...Bxe7
40.d7

Winning an exchange, and now it is all over.

40...Kf8 41.dxe8=Q+ Kxe8 42.Nd6+ Kd7 43.Nxf7 Ke6 44.Nd8+ Kd7 45.Nb7 Ne2+ 46.Kh2 Ke6 47.Ra4 Bf6
48.Nc5+ Kf5 49.Nd3 Kg5 50.Ra5+ Kh6 51.f4 Kg6 52.Kg2 h6 53.Kf3 Nd4+ 54.Kg4 Kf7 55.Ra7+ Ke6 56.f5+ Kd6
57.Ra6+ Ke7 58.Nf4 Kf7 59.Ra7+ Kg8 60.Nh5 1-0

Karpov also played in three minor Soviet team events in 1970. Altogether he played 10 games, winning 4 and drawing
6. He performed respectably but not spectacularly in these events.
In 1971 Karpov recorded an official ELO rating of 2540. This was one of his most active years. He tied a match with
Kortchnoi 3-3 and won the semi-finals of the USSR Championship with an impressive +9 =8. On the 18th Student
Olympiad in Puerto Rico he scored a superb 7½ out of 8. In the USSR Championship final he ended on plus four. The
games with Taimanov and Tukmakov are taken from this event.
Karpov shared a magnificent first place with Stein at the Alekhine Memorial in Moscow, but he played no endings there
that are suitable for this book.

GAME 27
Karpov, Anatoly
Klovans, Janis
Daugavpils USSR Ch. sf 1971 (4)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.0-0 f6 6.d4 exd4 7.Nxd4 Ne7 8.Be3 Ng6 9.Nd2 Bd6 10.c3 0-0 11.Qb3+
Kh8 12.Nf5 Bxf5 13.exf5 Nh4 14.Qxb7 Qd7 15.Qb3 Nxf5 16.Nc4 Rfe8 17.Rad1 Rab8 18.Qc2 Rb5 19.Rfe1 Nxe3
20.Nxe3 Rbe5
21.g3!

Black has a seriously damaged pawn structure on the queenside. However, his rooks are highly effective on the e-file.
The text move creates a luft and prepares to ease the pressure on the e-file. White places a fifth pawn on the colour of
the opponent’s bishop. In this case this constitutes no real danger as, rather than becoming targets, they restrict the
bishop in its movements.

21...Qe6?!

Black’s situation is not easy. He has to make an important decision: what is the best way to create counterplay to
compensate for his damaged pawn structure? It is hard to see the consequences of this queen move. Black wins a tempo
by hitting the a-pawn, and if he can open the e-file he hopes to invade on e2.
A) Black could have tried to open up the kingside with 21...f5, and now:
A1) In case of 22.Ng2 Rxe1+ 23.Nxe1 Qe6 24.b3 f4 25.Nf3 fxg3 (25...Qe2?? 26.Re1 wins) 26.hxg3 Qf6 Black can
open up the position enough to equalize;
A2) 22.Qa4! keeps an eye on the queenside and prevents 22...f4, e.g. 22...a5 (22...Qf7 23.Ng2) 23.Ng2
analysis diagram

... and White keeps the better prospects.

B) With 21...c5! Black could aim to gain further space with ...c4 and then possibly exert pressure on the b-file. This
move raises the stakes. Black either achieves what he wants on the queenside or he weakens his pawns further:
B1) 22.c4 f5 23.Ng2 Rxe1+ (after 23...Re4 Black’s position also looks tenable) 24.Rxe1 Rxe1+ 25.Nxe1 Qe6 and
Black should hold, as White’s king will be exposed to checks;
B2) 22.b3 and now:
B21) 22...c4 23.b4 (in the event of 23.bxc4 Qc6 Black easily has enough compensation for the pawn) 23...a5 (after
23...Re4 24.Rd5 Qe6 25.Red1 Kg8 26.R5d4 White consolidates) 24.Rd4 (or also 24.a3 Qe6 25.Rd4 c5 26.Rxc4 and
Black has little for the pawn) 24...axb4 25.cxb4 Qe7 26.Rh4 and by finding this surprising new target White keeps the
advantage;
B22) 22...a5 23.c4 (23.a4 c4 24.b4 f5 (after 24...axb4 25.cxb4 Re4 Black’s pieces are also active) 25.Rd4 Qe6 26.Red1
axb4 27.cxb4 Bxb4 28.Qxc4 Bc5 29.Qxe6 R5xe6 30.Rc4 Bxe3 and Black holds) 23...a4 (also with 23...Qc6 24.Rd5
R5e6 25.Red1 Be5, Black’s chances would be better than in the game) 24.bxa4 Ra8 25.Rb1 (in case of 25.f4 Ree8
26.Nd5 Rxa4 27.Re2 Rxe2 (27...Rea8? 28.Nxc7) 28.Qxe2 Ra8 Black is just a little worse) 25...Ree8 26.Rb5 c6 27.Rb6
Be5 28.Reb1 Bd4 29.Rb7 Qe6 and it is difficult for White to do something constructive with his extra pawn.
B3) 22.Re2
analysis diagram

B31) The thematic idea here is to push the pawn: 22...c4 23.Red2 (after 23.Rde1 Qe6 24.Kf1 Bc5 25.Ng2 Qd5 Black’s
pieces are working nicely) 23...Qc6 24.Rd5 Rxd5 25.Nxd5 Qb5 and Black is not at all worse;
B32) Another option is 22...f5 23.Rd5 (23.Rde1 Qe6) 23...f4 24.Rxe5 Rxe5 25.Nc4 Rxe2 26.Qxe2 fxg3 27.hxg3 h6 and
White is slightly better.

22.b3 Kg8

Black just prepares to centralize his king. However, this allows White to liquidate on the e-file. Active play with
22...Bc5 would not be convincing either: 23.Ng2 a5 (in case of 23...Kg8 24.Rxe5 fxe5 25.Qe4 Rf8 26.Rd2 White
neutralizes Black’s direct play and slowly builds up the pressure on Black’s pawns) 24.Rxe5 Qxe5 25.Re1 Qh5
26.Rxe8+ Qxe8 27.Qf5 Bb6 28.c4 and the ending is unpleasant for Black.
23.Ng2!

Karpov has rightly envisaged that what is often a deadly weapon – heavy artillery on an open file – can be easily
defused this time. This is not the last time that Karpov plays a revelatory Ng2.

23...Re2 24.Rxe2 Qxe2 25.Rd2 Qf3

26.Kf1!!

This is a very subtle move. In endgames, centralization of the king is a common event. At first glance it seems
premature here, with so many pieces on the board. In fact, the text has a most unusual purpose.
26...Re5

A) 26...c5 27.Qd3 Qxd3+ 28.Rxd3 c4 29.bxc4 Rb8 30.Ke2 Kf7 31.Ne3 and White is better in spite of the fact that his
extra pawn is a doubled one;
B) 26...Qh5 27.h4 (after 27.Re2 Re5 28.h4 Qf3 White is tied up) 27...Re5 28.Qd3 and White is slowly putting Black’s
pawns under close scrutiny.

27.Qd3!

Now we can see the depth of Karpov’s king move. The second rank has been protected from being invaded.

27...Qxd3+

After 27...Qh5 28.Qxa6 Qxh2 29.Qc4+ Kf8 30.Qh4, despite the fact that the pawns are on both sides of the board and
Black possesses the bishop, the endgame still favours White as he already has a passed pawn and Black has the doubled
c-pawns.

28.Rxd3

With very neat play White has neutralized Black’s activity. Now he can concentrate on Black’s queenside pawns.

28...Kf7 29.Ne3

Karpov starts looking for the best square for his knight.

29...Ke6 30.Nc4 Rh5 31.h4

Four white pawns are on the Black bishop’s colour, but they are not really vulnerable.

31...Bc5
32.Nb2!?

Karpov improves his knight and consolidates his position before he launches his initiative.
Going for the win of the a-pawn with 32.Rd8!? would have given excellent winning chances as well, e.g.:
A) 32...Rd5 33.Re8+ Kf7 34.Ra8 Rd3 35.Rxa6 Bb6 (35...Rxc3 36.Rxc6) 36.Nxb6 cxb6 37.Rxb6 Rxc3 38.Ke2 Rc2+
39.Ke3 Rxa2 40.Rxc6 and White has decent winning chances;
B) 32...Rf5 33.f4 g5 34.Re8+ Kf7 35.Rc8 Bb6 36.Nxb6 cxb6 37.Rc7+ and White wins;
C) Black’s best option is to look after the queenside first: 32...a5! 33.Ra8 Rf5 34.f4 Bb6 (not 34...g5? 35.Rxa5 gxh4
36.b4 Be7 37.Rxf5 Kxf5 38.a4 and the a-pawn promotes) 35.Ke2 and White is better, but Black is not yet lost.

32...Rf5?!

Black doesn’t smell a rat. The rook looks so active on the fifth rank...
Black could have ‘freed’ the rook from the fifth rank with 32...Rd5!?. There could follow 33.c4 Rd6 34.Ke2 and now:
A) With 34...Kf5 35.Rf3+ Kg4 – or alternatively 35...Ke6!? 36.Nd3 Bd4, and now the black pieces are rather active –
the king becomes active, but it also becomes too much of a target: 36.Rf4+ Kh3 37.Nd3 Bd4 38.h5 g5 39.hxg6 hxg6
40.Rh4+ Kg2 41.Ne1+ Kg1 42.Nf3+ Kg2 43.Rh2 mate;
B) 34...h5!? 35.Rxd6+ (after 35.Rd1 Kf5 Black obtains some counterplay) 35...cxd6 36.Nd3

analysis diagram

36...Kf5 and Black is not in danger in this ending.

33.Rd2 h5

Black gives up the a7-g1 diagonal.

It was possible to maintain the bishop on the diagonal with the aid of 33...Rd5, e.g. 34.Nd3 Kf5 35.Ke2 Bb6 36.b4! (not
36.Nb4? Rxd2+ 37.Kxd2 Bxf2; or 36.c4 Rd4) 36...Rd6 (White keeps an extra pawn after 36...c5 37.bxc5 Ba5 38.Nb4
Rxc5 39.Nxa6, although it is not easy to convert it into a win) 37.c4 Bd4 38.c5 Rd7 39.Kf3 and White is somewhat
better.

34.Nd3!

It was not that difficult to position the knight here, but to understand fully the power it will wield in the remainder of the
game required very subtle judgement.

34...Bd6

Not 34...Bb6? 35.Nb4.

35.Re2+ Kd7

36.Re3!

It appears that White is simply defending his pawn. However, there is more to the text than that. Karpov sets up a most
wicked trap at the same time.

36...g5?

Black is still harbouring dreams of activity on the kingside.

A) After 36...Rd5 37.Ke2! (37.c4?! Rd4) 37...g5 38.Re4 Black’s rook would be in danger;
B) Better would have been 36...Rb5! (Black’s rook must leave the fifth rank at some stage) 37.Re4! (after 37.c4 Rb8
White is obviously better, but Black is not without defensive chances) 37...c5 38.Ke2 and White will cause a lot of
problems on the queenside with his rook and knight. It is hard to tell whether he has a clear win, but it is easy to see that
Black will be made to suffer.
37.c4!!

Out of the blue White prepares to trap Black’s rook. This is a fantastic idea. Black still has squares on the fifth rank, but
not for too long.

37...c5

This stifles the rook further, but the white moves b4 and a3 would have had the same effect anyway. If 37...gxh4
38.gxh4; or 37...g4 38.Ke2 Ra5 39.a4 f5 40.f4 and Black’s rook will be hunted down by Ra2 and b3-b4.

38.Kg2 c6
39.f3!

Suddenly the threat is 40.g4, trapping the rook.

39...gxh4 40.gxh4 Bf4

This merely postpones f3-f4 with the king’s inevitable march to e4.

41.Re4 Bd6 42.f4 1-0

Black loses the rook with Kf3 after which the white knight attacks the rook. Karpov released Black’s grip by the rooks
and the queen beautifully. His entrapment of the rook was accomplished with a stunning idea. This game is a true
masterpiece!

GAME 28
Karpov, Anatoly
Taimanov, Mark
Leningrad USSR Championship 1971 (1)

1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 5.Nb5 d6 6.Bf4 e5 7.Be3 Nf6 8.Bg5 Be6 9.N1c3 a6 10.Bxf6 gxf6 11.Na3
Ne7 12.Nc4 d5 13.exd5 Nxd5 14.Nxd5 Bxd5 15.Ne3 Bc6 16.Bc4 Qxd1+ 17.Rxd1 Rc8 18.Bd5 Bxd5 19.Rxd5

19...Ke7!?

Taimanov centralizes his king with a remarkable plan. Black would still have been somewhat worse if he had played
19...Bc5. One has to praise Taimanov for adopting this positive approach. Before this game he had lost 6 times in a row
to Fischer during the World Championship Candidates’ match and according to the database he had lost one against
Spassky in USSR and one to Stein before that, before drawing with Smyslov. He had faced some really great players,
yet all these losses must have been hard to take.
20.Ke2

White has no time to fix the f6-pawn with 20.g4, because of 20...Ke6 21.Rd2 h5 and Black becomes very active.

20...Ke6! 21.Rhd1 f5!

Black has activated his king and he has play in the centre.

22.g3 f4!

Opening up the position for his bishop. 22...Rg8 would have allowed White time to consolidate with 23.c3, and Black
does not manage to get rid of the doubled pawns. Also after 23.f4 f6 24.c4 b6 25.b3 White would be slightly better.

23.gxf4 exf4 24.Ng2 Rxc2+

Black discards the weak pawn and manages to open up the position. But it has cost him many tempi.

25.Kf3!

Karpov has brought his pieces into the fray. He can now press, as he is better developed.

25...Bc5

Taimanov keeps playing actively and catches up with his development. However, it was possible to grab the pawn as
well: 25...Rxb2 26.Nxf4+ Kf6 27.Re1! and now:
A) 27...Rxa2?? – taking the second pawn is cruelly punished. With 28.Nh5+! Kg6 29.Rg1+ Kh6 30.Nf4 White catches
Black’s king;
B) 27...Rc2 (Black holds by improving this rook) 28.Nh5+ Kg6 29.Rg1+ Kh6 30.Nf4 Rc5! 31.Rd7 Rc3+ 32.Ke4 Rc4+
and Black can keep checking;
C) 27...Rb5 28.Nh5+ Kg6 29.Rg1+ Kh6 and White cannot catch Black’s king.
26.Nxf4+ Kf6 27.Nd3

This placement of the knight is typical for Karpov.

27...Rc8 28.Rd7!

Rooks to the seventh!

28...b5

29.Re1!

Karpov improves his second rook as well. This makes two moves extremely unpleasant to deal with for Black. White’s
main idea is Re4, but the text also enables Ne5 in certain cases. The position is extremely rich in tactical motifs.

29...Kg7?

Karpov’s powerful moves have generated enough pressure to force a mistake from Taimanov. Black wants to avoid the
harassment of his king, but he achieves this aim only temporarily. Having brought his king into the centre, he retreats
and now his downfall is imminent. The punishment for this mistake is surprisingly harsh. The legacy of recent defeats
by Taimanov may have played a role. One ending against the great American, for example, had been very unfortunate.
A) Preventing Re4 with 29...Rc4 would not have solved the problems, because 30.Ne5 doesn’t work at once; after
30...Bb4! 31.Rxf7+ (or 31.Re2 R4c7 and Black is safe) 31...Ke6 32.Re2 (32.Re3 R4c5!) 32...Rc2! (32...R4c5 33.Rf4!)
33.Re4 Bc5 34.Rxh7 Rxf2+ 35.Kg3 Kd5 36.Nd3 Kxe4 the position is drawish. But 30.a3!, preparing 31.Ne5, gives
White an edge.
B) Moving the pawn from the seventh rank with 29...h6 was possible as well. After 30.Re4 Rc4 Black holds;
C) Black can also improve his rook with 29...Rc6, which enables the bishop to help. Most importantly, it keeps the king
in the centre as well.
C1) Now, after 30.Re4 Bd6! (not 30...Rc4 in view of 31.Ne5!, forking the rooks) 31.Rh4 Kg7 (it is very hard to spot
why the pawn has to be defended from this square, but after 31...Kg6? 32.Rd4! Bxh2 33.Kg2 Bb8 34.Nb4 White
would win the exchange) 32.Rd4? (32.Rg4+ is equal) 32...Bxh2 33.Kg2 Bb8 34.Nb4? and this is the difference: here,
34...Rg6+! is possible.
C2) 30.Ne5 Rxf2+ 31.Kg3

analysis diagram

C21) After 31...Rxb2 32.Rxf7+ Ke6 33.Rf4! White wins an exchange by brilliantly protecting the king from the bishop
with the rook;
C22) 31...Rc2!! 32.Rxf7+ (32.Kf3 Rf2+=) 32...Ke6 33.Rf4 Rc7 (now it becomes clear that 31...Rc2 was played to
reduce the power of the discovered knight check) 34.b4 Bd6 35.Nf7+ Kd7 36.Rd1 R7c6. Both sides have pinned pieces
and neither player can take charge in this remarkable position.

30.Re4 Rc4 31.Ne5 Rxe4 32.Kxe4 Kg8?!

As we have seen before, Taimanov is looking for active play. After 32...Rf8 33.f4 Kg8 Black’s position would have
become passive. However, this might well have been better than the text.
33.f4!!

Karpov postpones the pawn capture and avoids the exchange of the knight for the bishop. Why does he want to keep the
knight when the pawns are far apart and the bishop is usually superior in such an endgame? Soon we will see. After
33.Rxf7 Bd6! 34.Ra7 Bxe5 35.Kxe5 Rc2 Black has excellent chances to draw the rook ending.

33...Bf8 34.Nxf7 Rc2 35.Ng5!

Karpov needs the knight for a successful attack on the black king.

35...Bh6

In case of 35...Rxh2, the pinning 36.Rd8 wins.

36.Ne6 Rxh2
37.Kf5!

Now White catches Black’s king.

37...Bxf4

This only postpones the end. Materially speaking, Black has a chance as he has hopes of removing all the pawns, but his
king only enjoys a brief respite. If 37...Rg2, 38.Kf6 Rg6+ 39.Ke7 wins.

38.Kxf4

This is not the most direct win. The more accurate recapture 38.Nxf4! would have been much faster, as Furman,
Karpov’s trainer at the time, has pointed out.
After the capture with the knight, White checkmates after 38...Rxb2 39.Kf6 h6 40.Rg7+ Kf8 41.Ne6+ Ke8 with 42.Re7.

38...Rxb2 39.Rg7+!

With the help of the mating net White gains an important tempo.

39...Kh8 40.Ra7 h5 41.Rxa6 b4 42.Nd4

Going after the king with 42.Kg5 b3 43.axb3 Rxb3 44.Kg6! (not 44.Ra8+ Kh7 45.Ra7+ Kg8 46.Kg6 Rg3+ 47.Ng5
Kf8 and Black holds) 44...Rb8 (44...Rg3+ 45.Ng5) 45.Kxh5 wins immediately, as computer programs demonstrate.
But Karpov understandably wants to keep the pawns on the board.

42...Rg2 43.Kf3 Rd2 44.Ke3 Rb2 45.Kf4 Rd2

45...Rg2 would have prolonged the resistance.

46.Nf5 Rb2?
46...Rg2 would have avoided immediate collapse.

47.Kg5 b3 48.Rh6+!

By 48.axb3 White could also win without the pawn: 48...Rxb3 49.Kg6 Rb8 50.Ne7! h4 (50...Re8 51.Kf7) 51.Kh6 h3
52.Ng6+ Kg8 53.Ra7 and Black gets checkmated.

48...Kg8 49.Kf6 1-0

Taimanov played very imaginatively with his king for quite some time. One mistake with the king, however, brought
about disaster.
 GAME 29
Karpov, Anatoly
Tukmakov,Vladimir
Leningrad USSR Championship 1971 (18)

1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 5.Nb5 d6 6.c4 Nf6 7.N1c3 a6 8.Na3 Be7 9.Be2 0-0 10.0-0 b6 11.Be3 Bb7
12.Rc1 Ne5 13.Qd4 Ned7 14.f3 d5 15.exd5 Bc5 16.Qd2 Bxe3+ 17.Qxe3 exd5 18.Rfd1 Re8 19.Qf2 Rc8 20.Rc2 Qe7
21.Bf1 Qd6 22.Rcd2 b5 23.Nxd5 Nxd5 24.cxd5 Nf6 25.Bd3 Bxd5 26.Be4 Re5 27.Bxd5 Rxd5 28.Rxd5 Nxd5
29.Qd4 Qb6 30.Qxb6 Nxb6 31.Rd6 Rb8 32.Rc6 Kf8 33.Nc2 Rc8 34.Rxc8+ Nxc8 35.Nb4 a5 36.Nc6 a4 37.Kf2 Ke8
38.Ke3 Kd7 39.Ne5+ Ke6

40.Kd4

White’s king is stronger than Black’s, but it is a very hard task to exploit this advantage. Actually White will need a bit
of help.

40...f6

If 40...Ne7 41.f4, and Black has to move the f-pawn at some stage.

41.Nd3 Kd6 42.Nf4 g6

Black could have done without ...g7-g6, but he cannot solve all of his problems. After 42...f5 43.Nd5 Na7 44.Nb4 the
position is still unpleasant for him, for example 44...g5 45.g4 fxg4 (after 45...f4 46.Ke4 Ke6 47.Nd5 White is still
pressing) 46.fxg4 h6 47.h3 Nc6+ 48.Nxc6 Kxc6 49.b4 and White wins. Of course Black can play differently, but this
variation shows that it is not easy for him.

43.Nd5 f5 44.g4!

The pawn moves closer to the promotion square and plans to fix the h7-pawn on its square. It certainly puts the pressure
on Black.

44...Ne7?

One might think this was not actually played in the game – just a mistake with the entering of the moves. It is very
likely that Tukmakov played 44...Na7 here. We’ll make that assumption.

45.Nb4

If Black had actually played 44...Ne7, then 45.Nxe7! would have brought an elementary win. Players of this calibre
hardly ever miss something as simple as this. After 45...Kxe7 46.Kc5 (or also 46.gxf5 Kf6 (46...gxf5 47.Ke5) 47.fxg6
hxg6 48.Kc5 and wins) 46...fxg4 47.fxg4 Kf6 48.Kxb5 White would easily win the race.

45...fxg4 46.fxg4 Nc8

The knight probably went here from a7; if it had been on e7, 46...h5 or 46...g5 would have equalized.

47.Nd3 Ne7

With 47...Nb6 48.b3 axb3 49.axb3 Nd5 50.Ne5 Nf4 51.Nf7+ Ke7 52.Ng5 h6 53.Ne4 Ke6 54.Kc5 Ke5, Black should
hold as well with his active king.

48.Ne5
48...Ke6

48...h5 probably draws, as it reduces the material in Black’s favour. However, the prospects were not pleasant either, as
it is hard to calculate the lines with the clock ticking: 49.h3 (after 49.g5 Ke6 50.Ke4 Kd6 51.Nf3 Nd5 52.a3 Nc7
53.Ne5 (53.h4 Ne6) 53...Ne6 54.h4 Nc5+ Black keeps a perpetual) 49...hxg4 50.hxg4 Ke6 51.a3 (51.Nxg6 Nxg6
52.Kc5 Ne5; and in case of 51.Nf3 Kd6 52.Ng5 Nc6+ 53.Ke3 Kd5 Black is safe) 51...Kd6 52.Nf7+ Ke6 53.Ng5+
Kd6 54.Ne4+ Kc6 55.Ke5 Nd5 56.g5 b4 (not 56...Ne3 57.Kf6 Nc4 58.Kxg6 and Black is in trouble) 57.axb4 Nxb4
58.Nc3 (58.Kf6 Nd3) 58...Nd3+ 59.Kf6 Nxb2 60.Kxg6 a3 61.Kf5 (61.Kf7 Nd1 62.Na2 Ne3 63.g6 Nf5 is a draw)
61...Nd1 62.Na2 Ne3+ and Black holds.

49.a3 Nd5?

This loses, but Black’s position is not easy to handle.

A) Better was 49...h5 and because of the limited material on the board, Black can survive this unpleasant ending: 50.h3
(50.g5 Kf5) 50...hxg4 51.hxg4 Kd6 52.Nf7+ Ke6 53.Ng5+ Kd6 54.Ne4+ Kc6 (in case of 54...Ke6 55.Kc5 Ke5
56.Nf2 Kf4 57.Kxb5 White obtains winning chances) 55.Ke5 Nd5 (not 55...b4 56.axb4 Kb5 57.Kf6 Nd5+ 58.Kxg6
Kxb4 59.g5 Kb3 60.Nc5+ Kxb2 61.Nxa4+ and White wins) 56.g5 b4 57.axb4 Nxb4 58.Nc3 Nd3+ 59.Kf6 Nxb2
60.Nxa4 Nxa4 and Black holds;
B) 49...Kd6 50.Nf7+ (in the event of 50.g5 Nf5+ 51.Ke4 Kc5 Black’s king is also active) 50...Kc6 51.Ke5 (51.h4 h5;
and after 51.h3 Nd5 52.Ne5+ Kd6 53.Nd7 Nf4 Black holds) 51...Kc5 52.Kf6 Nc8 53.Kg7 Kc4 and Black is not
slower in the pawn race.
50.Nd7!!

Suddenly White’s advantage can be transformed into a win. Tukmakov may have missed this nice move.

50...Kd6

The attempt to trap the knight with 50...Nf4 51.Nf8+ Kf7 is not effective: 52.Nxh7 Nh3 53.Kc5 Kg7 54.Kxb5 Kxh7
55.Kxa4 Nf2 56.Kb5 and White’s pawns decide the issue.

51.g5!

Fixing the h7-pawn.

51...Ne7

The alternative 51...Ne3 52.Kxe3 Kxd7 53.Kd4 Kd6 54.h4 would force Black into zugzwang: 54...Kc6 (54...Ke6
55.Kc5) 55.Ke5 and White wins the race.

52.Nf8

This wins the pawn and proves to be decisive.

52...Nc6+ 53.Kc3 Ne5 54.Nxh7 Nf3 55.Kb4 1-0

GAME 30
Karpov, Anatoly
Markland, Peter
Hastings 1971/72 (15)

1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3+ 6.bxc3 Qc7 7.Nf3 Ne7 8.a4 b6 9.Bb5+ Bd7 10.Bd3 Nbc6 11.0-0 h6
12.Re1 Na5 13.Qd2 Rc8 14.h4 0-0 15.Qf4 f5 16.exf6 Rxf6 17.Qxc7 Rxc7 18.dxc5 bxc5 19.Ne5 Bc8

20.c4!

As usual Karpov plays consistently, he opens the position further.

20...Nac6 21.Bb2 Nb4

22.a5!

Did Karpov already anticipate the glorious role of this pawn, or did he just want to prevent ...a7-a5? Who knows? We
have already seen the Sazontiev game, where he employed the same plan, but here it is even less automatic.

22...Rf8 23.Ba3!

White sacrifices a pawn. The reason is not immediately obvious.

23...dxc4 24.Nxc4 Rf4 25.Nd6

After 25.g3 Rxc4 26.Bxc4 Nxc2 Black is safe.

25...Nxd3 26.cxd3 Rxh4 27.Ne4 Rh5 28.Rec1 Bb7?

With 28...Rd5 Black could bring back his rook, e.g. 29.Rc3 a6 (in some cases Black must put the pawn on his own
bishop’s colour; also in case of 29...Nf5 30.Rxc5 Rcxc5 31.Bxc5 a6) 30.Bxc5 Ng6 31.g3 e5. In both these lines White
has a very small, if any, advantage.

29.Nxc5 Bd5

The bishop appears to be more active on d5 than on c8, but in fact it would have been better to keep it on the back rank.

30.f3 Rf5

The rook has to return from the edge sometime.

After 30...Nc6 31.a6 Nd4 32.Kf2 e5 33.Rab1 Rf5 34.Rb8+ Kh7 35.Rd8 White would be slightly better.

31.a6!!

This is the introduction of a very deep plan. White fixes the a7-pawn on the colour of his bishop.

31...Rf7

A) 31...Rf8 32.Nxe6 Rxc1+ 33.Rxc1 Bxe6 34.Bxe7 is also difficult for Black;
B) After 31...Nc6 32.Ne4 Rff7 33.Rab1 (33.Bd6 Rcd7 34.Bc5 slightly favours White) 33...Nd4 34.Kf2 White is better,
as he will gain control of the c-file with Bd6. The a6-pawn is strong.

32.Ne4 Nf5 33.Bc5 Rc8

Black can try to bring his knight to d4 in order to support the queenside, by playing 33...e5. Now, 34.Nd6 Nxd6
35.Bxd6 Rxc1+ 36.Rxc1 Rf6 37.Bxe5 Rxa6 38.Rc7 Bf7, or 34.Bf2 Nd4!, is OK for Black. But with 34.Bb4! White
uncovers a surprising weakness: 34...Rfc7 35.Rxc7 Rxc7 36.Ra5 and Black is in trouble along the fifth rank.

34.Bf2!

The continuation of an extremely deep plan. Karpov exchanges a pair of rooks in order to approach the weak a7-pawn.

34...Rfc7 35.Rxc7 Rxc7 36.Rb1! Ne7 37.Rb8+ Kh7

38.Kh2!

Not only is the plan subtle, its execution is clever as well. Again Karpov makes use of a tool that is typical for his style.
Anatoly improves his king, after which Black cannot gain a tempo by giving check and then attacking the a6-pawn.
Kasparov learned a lot about the endgame by playing against Karpov. A nice echo of the text features in the following
game, though Kasparov’s sidestep with the king has a tactical motivation.

Kasparov - Jobava
Rethymnon tt 2003 (3)
41.Kh2!

(as Jeroen Bosch has indicated in Yearbook 70, the immediate 41.Rxd5 would fail to 41...Rg4!;
and 41.Bxd5 exd5 42.Rc3 to 42...Qe4 43.Qxe4 dxe4 44.Rc8 Bd8)

41.Kh2! b4 42.Bd5 ed5 43.Rd5 Rg6 44.Rd3 Rc6 45.e6?!

45.Bxb4! Bxb4 46.e6+– Kasparov

45...Qe6 46.Re3 Qc4 47.Bb4 Re6 48.Be7 Ke7 49.Rc3 Qg4 50.Rc7 Kd6 51.Qc6 Ke5 52.Qc3 Kd5 1-0

Back to the Karpov-Markland game.

38.Kh2! Ng6?

Black is helping by diverting his own knight from the queenside.

After 38...Kg6!? Black’s position would have been unpleasant anyway. Still, this looks better than the text: 39.Nc5 (just
one of the ways to squeeze) 39...e5 40.Rd8! (not 40.Rb7?! Bxb7 41.axb7 Nc6 42.Bg3 Rf7 43.Bxe5 Rxb7 and Black is
safe) 40...Kf5 41.Rd6 and White is controlling the entire board.

39.Nc5 Rc6

This loses a tempo, but Black can do nothing useful. If 39...Ne7 40.Rb7! Bxb7 41.axb7 and the passed pawn provides
White with excellent winning chances, for example 41...Nc6 42.d4, or 42.Bg3 e5 43.d4.
Alternatively, 39...Ne5 40.Rb7!.

40.Rd8 Rc7

41.Rd7!

This is the final exchange of pieces in a remarkable series.

41...Rxd7 42.Nxd7 Bc6 43.Nb8 Bb5 44.Bxa7

Mission accomplished. The ripe apple falls from the tree. The cultivation of the orchard reveals the hand of the endgame
maestro. The rest is simple.

44...Ne7 45.Bb6 Nc8 46.Bc5 Kg6 47.a7 Nxa7 48.Bxa7

Black’s bishop traps the knight, which is usually devastating. White cannot rescue it, but the bishop is there to protect it.

48...e5 49.d4 exd4 50.Bxd4 Kf7 51.f4 g5 52.fxg5 hxg5 53.Kg3 Kg6 54.Kf3 Kf5 55.g3 1-0

Black finds himself in zugzwang.

Karpov, who had by now achieved an Elo rating of 2630, played little in 1972. At the Skopje Olympiad he showed his
endgame class several times, as well as in San Antonio, where he shared the tournament victory with Petrosian and
Portisch.

GAME 31
Karpov, Anatoly
Wirthensohn, Heinz
Skopje Olympiad Final-A 1972 (8)

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.f4 e5 7.Nf3 Nbd7 8.Bc4 Be7 9.a4 0-0 10.f5 b6 11.Be3 Bb7
12.Nd2 d5 13.Nxd5 Nxd5 14.Bxd5 Bxd5 15.exd5 Bc5 16.Qe2 Qh4+ 17.g3 Qb4 18.0-0 Qxb2 19.c4 f6 20.Kh1 Rfc8
21.Qd3 Bxe3 22.Qxe3 a5 23.Rfc1 Qb4 24.Rcb1 Qc5 25.Qxc5 Rxc5

26.Kg2! Kf8

The Swiss player decides to sit and wait. Such a plan seldom works against a world-class player. 26...g6! would have
given Black some space and kept White more busy on the kingside as well. However, White has an edge anyway: 27.g4
h5 28.h3 Kg7 (or 28...hxg4 29.hxg4 Kg7) 29.Rf1 and Black is slightly worse.

27.Kf3 Ke7 28.g4 Rh8?!

Black is consistent in the way he rejects to search for counterplay. The more active 28...h5 29.gxh5 Rh8 30.Rg1 Kf8
would have been better.

29.h4 Rf8?!

Black hopes to stop g4-g5 this way. It is not obvious at this stage that Karpov will ultimately force the rook away from
this square and create a zugzwang. In case of 29...h5 30.g5!? (White would almost overpress with this move; 30.Rg1 is
also better for him) 30...fxg5 31.hxg5 Rcc8!, suddenly it looks as though Black can obtain play against the f5-pawn, but
32.Kg3! (32.Rh1 Nc5) 32...Rcf8 33.Rf1 Rf7 34.Ra3! gives White the edge. He can also play actively with his king, i.e.
34.Kh4!.

30.Ke3

As usual, Karpov starts with small improvements.

30...h6 31.Rg1 Rcc8 32.Rab1 Rb8 33.Ne4 Rbc8

34.Rgc1!

Hereby White signals that c4-c5 may be played at any moment. But that is not the only problem for Black in this
position.

34...Rfe8

If Black wants to retain the rook on f8, the king is the only piece he can move, e.g. 34...Kd8 35.Rb5 Ke7 36.c5 and
Black is in trouble.

35.Rb5 Red8 36.Kd3 Kf8

37.g5!

This is an exemplary demonstration of how to exploit space. Karpov has first made improvements on the queenside and
now he breaks through on the kingside. With little manoeuvring space, the chances of the defender are much better if
there is only one wing to guard.

37...hxg5 38.hxg5 Ke7 39.Rg1 Rh8

In the event of 39...Rg8 40.d6+ Kf8, with his space surplus White attacks on the queenside again: 41.c5! bxc5 42.Kc4!
Rh8 43.Kd5 and White wins.

40.gxf6+ gxf6 41.Rg7+ Kd8 42.Nd6 Rh3+ 43.Ke2 Rh2+ 44.Ke3 1-0

GAME 32
Karpov, Anatoly
Browne, Walter
San Antonio 1972 (2)

1.c4 c5 2.b3 Nf6 3.Bb2 g6 4.Bxf6 exf6 5.Nc3 Bg7 6.g3 Nc6 7.Bg2 f5 8.e3 0-0 9.Nge2 a6 10.Rc1 b5 11.d3 Bb7 12.0-0
d6 13.Qd2 Qa5 14.Rfd1 Rab8 15.Nd5 Qxd2 16.Rxd2 b4 17.d4 Rfd8 18.Rcd1 cxd4 19.exd4 Kf8 20.c5 Na7 21.Ne3
Bxg2 22.Kxg2 dxc5 23.dxc5 Rxd2 24.Rxd2 Rc8 25.Nd5 Rxc5 26.Nxb4 a5 27.Nd5 Rc6 28.Ne3 Rc5 29.Nf4
29...Bh6?

This is an ill-fated move. Black should have kept as many pieces on the board as he could, to reduce the role of White’s
queenside majority.
A) After 29...Nb5 30.Rd5 Rxd5 31.Nfxd5 Ke8 32.Nc4 Nc3 33.Nxc3 Bxc3 34.Kf3 Kd7 35.Ke3 Ke6 36.Kd3 Be1
37.f4! Black would also suffer. White puts the pawn on the bishop’s colour. Normally this should be avoided, but here
it restricts the bishop:
A1) 37...Kd5 38.a3 f6 39.Ne3+ Kc5

analysis diagram
40.Nd1! and White traps the bishop in an unusual way;
A2) 37...h5 38.a3 f6 39.Nd2 Kd5 (after 39...h4 40.gxh4 Bxh4 41.Ke2 the bishop is trapped again) 40.Nf3 Bf2 and
now:
A21) 41.Nh4 Bg1 (41...Ke6 42.Nxg6 Bg1 43.Ke2!) 42.h3 Bf2 43.Nxg6 Bxg3 44.Ne7+ Ke6 45.Nc6 Bxf4 46.Nxa5
and it is hard to tell whether White can win this or not;
A22) 41.b4 axb4 42.axb4 Ke6 and White is pressing.
B) 29...Nc6! (Black defends the a5-pawn and keeps the bishop on the board) 30.Rd5 Bd4 31.Rxc5 Bxc5 32.Nfd5 Ke8
and it is quite possible that Black can hold.

30.Rd5!

Obviously, White gladly exchanges the rooks.

30...Rxd5 31.Nfxd5 Bxe3?

The American grandmaster wants to bring his king into play. He achieves this, but giving up the bishop is too high a
price. 31...Nc6 was clearly superior.

32.Nxe3

Now it is very hard to obtain compensation for White’s extra pawn on the queenside.

32...Ke7

33.Nc4!

Karpov pins Black’s knight to the defence of the a-pawn. It has been said that a good way to evaluate a knight endgame
is to remove the knights. In this case, without knights White would win easily.

33...Nc6 34.Kf3 Ke6 35.Ke3 Kd5

36.a3!

Maybe at this point Browne realized the problem with his strategy. He either has to give up the excellent placement of
his king or move his pawns.

36...Ke6

Not 36...a4? 37.Nb6+.

37.Kd3 Kd5 38.f3 h6?!

Black decides to move a kingside pawn.

Marking time with 38...Ke6 was no fun either. However, it may well have been preferable; after 39.Kc3 Kd5 40.b4
axb4+ 41.axb4 Na7 42.Kd3 Nb5 43.Ne3+ it is hard to say how good White’s prospects really are.

39.Kc3 h5 40.Kd3 f6 41.f4 g5

Now we see what the problem is with moving the pawns on the kingside: eventually the barrel runs dry. With this move
Black weakens himself terminally. In case of 41...Ke6 42.Ne3, White will invade decisively.
42.Ne3+!

Now the black king is tied to the defence of the f5-pawn and diverted from the centre. Perhaps now Browne regretted
giving up the bishop to reach d4 quickly.

42...Ke6 43.h4!

This blocks the kingside and prevents the exchange of too many pawns.

43...gxh4 44.gxh4 Ne7 45.Kc4 Ng6 46.Ng2 Kd6 47.Kb5 Kd5 48.Kxa5 Ke4

Black has counterplay, but it is not sufficient to save the game.

49.b4 Kf3
50.b5!

Now White’s passed pawns are triumphant.

50...Kxg2 51.b6 Nf8 52.Kb5 Nd7 53.a4 Nxb6 54.Kxb6 Kf3 55.a5 Kxf4 56.a6 Ke3 57.a7 f4 58.a8=Q f3 59.Qe8+ 1-
0

GAME 33
Karpov, Anatoly
Smith, Kenneth
San Antonio 1972 (10)

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e6 6.g4 h6 7.g5 hxg5 8.Bxg5 a6 9.Qd2 Bd7 10.0-0-0 Nc6 11.h4 Qc7
12.Be2 0-0-0 13.f4 Be7 14.h5 Kb8 15.Kb1 Be8 16.Bf3 Na5 17.Qe2 Nc4 18.Rhe1 Rc8 19.Rd3 Ng8 20.Qg2 Bf8
21.Rh1 Ne7 22.b3 Na3+ 23.Kb2 Nb5 24.Ncxb5 axb5 25.Qd2 Qb6 26.Bh4 b4 27.Bf2 Qa5 28.Be1 e5 29.Ne2 Nc6
30.f5 f6 31.Rd5 Qa3+ 32.Kb1 Bf7 33.Rd3 b6 34.Bf2 Kb7 35.c3 bxc3 36.Rxc3 Be7 37.Rg1 Rhg8 38.Qb2 Qxb2+
39.Kxb2 Na7 40.Rxc8 Nxc8 41.Nc3 Bd8 42.Be2 Ne7 43.Bc4 Bxc4 44.bxc4 Rh8 45.Rh1 Kc6 46.Kb3 Rh7 47.Be3
Rh8 48.Kb4 Rh7 49.Nd5 Nxd5+ 50.cxd5+ Kb7
51.Kb5

White pushes Black further back. Not only does Black have little room to move, but all his pawns are on the same
colour as Karpov’s bishop.

51...Bc7 52.a4 Rh8 53.Bd2

Karpov keeps improving his position and putting the squeeze on Black.

53...Rh7 54.Bb4 Rh8 55.Rg1

There is nothing more to improve, so White starts to prise open the position.

55...Rh7 56.Rg6 Bb8

This is the only move, which still resists, though not for long.
57.h6!

Removing the cornerstone.

57...gxh6 58.Rxf6

White has enough pawns, so he can allow Black to push his passed h-pawn.

58...h5 59.Bxd6 Bxd6 60.Rxd6 h4 61.Rxb6+ Ka7 62.Rg6 h3 63.Rg1 Rh4

After 63...h2 64.Rh1 Rh4 65.d6 Rxe4 66.Rxh2 White wins.

64.Rg7+ Ka8 65.f6

White’s f-pawn is not weaker than Black’s h-pawn.

65...Rf4 66.f7 h2 67.Rh7 Rxf7 68.Rxh2 Rf4 69.d6 Rxe4 70.Rh8+ 1-0
The Road to the World Title (1973-1974)

Before he set out on his quest for the world crown, Karpov took second place in a super strong tournament in Budapest.
At the Leningrad Interzonal he shared first place with Kortchnoi, producing a fine endgame against Smejkal which
features as Game 36. In the strong Soviet Championship final he came 2nd-6th behind Spassky.
In the Candidates’ matches in 1974, Karpov played superbly. He showed that he was already superior to Polugaevsky
(+3 =5) and next he overcame Spassky, whose style suited Karpov very well. Their overall score is 15-2, an incredible
achievement and a record against a World Champion. This match was the most impressive of the cycle. In the final he
jumped to 3-0. Then he got tired, as he would in later long matches, where this would show especially in his endgames.
Kortchnoi came back to 3-2, but in the end Karpov won by this narrow margin and became Fischer’s challenger.

GAME 34
Karpov, Anatoly
Hort, Vlastimil
Budapest Tungsram 1973 (2)

1.e4 e6 2.d4 d5 3.Nd2 Nf6 4.e5 Nfd7 5.c3 c5 6.Bd3 Nc6 7.Ne2 Qb6 8.Nf3 cxd4 9.cxd4 f6 10.exf6 Nxf6 11.0-0 Bd6
12.Nc3 0-0 13.Be3 Qd8 14.Bg5 Bd7 15.Re1 Qb8 16.Bh4 a6 17.Rc1 b5 18.Bb1 Bf4 19.Bg3 Bxg3 20.hxg3 Qb6
21.Ne2 Rae8 22.Nf4 Nxd4 23.Qxd4 Qxd4 24.Nxd4 e5 25.Nfe6 Bxe6 26.Rxe5 Bd7 27.Rxe8 Rxe8

28.f3!

Simple and very strong. White prepares to bring his king to the centre and covers the e4-square at the same time.

28...Rc8 29.Rxc8+ Bxc8 30.Kf2 Kf7 31.Ke3 Ke7

32.b4!

Black had no chance to prevent this fixture of his queenside pawns by playing 31...b4 himself, as White would have
collected the pawn with 32.Nc6.

32...g6?

This is a mistake, and a surprising one for a player of Hort’s calibre. He should not have put more pawns on the colour
of his opponent’s bishop.

33.g4!

Karpov does the same, but his move gains space and aims to fix Black’s pawns.

33...Nd7 34.f4 Nf8?

The winner of the game gives this move a question mark.

A) Better was 34...Nb6 35.Nc6+ Kd6 36.Ne5 Nc4+ when after 37.Kd4 White’s advantage is smaller than in the game.
Karpov evaluates the position as plus/minus;
B) If 34...Kd6! 35.g5 (35.f5 Ne5 36.Kf4 Bd7) 35...Nb6 36.Nb3 (Karpov evaluates the position after 36.Bd3 as slightly
better for White. It would also be possible to move the knight to e5 by means of 36.Nf3, e.g. 36...Nc4+ 37.Kd4 a5
38.bxa5 Nxa5 39.Ne5 Be6 40.g4 Nc4 41.Nxc4+ dxc4 42.f5) 36...Nc4+ 37.Kd4 Nb6 38.Nc5, White is pressing in all
lines.

35.g5 Kd6 36.Kf3 Ne6?

This mistake confirms that Black will lose. Black has too many pawns on white squares. Ironically, if he had more
pawns he might have been able to hold, as White would have problems invading.
37.Nxe6!

Naturally, White exchanges knights.

37...Bxe6 38.Ke3 Bg4 39.Bd3 Be6 40.Kd4

It is clear that very soon White will set up a zugzwang.

40...Bg4 41.Bc2 Be6 42.Bb3

Here is the first zugzwang.

42...Bf7 43.Bd1 Be6 44.Bf3

Setting up the second, and Black has no hope of survival.

44...Bf7 45.Bg4 1-0

GAME 35
Karpov, Anatoly
Vaganian, Rafael
Budapest Tungsram 1973 (12)

1.e4 e6 2.d4 d5 3.Nd2 c5 4.Ngf3 Nc6 5.exd5 exd5 6.Bb5 Bd6 7.dxc5 Bxc5 8.0-0 Ne7 9.Nb3 Bb6 10.Re1 0-0 11.Bg5
h6 12.Bh4 g5 13.Bg3 Nf5 14.Qd2 Nxg3 15.hxg3 Qf6 16.c3 Bf5 17.Qxd5 Rad8 18.Qc4 Bd3 19.Qa4 Bxb5 20.Qxb5
g4 21.Nfd4 Nxd4 22.cxd4 a6 23.Qh5 Bxd4 24.Qxg4+ Qg7 25.Qf3 Bxb2 26.Rad1 b6 27.Qb7 Rxd1 28.Rxd1 Qg4
29.Rb1 Rd8 30.Qxa6 Rd1+ 31.Rxd1 Qxd1+ 32.Qf1 Qc2
33.Qb5

White is a pawn up and Black’s kingside pawns are disjointed. However, White has the problem of how to create a
passed pawn.

33...Ba3

After 33...Qb1+ 34.Kh2 Qxa2 35.Qxb6 Bc3 White has to bring the knight to the kingside: 36.Nc1 Qd2 37.Qb8+ Kg7
38.Qf4 with winning chances.

34.Qd5 Bf8

Karpov recommends 34...Bc5! – an excellent move – as the best option for Black, e.g. 35.Qd2 (in case of 35.Nxc5 bxc5
36.a4 Qxa4 37.Qxc5 Black can hold) 35...Qf5 (35...Qe4 36.Kh2) 36.Qe2 (after 36.Nxc5 bxc5 37.Qxh6 c4 38.Qc1
Qd3 Black can draw despite the two pawns’ deficit) 36...Qd5 37.g4 Bd6 38.g3 and White can still press. It is an open
question whether he can win, though.

35.Qd2 Qe4 36.Kh2!

The typical Karpovian ‘side-step’ with the king.

36...Bc5 37.Nc1!

White brings the knight to the kingside.

37...Kg7 38.Nd3 Qd4 39.Qe2 Bd6

Karpov recommends 39...Qc4!? – a motif he likes to employ himself: setting up a pin.

40.Kh3!

White threatens to exchange queens in an unusual way.

40...Qd5 41.Nf4 Bxf4

Hanging on to the bishop with 41...Qa5 would have given Black a hard time. After 42.Qg4+ Kf8 43.a4 Qc5 (43...Ke7
44.Qh4+) 44.Qf3 Qb4 (in case of 44...Kg7 45.Nd5 Be5 46.Ne3 the knight reaches the f5-square, which should be
decisive) 45.Qa8+ Kg7 46.Nh5+ Kh7 47.Nf6+ Kg6 48.Nd5 the black king is too exposed.

42.gxf4

The queen ending offers decent winning chances.

42...Kf8 43.Kg3 b5 44.Qb2 Qd3+ 45.Kh4 Qd8+ 46.Kg3 Qd3+ 47.Kh2

The king needs a shelter.

47...Kg8 48.a3 Qd6 49.Qb4 Qf6

50.f3!

From now on White must guard against perpetual checks. The text has its merits, though. He may be able to exchange
queens from the e4-square.

50...Qh4+

The Armenian grandmaster moves the queen to the edge of the board. The move is feasible, but there must be a
particular reason for playing it. Also, he must ensure that the lady can return in time.
A) There is no clear-cut path to victory for White after 50...Qf5 51.Qd6 Kh7 (51...Qh5+ 52.Kg3 Kh7 53.Qe5 Qg6+
54.Kh3 Qa6 does not give White a forced win) 52.g4 Qc2+ 53.Kg3 Qe2 54.Qb4 Qf1 55.Qe4+ Kg8 and Black is
struggling, but he is not swept away;
B) After 50...Qc6 51.Qe7 Qc4 52.Qe8+ Kg7 53.Qe5+ Kh7 54.f5 (54.Qe4+ Kg7) 54...Qh4+ (in case of 54...b4 55.Qg3
Qc5 56.axb4 Qxb4 57.f6 Qf8 58.Kh3 White wins, I think) 55.Kg1 Qc4 White can press on.

51.Kg1 Qh5?

Black finds an awkward location for the queen. It was best to keep it where it was. In case of 51...Qf6 White has no
forced win; after 52.Qxb5 Qa1+ 53.Kf2 Qxa3 Black holds.
52.Qe7!!

Karpov tightens the noose around the black king, but that is not the only idea behind this perceptive incursion.

52...Kh7?

A) If he chooses 52...Qd5 White can swap queens with 53.Qe8+;

B) 52...Qg6 53.Qe8+ Kg7 54.Qxb5 would lose a second pawn;
C) But there was still a chance to stay in the game after Karpov’s magical idea with 52...Qf5!. Karpov gives a precisely
calculated win in his Informant analysis, but Black is not forced to follow his line all the way: 53.Qe8+ Kg7 54.Qe5+
Kg6! and Black is living dangerously as he is on the ropes now.
After 54...Qxe5 55.fxe5 Kg6, on the other hand, White wins the pawn ending, for instance 56.f4 Kf5 57.g3 h5 58.Kf2
Kg4 59.Ke3! Kxg3 60.f5 h4 61.e6 fxe6 62.fxe6 h3 63.e7 h2 64.e8=Q h1=Q 65.Qg8+.
After 54...Kg6! there is:
C1) Going for the pawn ending straight away with 55.Qxf5+ does not win: 55...Kxf5 56.g3 Ke6 57.Kf2 (or 57.g4 Kd5
58.Kf2 Kc4 59.g5 hxg5 60.fxg5 Kb3) 57...Kd5 58.Ke3 Kc4 59.g4 Kb3 60.g5 hxg5 61.fxg5 Kxa3 62.f4 b4 and the
race is even;
C2) 55.Qd6+ Kh7 and White is unlikely to have a forced win. The alternative 55...Qf6 is difficult for Black, but it
certainly creates an obstacle for White;
C3) It is possible that White has a forced win after 55.Qe4, but it is not guaranteed.
53.g4!!

Karpov virtually traps the queen, even though it doesn’t look as though it is in a cage without an escape.

53...Qh3

Black can hold on to the queen but not to the game – at least not for long.
After 53...Qg6 54.Qe4 f5 (after 54...Kg7, 55.Qe5+ is the simplest win) 55.Qd3! Qb6+ 56.Kg2 Qc5 57.Qd7+ Kg6
58.Qe6+ Kh7 59.g5 White wins.

54.Qxf7+

Two pawns are two too many.

54...Kh8 55.Qe8+ Kh7 56.Qe4+

Nicely centralizing the queen.

56...Kg8 57.f5 Qg3+ 58.Kf1 Qh3+ 59.Ke2 Qg2+ 1-0

GAME 36
Smejkal, Jan
Karpov, Anatoly
Leningrad Interzonal 1973 (16)

1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 5.Nc3 Qc7 6.Be2 a6 7.0-0 Nf6 8.Be3 Bb4 9.Na4 0-0 10.Nxc6 bxc6 11.Nb6
Rb8 12.Nxc8 Rfxc8 13.Bxa6 Rd8 14.Bd3 Bd6 15.Kh1 Be5 16.c3 Rxb2 17.Qc1 Ng4 18.f4 Nxe3 19.Qxb2 Bxf4
20.Qf2 Nxf1 21.Rxf1 e5 22.g3 Qd6 23.Be2 Bg5 24.Qxf7+ Kh8 25.a4 Be7 26.a5 Rf8 27.Qc4 Rxf1+ 28.Bxf1 Qf6
29.Kg2 Qf8 30.Be2 Bc5 31.Bg4 Qf2+ 32.Kh3 d6
33.Bd7

White can try to go after Black’s king with 33.Qe6!. However, this would require great alertness and precision. In his
analysis of his best games, Karpov has corrected some of the commentators who assessed this position as winning for
White. He also mentioned that with so little time left on the clock, it was impossible for Smejkal to play such a
complicated position.
The position is so complicated that a mistake, which changes the evaluation, also appeared in Karpov’s analysis.
After 33...g6 34.Qe7

analysis diagram
A) In the event of 34...h6 35.Be6 d5 (35...Qf1+ loses to 36.Kh4 d5 37.Qe8+ Qf8 38.Qxg6) 36.Qe8+ Qf8 37.Qxg6
Black is in trouble;
B) The natural 34...Be3? loses:
B1) 35.Qxd6 Qf1+ 36.Kh4 g5+ 37.Kh5 Qf7+ 38.Kh6 Qg7+ 39.Kh5 Qf7+ is a perpetual check;
B2) 35.Be6 Qf1+ 36.Kg4 (36.Kh4 g5+ 37.Kg4 (37.Kh5? Qe2+ 38.Bg4 Qxh2+ 39.Bh3 Qxh3 mate) 37...Qe2+
38.Kh3=) 36...Qe2+ (36...h5+?? 37.Kh4) 37.Kh3 Qf1+ and the white king cannot escape from the checks;
B3) 35.a6 Qf1+ 36.Kh4 h6 37.Qe8+ Kg7 38.Qd7+ Kh8 39.Qc8+ Kg7 40.Qc7+ and this time it is White who keeps a
perpetual check in hand;
B4) 35.Be2!! (White is going to place the bishop on c4, where it defends the f1-square and attacks the enemy king as
well. In those days there were no computer programs, and even the endgame maestro Karpov misses this very subtle
move in his analysis. Karpov is much more a player than a scientist. He probably analysed the game for his own
benefit, rather than to draw conclusions. In his 2006 book he does consider this bishop move. All he has to say is ‘the
fight begins again’) 35...h5 (35...Bc5 loses to 36.Bc4 Qf3 37.a6 Be3 38.Qxd6) 36.Bc4 Qf3 (in the event of 36...g5
37.Qe8+ Kg7 38.Qxh5 d5 39.Be2 White probably wins)

analysis diagram

37.Be6! (this makes it very hard for Black to avoid an immediate loss. Instead, with 37.Qe8+ Kg7 38.Qg8+ Kh6
39.Qh8+ White cannot force a win):
B41) The black queen cannot move now: 37...Qf1+ loses to 38.Kh4 Bh6 39.Qe8+ Qf8 40.Qxg6;
B42) On 37...Bg1, 38.Kh4 wins;
B43) 37...Bc5 38.Kh4 Qxe4+ 39.Kg5 Be3+ 40.Kf6 and Black gets checkmated;
B44) 37...Bd2 (this prevents Kh4, but loses anyway) 38.a6 g5 (38...Be3 loses to 39.a7 Bxa7 40.Kh4) 39.Qe8+ Kg7
40.Qg8+ Kh6 41.Qh8+ Kg6 42.Bf5+ Kf7 43.Qh7+ Kf8 (43...Kf6 44.Qg6+) 44.Qh6+ Ke7 (44...Kf7 45.Qe6+ Kg7
46.Qg6+ Kf8 47.Qf6+) 45.Qe6+ Kf8 46.Qf6+ and White wins the queen;
B45) 37...Bf2
analysis diagram

B451) If now 38.Kh4 Qf4+ and Black holds – this is the point of putting the bishop on f2;
B452) 38.a6 Bxg3! 39.Qe8+ Kg7 40.Qf7+ Qxf7 41.Bxf7 Bf2 42.Be8 d5 and Black holds again;
B453) If White tries to set up a mating net with 38.Bf7, there is time for a perpetual check: 38...Qg4+ 39.Kg2 Qxe4+
40.Kxf2 Qc2+;
B454) White has to exchange queens with 38.Qf7! in order to win. These positions are hard to evaluate. Even a player
of Smejkal’s calibre had no chance to calculate this deep during time-trouble. There follows 38...Qxf7 39.Bxf7 and
Black loses because he has no time to put his h-pawn on h4. White fixes it and as Black king has to guard it, White’s
counterpart can invade on the queenside: 39...Kg7 40.Be8 d5 (40...Kf6 41.Bxc6 g5 42.Kg2 Bc5
analysis diagram

43.h4! – fixing the pawn on h5 is the way for White to win this position) 41.Bxc6 dxe4 42.Bxe4 Kf6 (42...Be1 loses to
43.a6 Bf2 44.Kg2 Ba7 45.h4, and 42...g5 to 43.Kg2 Bc5 44.h4) 43.Kg2 Bc5 44.h4 and White wins.
Back to the position after 33.Qe6! g6 34.Qe7:
C) 34...Qf1+!! – the black queen has to occupy the f1-square. It appears that Black can achieve this at any stage, but
chess – as we all know – can be very surprising and, fortunately for us, entertaining as well. Karpov misses this check
in his 2006 book My Best Games.

analysis diagram

35.Kh4 Be3 36.Qxd6 (the pawn capture allows a perpetual, but after 36.a6 h6! the checkmate threat would force White
to give perpetual check by 37.Qe8+ Kg7 38.Qd7+ Kh8 39.Qe8+) 36...g5+ 37.Kh5 Qf7+ 38.Kh6 Qg7+ and the White
king has no way to escape from the checks.
Back to the game.
33...g6 34.Bxc6 Kg7 35.Bb5 Qb2 36.a6 Bg1 37.Qe2 Qxc3

Keeping the queens on gives better practical chances. Czech grandmaster Smejkal was a very fine player, but he often
got into time-trouble. The importance of this particular game was extremely great – an almost certain place in the
Candidates’ matches was at stake. Maybe the added tension made it more difficult to play well in zeitnot.
Alternatively, 37...Qxe2 38.Bxe2 Kf6 39.Kg2 Bc5 40.Bc4 Ke7 41.Kf3 Kf6 (after 41...Kd7 42.Bg8 h6 43.Bh7 g5
44.Kg4 the white king penetrates) 42.Ke2 Ke7 43.Kd3 Kf6 44.Bd5 Ke7 45.Kc4 Kd7 46.Bb7 and White invades on
one of the wings after Kd5.

38.Bc4 Qc1
39.Qf1?

This blunder betrays that White was indeed in time-trouble. After 39.Kg2 Black would have faced prolonged torture,
but he would have had realistic drawing chances.

39...Qh6+ 40.Kg2 Qxh2+

Black takes the free pawn.

41.Kf3 Qh5+

Black is a pawn up now, but winning still requires true endgame skill.

42.Kg2 Qh2+

The game was adjourned at this point, and there is no doubt that many Soviet players helped Karpov. The position is
very hard to analyse, especially without a computer. Karpov wrote that he was by no means certain of the win, while
Smejkal, in a conversation with a certain master, asserted that at any rate he would not lose the game.
Incidentally, Karpov was late for the next session, as his car had broken down and he had to take the metro and then
walk.

43.Kf3 Bd4 44.Bd5 Bc5 45.Bc6 Bd4

Karpov was probably looking to adjourn the game for a second time. There is no doubt that several Soviet grandmasters
were analysing the position. Soviet chess needed the young player to take on the mighty Fischer.

46.Bb7?

It is not easy to see what is wrong with this move.

Better was 46.Bd5!! g5 (after 46...h5, 47.Qg2 holds easily) 47.Kg4 h5+ 48.Kf5 Qxg3 49.Ke6 and now:
A) 49...Qf4 50.Qxf4 exf4 (after 50...gxf4 51.Kf5 White holds easily) 51.Kxd6 Kf6 (in the event of 51...f3 52.e5 f2
53.Bg2 White even wins) 52.e5+ (52.a7 Bxa7 53.e5+ Kf5 54.e6 Kf6 55.e7 Bc5+ 56.Kxc5 Kxe7 57.Kd4 also draws)
52...Bxe5+ 53.Kc5 Bb8 54.Kb6 Be5 55.Kc5 and both sides are forced to repeat moves;
B) 49...Qf2 50.Qxf2! (50.Qb5 loses: 50...g4 51.Qd7+ Kg6 52.Qe8+ Kg5 53.Qe7+ Kf4 54.Qf7+ Ke3 55.Qxh5 g3)
50...Bxf2 51.Kf5 g4 52.Bc4! and now there is no pawn blocking the way of the bishop, so it can stop the g-pawn.

46...g5!

This is an excellent idea. Black aims to expose White’s king.

47.Kg4

If 47.Bd5 h5!.

47...h5+ 48.Kf5

White can consolidate his king, but it costs him a second pawn.
Not 48.Kxg5?? Qxg3+ 49.Kxh5 Bf2 and White must give up his queen to avoid the mate.

48...Qxg3

From now on White is clearly losing in all opposite-coloured bishop endings because of Black’s two connected passed
pawns.

49.Ke6

White still has hopes of a perpetual.

49...Qf2 50.Qb5 Qf6+ 51.Kd5

51...g4!
Taking another step towards promotion and opening a route for his king to move forward.

52.Bc8 Qe7 53.Bf5 Kh6 54.Qf1 Qc7 55.Qe2 Qc5+

Thus, Black nicely blocks the check on the c1-h6 diagonal.

56.Ke6 Kg5!

The black king becomes active.

57.Qf1 Qa3 58.Qe2 Bc5 59.Qd2+ Qe3 60.Qa5 Bb6 61.Qa2 Qf2 62.Qb1 g3 63.Bh3 Kh4 64.Bg2 Qg1!

Exchanging the queens forces the win.

65.Qxg1 Bxg1 66.Kxd6 Bd4 67.a7 Bxa7 68.Kxe5 Kg4 69.Kd5 h4

With 69...Bb8 70.e5 Bc7! (70...Bxe5 was not winning; after 71.Kxe5 h4 72.Ke4 h3 73.Bf3+! White holds) 71.e6 Bd8
72.Ke4 h4 73.Ke3 h3 Black would also win.

70.e5 h3 71.Bxh3+

After 71.e6 hxg2 72.e7 g1=Q 73.e8=Q Qc5+ forces the queen exchange and wins.

71...Kxh3 72.e6

72...Bc5!! 0-1

This is a very neat win. The g-pawn will promote with check. Black badly needed this one tempo. This win was highly
significant for Karpov’s blossoming career.

GAME 37
Karpov, Anatoly
 Ribli, Zoltan
Bath Eur. Team Championship 1973 (7)

1.e4 c5 2.Nf3 d6 3.Bb5+ Bd7 4.Bxd7+ Qxd7 5.c4 e5 6.Nc3 Nc6 7.0-0 Nge7 8.Nd5 Nxd5 9.cxd5 Nd4 10.Nxd4 cxd4
11.d3 Be7 12.Qb3 0-0 13.f4 Rac8 14.Bd2 f6 15.h3 Rc7 16.Rf2 Rfc8 17.Raf1 Rc2 18.g4 a6 19.a4 R8c5 20.Bb4 Rxf2
21.Kxf2 Rc7 22.Bd2 Qc8 23.Ke2 Rc5 24.f5 Bd8 25.Qa3 Kf7 26.Rb1 Ke7 27.Kd1 Qd7 28.b4 Rc8 29.Qa2 Bb6
30.Be1 Kf7 31.Ra1 Bd8 32.Bd2 Bb6 33.a5 Bd8 34.Qa4 Qxa4+ 35.Rxa4 Ke8 36.h4 h6 37.Ra2 Kd7 38.Be1 Be7
39.Rg2 Bd8 40.Bd2 Be7 41.Rg3 Rh8 42.Kc2 Rc8+ 43.Kb2 Rh8 44.Rg2 Rc8 45.Rg1 Rh8 46.Kc2 Rc8+ 47.Kd1 Rh8
48.Be1 Rc8 49.Rg2 Rh8

50.Rb2!

Karpov returns to the queenside without having achieved anything significant on the kingside. But his threats have not
been in vain, they have tired the opponent.

50...Rc8 51.Bd2 Bd8 52.b5!

Karpov again assesses that exchanging in order to invade is more effective than squeezing the opponent.

52...axb5 53.Rxb5 Kc7 54.Ke2 Kb8?!

Black’s position is very difficult, if not losing, and allowing the exchange of rooks gives White fewer chances to go
wrong. Normally, same-coloured bishop endings should be avoided by the player who is at a disadvantage.
However, 54...Ra8 55.Kf3 would not be a great improvement either.

55.Bb4

Karpov allows the black rook to invade, but not for free. The d6-pawn has more value than the h- and g-pawns together.

55...Bc7
Ribli keeps playing passively, hoping that White cannot crack his position. Active play would also have led to a loss
eventually. After 55...Rc2+ 56.Kd1 Rg2 57.Bxd6+ Kc8 58.Bf8! Rxg4 59.d6! Rxh4 60.Be7! (with these two moves
White conquers the b7-pawn) 60...Rh1+ 61.Kc2 Ra1 (61...h5 loses to 62.Kb2! h4 63.Bxd8 Kxd8 64.Rxb7) 62.Bxd8
Kxd8

analysis diagram

Now Black would hold in the event of 63.Rxb7 Rxa5 64.Rxg7 Ra6. Therefore: 63.Kb2!! – Karpov had to see this subtle
zwischenzug in advance. It wins a vital tempo for the invasion of his king, e.g. 63...Ra4 64.Kb3 Ra1 65.Kc4 and on e6
the king will decide the issue swiftly.
Back to the game.
56.a6!

Here is another example of the fascinating versatility of the game of chess. Limiting the pawns to one part of the board
usually helps the weaker side, as the latter’s chances to build a fortress grow significantly. However, in this case the a5-
pawn was a target for Black. That is why White exchanges it.

56...Ka7 57.axb7 Rb8 58.Bd2 Rxb7 59.Rxb7+

Same-coloured bishop endings contain excellent winning chances for the side that already has the advantage.

59...Kxb7 60.g5

White refocuses on the kingside once again.

60...hxg5

Not 60...h5? in view of 61.g6 Bd8 62.Bh6.

61.hxg5 Bd8

Otherwise g5-g6 and Bh6 would win.

62.Kf3 Kc8 63.Kg4 Kd7 64.Kh5 Ke8 65.Bb4

The knock-out blow falls on the queenside.

65...fxg5

After 65...Be7 66.gxf6 gxf6 67.Kg6 White wins the f6-pawn.

66.Bxd6 Bf6 67.Bb4 Kf7 68.Bd2 Be7 69.Bxg5 Ba3 70.Bd8 Bd6 71.Kg5 1-0

White can invade after playing f5-f6 on the next move.

Ribli was especially strong in endgames, nothing can emphasize this more than the fact that he beat Karpov twice in the
endgame. This game is reminiscent of Tarrasch or Rubinstein endgames where the hapless opponents went for
desperately passive positions against the old masters.

GAME 38
Karpov, Anatoly
Uhlmann, Wolfgang
Madrid 1973 (12)

1.e4 e6 2.d4 d5 3.Nd2 c5 4.exd5 exd5 5.Ngf3 Nc6 6.Bb5 Bd6 7.dxc5 Bxc5 8.0-0 Ne7 9.Nb3 Bd6 10.Bg5 0-0 11.Bh4
Bg4 12.Be2 Bh5 13.Re1 Qb6 14.Nfd4 Bg6 15.c3 Rfe8 16.Bf1 Be4 17.Bg3 Bxg3 18.hxg3 a5 19.a4 Nxd4 20.Nxd4
Nc6 21.Bb5 Red8 22.g4 Nxd4 23.Qxd4 Qxd4 24.cxd4
24...Rac8?!

Rooks belong on open files, yet this may not be the best, as White manages to double his rooks fairly easily.
24...Kf8!? would have postponed the invasion, but not really stopped it: 25.Re2 f6 26.Rae1 and now:
A) If 26...Rac8 27.f3 Bg6 28.Re7 b6 (exchanging the bishop with 28...Be8 is too high a price to pay: 29.Rxb7 Bxb5
30.axb5 Rb8 31.Ree7 and Black is in trouble) 29.Rb7 Rc2 30.Ree7 Rxb2 31.Rxg7 Rd2 32.Rgc7 Rxd4 33.Rxb6 and
White wins;
B) 26...Bg6 27.Re7 Rab8 28.Rc7 Rdc8 29.Ree7 Rxc7 30.Rxc7 b6 and Black’s position is passive and unattractive, yet
probably this was his best bet.

25.f3 Bg6 26.Re7! b6

Karpov mentions that it was time to play actively with 26...Rc2. However, Kasparov suggests a nice variation that
enables White to keep his advantage. 27.Rae1! h6 28.Rxb7 Rxb2 29.Ree7 Rc8
analysis diagram

30.Kh2!! (the right moment to improve the king) 30...Rd2 (after 30...Rcc2?! 31.Kg3 Rxg2+ 32.Kf4 White is clearly
better) 31.Be8 (31.Kg3 Rxd4 32.Be8 Kh8 is less convincing) 31...Kh8 32.Bxf7 Bxf7 33.Rxf7 Rg8 34.Rb5 Rxd4
35.Rxa5 Rc4 36.Raa7 and White has excellent chances to win.

27.Rae1 h6?

After 27...Kf8 28.Rb7 Rb8 29.Ree7 Rxb7 30.Rxb7 Rd6 31.Kf2 White also has excellent winning chances.

28.Rb7 Rd6

According to Dragoljub Minic, 28...Rc2 was slightly better.

A) Now, going for the kill with 29.Ree7 is very dangerous for Black, but he can manage to stabilize his position:
29...Rxb2 (29...Kf8 30.f4) 30.Be8 Rd6 31.f4 (in the event of 31.Bxf7+ Bxf7 32.Rxf7 Rg6 Black probably escapes)
31...Be4 32.Bxf7+ Kh7 and now, after 33.Bh5 Rc6 34.Rxg7+ Kh8 35.Rgc7 Rxg2+ 36.Kf1 Rxc7 37.Rxc7 Rd2 the f-
pawn will be strong, but Black can live with it. In case of 33.Be6, Black will not be worse after the exchange sacrifice
33...Rxe6 34.Rxe6 Rxg2+ 35.Kf1 Rxg4;
B) 29.Re2! offers the simplest and best practical chance. It prevents Black from obtaining counterplay, even though
White loses the opportunity to create a double rook invasion on the seventh rank, e.g. 29...Rxe2 30.Bxe2 Rd6 and now:
B1) After 31.Bb5 White has a very nice advantage, as Black has no counterplay. Kasparov noticed, however, that Black
has chances to build a fortress: 31...Kf8 32.Kf2 Bc2 33.Rc7 Bh7! 34.Ke3
analysis diagram

34...g5 and Kasparov thinks that Black can survive;

B2) 31.Rb8+!?, sending the black king to the edge of the board instead of allowing it access to the centre, is a small
improvement on Kasparov’s line. Nevertheless, his superb fortress idea may still be good enough for a draw: 31...Kh7
32.Kf2 Bb1 33.Ke3 g5 and Black’s position is not an easy nut to crack.

29.Ree7 h5?

This accelerates White’s play on the seventh rank and, interestingly, removes the white g2-pawn which could
sometimes have become a target.
A) Karpov mentions 29...Rc2 as a possibility that might resist more effectively, for example 30.Rb8+ Kh7 31.Ree8
Rc1+ 32.Kh2 Bb1 33.f4 Kg6! 34.Re7 and White is still much better;
B) 29...Rc1+, driving away the king, looks like Black’s best chance to survive: 30.Kh2 Kh7 31.f4 Be4 32.f5 Rf6
33.Be8 h5 34.Bxf7 hxg4 35.Be6 Bxf5 36.Rxg7+ Kh6 37.Bxf5 Rxf5 38.Rxg4 Rf6 and White is a pawn up, but it is hard
to tell whether it is enough for a win.

30.gxh5 Bxh5 31.g4 Bg6

32.f4!

This demonstrates that Black’s position has become hopeless. The domination of the white rooks along the seventh rank
becomes unbearable.

32...Rc1+ 33.Kf2 Rc2+ 34.Ke3 Be4 35.Rxf7

This wins, but he had other choices, e.g. 35.f5 Rf6 36.Be8 Kh7 37.Bxf7.

35...Rg6 36.g5 Kh7

This obliges White to take some care, but it loses anyway.

37.Rfe7 Rxb2 38.Be8 Rb3+ 39.Ke2 Rb2+ 40.Ke1 Rd6 41.Rxg7+ Kh8 42.Rge7 1-0

The year that Karpov effectively earned the title of World Champion through his efforts over the chess board was 1974.
He just tied up the loose ends in 1975.

GAME 39
Karpov, Anatoly
Polugaevsky, Lev
Moscow Candidates Quarter Fin 1974 (4)

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 e5 7.Nb3 Be7 8.0-0 Be6 9.f4 Qc7 10.a4 Nbd7 11.Kh1 0-0
12.Be3 exf4 13.Rxf4 Ne5 14.Nd4 Rad8 15.Qg1 Rd7 16.Rd1 Re8 17.Nf5 Bd8 18.Nd4 Ng6 19.Rff1 Ne5 20.Bf4 Qc5
21.Nxe6 Qxg1+ 22.Rxg1 Rxe6 23.Bf3 Neg4 24.Rgf1 Bb6 25.Rd2 Be3 26.Bxe3 Nxe3 27.Rb1 Kf8 28.Kg1 Rc7
29.Kf2 Nc4 30.Rd3 g5 31.h3 h5 32.Nd5 Nxd5 33.Rxd5 Ne5 34.c3 h4 35.Rbd1 Ke7
36.R1d4 f6

This buries the rook on e6 for a while. Most commentators criticized this move; however, there is logic behind it. Black
wants to exchange the e-rook for the d5-rook. Black’s position is still all right. Interestingly, various chess giants have
recommended different solutions to the problem.
A) Botvinnik suggested 36...b6!? in order to gain space on the queenside before White does, e.g. 37.Rb4 Rc6 38.Be2 a5
39.Rb3 Kd8 and Black has nothing to worry about;
B) According to Kasparov in the My Great Predecessors series, 36...Rc5!? was best. It eases the pressure applied by the
two white rooks: 37.Be2 Rxd5 38.Rxd5 Nd7 39.Rxg5 Nc5 40.Rg4 Nxe4+ 41.Kf3 Nd2+ 42.Kf2 Ne4+ and Kasparov’s
move leads to a perpetual;
C) Polugaevsky considered 36...Rg6, attaching the marks ‘!?’. Kasparov does not like his move and recommends
37.Be2 in reply. However, Garry stops there and doesn’t mention what White should do against his own medicine – the
exchange of the rooks, easing the pressure: 37...Rc5!? 38.Rxc5 (in case of 38.Rb4 Rc7 39.a5 Rg8 40.Rb6 Rd8 this rook
stands better here than in the game) 38...dxc5 39.Rd5 Nd7 and Black is not worse at all. This is what was considered
before looking at Kasparov’s analysis.

37.a5!

The beginning of a plan which gains a lot of space on the queenside.

37...Rc6
38.Be2!

Once Karpov finds the right plan, he hardly ever loses focus. Polugaevsky gives ‘plus/minus’ here. It seems he just sent
to the Informant what he had calculated during the game. Actually, Black is not in trouble yet.

38...Kd8?!

It takes too much time to move the king to c7, and this is not necessarily a better location!
A) With 38...Ng6! followed by the exchange of the e-rook, Black could and should have eased his difficulties.
According to Botvinnik, Black consolidates his position: 39.Rb4 (the plan with 39.c4 that Karpov employs in the game
would not work so well here; after 39...Re5! Black stands well) 39...Rc7 40.Bg4 (40.Rb6 Nf4) 40...Re5 and Black eases
his position;
B) 38...b5! is Botvinnik’s suggestion. It is remarkable how highly the former World Champion valued the need to gain
space: 39.axb6 Rxb6 40.b4 Rc6 41.Ra5 (after 41.c4 Rc7 42.b5 a5 Black has a fortress) 41...Rxc3 42.Rxa6 Rc2 43.Ra7+
Kd8 44.b5 Nd7 45.Ke3 f5 and the position is equal, as the first Soviet World Champion has pointed out.
39.c4 Kc7?

Interestingly, the king is neither protecting the pawns, nor is it keeping the position together, but it is more of a target
itself now. Polugaevsky was already extremely short of time here.
A) Better was 39...b6!. Polugaevsky also attaches an exclamation mark to this move, but doesn’t analyse any further.
After 40.b4 bxa5 41.Rxa5 Kc7 42.b5 axb5 43.cxb5 Rc2 44.Ke3 Rc3+ 45.Kd2 Rg3 46.Bf1 Nd7 47.Ra7+ White is
better;
B) With 39...Kd7 Black could think of consolidating, but of course this doesn’t prevent White from enjoying the better
game. Also, it is hard to admit a mistake when the limitations of the plan are not obvious. 40.b4 Ke7 41.Ke3 gives
White a slight initiative, but Black is still very much in the game.

40.b4 Ng6 41.b5

With 41.Bd1 White could almost exploit the traffic jam on the queenside: 41...Kb8 42.Ba4 (42.b5 Rc5 43.bxa6 bxa6
44.Rxd6 Rxd6 45.Rxd6 Rxa5 46.Rxf6 Nf4 gives Black compensation for the pawn) 42...Rc7 43.Bb3 Ne7! 44.Rxd6 Nc6
and Black can still resist.

41...axb5

After 41...Rc5, 42.b6+ Kc6 43.Bh5 wins.

42.cxb5 Rc2
43.b6+

Black’s unfortunate king transfer allows this huge gain of tempo.

43...Kd7

After 43...Kc6 44.Ke3 Polugaevsky assesses the position as a clear win for White. 44...Rxe4+!? still creates an obstacle,
but after 45.Rxe4! (in the event of 45.Kxe4 Rxe2+ 46.Kf3 Re6 Black still resists) 45...Kxd5 46.Re8! and Polugaevsky
turns out to be correct after all. This move wins – it is followed by a5-a6.

44.Rd2!

Karpov removes the pin in the most effective way, initiating a straightforward win. Polugaevsky asks what is wrong
with stepping aside by 44.Ke3. But once again a mistake emerges in his analysis: 44...Nf4 45.Bb5+ Ke7 46.a6 Nxg2+
47.Kf3 Ne1+. Here Polugaevsky stops, evaluating the position as equal.
analysis diagram

But with 48.Kg4! Kf7 49.Kh5! (after 49.axb7? Kg6 White gets caught in a mating net) 49...Rc8 50.Rxd6 Rh8+ 51.Kg4
Nc2 52.Rxe6 Kxe6 (52...Nxd4 53.Re8!) 53.Rd2 White wins.

44...Rxd2 45.Rxd2 Re5

After 45...Rxe4 46.Bb5+ Kc8 47.Rc2+ (47.Rxd6 Re5 48.Bc4 also wins) 47...Kb8 48.a6 bxa6 49.Bxa6 Re8 50.b7 Ne7
51.Re2 wins according to Polugaevsky.

46.a6 Kc6

46...bxa6 47.Bxa6 Ne7 48.Rb2 Nc6 loses to 49.Bb5! Re8 50.Rc2.

47.Rb2 Nf4 48.a7 Ra5 49.Bc4 1-0

 GAME 40
Spassky, Boris
Karpov, Anatoly
Leningrad Cand. Semi-Final 1974 (6)

1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.Nf3 Nd7 7.Bd3 e6 8.0-0 Ngf6 9.c4 Bd6 10.b3 0-0 11.Bb2 Qc7
12.Bxg6 hxg6 13.Qe2 Rfe8 14.Ne4 Nxe4 15.Qxe4 Be7 16.Rad1 Rad8 17.Rfe1 Qa5 18.a3 Qf5 19.Qe2 g5 20.h3 g4
21.hxg4 Qxg4 22.d5 cxd5 23.cxd5 e5 24.d6 Bf6 25.Nd2 Qxe2 26.Rxe2

26...Rc8!

Karpov moves the rook to the open file. This is not just a case of occupying a file, but it also represents the initiation of
a very forceful regrouping. The d7-knight can’t be removed and as such it represents the key element of the position, as
Karpov has pointed out.

27.Ne4

Kasparov prefers a different deployment of the knight with 27.Nc4. After 27...Bg5 28.a4 f6 29.g3 the position is equal.

27...Bd8!

He improves the bishop and prepares to get closer to the centre with his king.

28.g4

With 28.Rd5 White can initiate an active plan on the queenside: 28...f6 29.Rb5 (if 29.b4, considering to remove the
superbly blocking d7-knight, Black has 29...Kf7 (after 29...b6 30.g4 Kf7 31.Kg2 the position is balanced) 30.Nc5
Nxc5 31.Rxc5 (after 31.bxc5 Rh8 32.Red2 White is tied up to defend his centre, but it is hard for Black to crack the
defence) 31...Rxc5 32.bxc5 Ke6 and Black has stabilized his centre) 29...Bb6 (29...b6 would restrict his own bishop;
30.g4 and the position is balanced) 30.a4 a6 31.Rb4 Kf7 32.a5 Bxa5 33.Rxb7 and the position is equal, even though the
a6-pawn is vulnerable.

28...f6 29.Kg2 Kf7

30.Rc1?

With this unfortunate mistake Spassky’s problems start. From now on he will be on the defensive – the one who has to
watch out. The way Karpov uncovers the drawbacks of this move betrays the hand of the endgame virtuoso.
Interestingly, no other commentators have attached a question mark to this innocent-looking rook move. It would be
quite interesting to know why Spassky exchanged the rook. There may be two reasons. One is that he thought he was
still pressing and wanted to exchange, to have a chance to invade. Alternatively, he may have been already short of
time and wanted to ease his position with an exchange. However, by playing this he facilitates the passage of Black’s
king to the centre. He will miss the force of his rook restricting Karpov’s king. White could probably still have reached
a draw on the way, though. While we were writing this text, Kasparov’s book on Karpov became available. Garry
assigns a ?! to 30.Rc1 and quotes Tal, who pointed out there was no square to invade on the c-file – a perceptive
comment.
After 30.Kg3 Bb6 (30...g6 31.Rh1 – White’s rooks are active, his position is fine) 31.f4 White is in no way worse.

30...Bb6 31.Rec2?!

The rook looks as though it is correctly positioned on the c-file, but actually it does very little there. Better was 31.Rxc8
Rxc8 32.b4 (32.Kg3 Nc5!) 32...Ke6 33.Kg3 and though White is slightly worse he can probably keep the balance.

31...Rxc2 32.Rxc2 Ke6 33.a4?!

It is understandable that Spassky wants to have the option of defending the pawn from a3. But this has a hidden
drawback.
A) If 33.Kg3 Kd5 34.Kf3 Nc5 Black gains access to the d6-pawn;
B) If 33.b4 in order to stop ...Nc5, then 33...Kd5 34.Kf3 g6. Kasparov considers this worse than the game continuation.
33...a5!!

Karpov not only stops b4, but he is planning a surprising manoeuvre.

34.Ba3

34...Rb8!!

This is vintage Karpov! It is another magnificent injection of power into the game. Black intends to open the b-file for
his rook.

35.Rc4?!

Spassky wants to redirect his knight to b5, but he cannot accomplish this plan immediately. 35.Nc3 would be bad
because of 35...Rc8 with an annoying pin. Interestingly, Karpov gives this move an exclamation mark. However, White
still has to suffer if he wants to save the game.
An alternative plan was to play 35.b4!? in order to reduce the number of pawns on the queenside. However, Black
retains the upper hand: 35...axb4 36.Bxb4 Ra8 (in case of 36...Bd4 37.a5 White stabilizes his slightly worse position)
37.Ra2 Kd5. Black’s king is menacing; however, it looks as if White has enough pieces to withstand it: 38.Kf3 Bd4
(38...Kc4 39.Be1 Kb3 (39...Bd4 40.Rc2+ Kd5 41.Rc7 and White holds) 40.Ra1 Rxa4 41.Nd2+ Kb4 42.Ne4+ Kb3
43.Nd2+ and White has a perpetual) 39.a5 Kc4 (or 39...b6 40.Ra3. White is just a bit worse here: 40...b5 41.Rb3 g6 and
Black can exert pressure, but White has decent chances to survive) 40.Be1 Nc5 41.Rc2+ Kd5 42.Nxc5 Bxc5 43.Rb2
Kc6 44.Ke4 Rd8 (44...Bxd6 45.Kf5) 45.Kf5 Rxd6 46.Kg6 Rd7 47.Bb4 and White’s active king gives him good
drawing chances.
35...Bd4! 36.f4

Spassky plays actively, but his position is already unpleasant to defend – it might even be lost. The text move keeps
Black busy but, like the alternatives, it has a drawback. It isolates the g-pawn. This is a rather obscure factor which
needs the touch of a maestro to exploit.
A) 36.Nc3 would produce a position in which White has chances to draw, but it would be unpleasant to play:
A1) If 36...Bxc3 37.Rxc3 b5 38.axb5 Rxb5 39.f3 (if White puts his king on the third rank with 39.Kf3 the rook must
defend the b3-pawn: 39...Nb6 40.Ke2 Nd5 41.Rc5 Kd7 and Black is better) 39...Nb6 40.Rc7 (after 40.Kf2 Nd5
41.Rd3 Kd7 White is having a hard time) 40...Rxb3 41.Re7+ Kd5 42.d7 Nxd7 43.Rxd7+ Kc6 44.Rxg7 Rxa3 45.Rf7
and Black is ahead in the race but White should not lose;
A2) 36...Nb6!. Black improves his knight, which has accomplished great deeds on d7 but now finds a better
square.37.Rc7 Bxc3 38.Rxc3
38...Nd5 (the knight not only prevents Rc7, but also intends to threaten White’s king) 39.Rc5 b6 40.Rc6 Kd7 41.Rc2
Rh8! (if 41...b5 42.axb5 (42.Rc5 Nf4+ 43.Kf1 bxa4 44.bxa4 Rb3 45.Rc7+ Kd8 46.Bc1 Rb1 wins) 42...Rxb5 43.Rc5
Nf4+ 44.Kh2 Rxc5 45.Bxc5 Nd3 46.Bb6 Kc6 47.Bxa5 Nxf2 Black has excellent winning chances) 42.Kg3 Rh1
43.Rd2 Nf4 44.f3 Re1 45.Kf2 Rb1 and Black wins;
B) After 36.b4 axb4 37.Bxb4 b6 38.Bc3 Bxc3 39.Rxc3 Ra8 40.Rc4 g6 White has a difficult position.

36...g6!

Not 36...b5? 37.axb5 Rxb5 38.f5+ Kd5 39.Nc3+ when the position is equal.

37.Ng3
The interesting 37.fxe5!? has been neglected by all commentators except Kasparov. He demonstrates a remarkably long
line: 37...Bxe5 38.Rc7 b6 39.g5 f5 40.Nd2 b5 (40...Bf4!? 41.Nf3 Be3 – Kasparov says Black can play for a win like
this. We think that White’s chances to hold are not bad) 41.Nf3 b4 42.Rxd7 bxa3 43.Re7+ Kxd6 44.Rxe5 a2 45.Re1
Rxb3 46.Ra1 Rb2+ 47.Kg3 Kc5 48.Ne5 Kb4 49.Nxg6 Kb3 50.Nf4 Rb1 51.Rxa2 Kxa2 52.g6 Ka3 53.g7 Rb8 54.Ne6
Rg8 55.Kf4 Kxa4 56.Kxf5 Kb4 57.Kf6 Kc4 58.Nf4 a4 59.Ne2 Kd3 60.Nc1+ Kc2 61.Na2 Kb2 and White draws.

37...exf4!

Less promising is 37...b5 38.f5+ gxf5 39.gxf5+ Kd5 40.Rc7 bxa4 (40...b4 41.Bxb4 Rxb4 42.Rxd7 Rxb3 43.Nh5 e4
44.Re7 Kxd6 45.Rxe4 and White has exchanged almost everything) 41.bxa4 Rb3 42.Rxd7 Rxa3 43.Nh5 e4 44.Re7
Ra2+ 45.Kh3 Kxd6 46.Rxe4 and White has avoided the worst.

38.Rxd4 fxg3 39.Kxg3

39...Rc8!

Karpov shows his flexibility, now he activates the rook without exchanging the b-pawns.

40.Rd3 g5!

For the second time in the game Karpov puts his pawn on the bishop’s colour. This time he fixes the g4-pawn. All the
time the f6-pawn (or ‘root’ pawn) remains well-protected!

41.Bb2 b6?!

This pawn move doesn’t spoil the position; however, it is probably not the best continuation. Karpov’s concentration
slips for a moment.
With 41...Rc6!? – going for an immediate win – he could have been even more accurate. Interestingly, nobody mentions
the possibility of 42.Ba3 Ne5 43.Re3 Rc2 44.Kh3 Kd7 45.Kg3 b6 46.Kh3 Ra2 when White is in deep trouble.
42.Bd4?

White has to give up the d6-pawn one way or the other. It is very hard to choose between candidate moves – even for an
exceptional player like Spassky. This was probably White’s final opportunity to save the game.
With 42.Rc3 White could try to exchange as many pawns and pieces as he can. It gives good practical chances but
ultimately it looks as if White will go down if Black plays precisely. 42...Rxc3+ (Karpov mentions that Black can keep
the tension with 42...Rh8. Kasparov was not impressed by this move. Unfortunately he gives no alternative. It doesn’t
look that easy for White) 43.Bxc3 Kxd6 (after 43...Nc5 44.Kf3 Nxb3 45.d7 Kxd7 46.Bxf6 Nc5 47.Bxg5 Nxa4,
according to Botvinnik White has good drawing chances) 44.b4 (Karpov could opt for this endgame later on, but he did
not. 44.Kf3 Ke6 45.Bd4 f5 46.gxf5+ Kxf5 wins for Black) and now:

A) 44...axb4 45.Bxb4+ Kd5 (45...Nc5 46.Kf3) 46.a5 (Kotov evaluates this position as a little better for Black) 46...b5
47.Kf3 (White has decent drawing chances here. Not 47.a6? when White cannot force matters anymore: 47...Kc6
48.Ba5 (48.Kf3 Ne5+) 48...Nc5 49.a7 Kb7 50.Bb6 Nd7 51.Bd4 b4 52.Kf3 Ne5+ 53.Ke4 (after 53.Kg3 b3 54.Kh3
Nd3 55.Bxf6 b2 56.Bxb2 Nxb2 57.Kg3 Nd3 58.Kf3 Nf4 Black wins) 53...Nxg4 54.Kf5 Ne5 55.Kxf6 Nf3 56.Bb2 g4
and Black wins) 47...Ne5+ (or 47...Nb8 48.Bc3 Ke6 49.Ke4 Na6 50.Bd4 b4 51.Bf2 b3 52.Kd3 and White saves the
game) 48.Kg3 Nd3 (after 48...Kc6 49.Be7 White resists. Or 48...Nc6 49.a6 Kc4 50.Be7 b4 51.Bxf6 b3 52.Kf3 Kd3
53.Bg7 and White holds) 49.a6 Kc6 50.Be7 b4 51.Bxf6 b3 52.Kf3 Kb6 53.Ke3 Nc5 54.Bd4 Kxa6 55.Kd2 Kb5
56.Kc3 Ne6 57.Bf6 and White can hold the draw;
B) 44...b5!! (a lovely breakthrough. Commentators, including Botvinnik, missed this at the time. Black creates a passed
a-pawn and obtains a winning position. It would have been a thrill to think we had presented this lovely move to the
readers first, but Kasparov also analyses it) 45.bxa5. We will first make a significant diversion to Kasparov’s analysis;
B1) He pays more attention to 45...bxa4 46.Bb4+ (White temporarily stops the a-pawn. Alternatives are 46.a6 Kc6
47.Kf3 (47.Bb2 Ne5) 47...Ne5+ 48.Ke4 Nxg4 49.Kf5 Nh6+ 50.Kxf6 g4 51.Bd4 a3 52.Kg5 a2 53.Kf4 g3 54.Kf3 Nf5
and Black wins and 46.Kf3 Ne5+ 47.Ke4 Nxg4 48.Kf5 (48.a6 Kc6 49.Kf5 Nh6+ 50.Kxf6 g4 loses) 48...Ne3+
49.Ke4 Nd5 50.Bb2

analysis diagram

50...Ke6 51.a6 f5+ 52.Kf3 Nc7 53.a7 Kd5 54.Bc1 f4 55.Bb2 Na8 56.Ba3 Ke5 57.Kg4 Ke4 58.Bc5 f3 59.Kg3 g4
60.Kf2 Kd5 wins according to Kasparov) 46...Kc6 47.Kf3 Ne5+ 48.Kg3 Nc4 49.Be7 a3 50.a6 a2 51.Bxf6 Kb6
52.Kf3 Kxa6 53.Ke2 (53.Ke4 Ka5 54.Kf5 Ka4 55.Kxg5 Kb3 56.Ba1 Kc2 57.Kf4 Kb1 58.g5 Kxa1 59.g6 Ne5
(Kasparov gives this move an exclamation mark for its beauty. He does not mention that it is not the only way to win.
There is also 59...Nb6 60.g7 (60.Kf5 Kb1 61.g7 Nc8 wins) 60...Nd5+ 61.Kf5 Ne7+ 62.Ke6 Ng8! 63.Kf7 Kb1) 60.g7
Ng6+ 61.Kf5 Ne7+ 62.Ke6 Ng8 63.Kf7 Kb1 – Kasparov) 53...Kb5 54.Kd3 (54.Kd1 Ne3+ 55.Kc1 Nxg4 56.Bg7
Nf2 57.Kb2 g4 58.Kxa2 g3 59.Be5 g2 60.Bh2 Kc4 61.Kb2 Kd3 62.Kc1 Ke2 63.Kc2 Kf1 and Black wins)
analysis diagram

54...Na3!! (Kasparov quite rightly gives this move a double exclamation mark) 55.Kd2 (55.Ke4 Kc4 56.Kf5 Kd3
57.Kxg5 Nc4 58.Kh6 Kc2 wins) 55...Kc4 56.Kc1 Kb3 57.Bg7 Nc4 58.Bf6 Ne3 wins;
B2) When I saw Kasparov’s book about Karpov I was wondering whether Kasparov had found another win, which had
taken quite some time to be found. Interestingly, he doesn’t discuss (after 44...b5!! 45.axb5) the natural moves 45...a4
46.Kf3.

analysis diagram

Playing, instead, 46.b6?!, White constrains the king: 46...a3 47.b5 a2 48.Kf3 Ne5+ 49.Kg3 and the king remains
passive, allowing White few chances (49.Ke4 Nxg4 50.Kf5 Ne3+ loses). 49...Kd7 50.Bb2 Kc8 51.Ba1 Kb7 52.Bd4
Nd3 53.Bxf6 Kxb6 54.Kf3 Nc5 55.Bb2 Kxb5 wins as Kasparov has pointed out. With ...Kc4-Kd5 the king
‘shoulders’ his way through and the knight wins the bishop from b3.
Black has several ways to try for a win here. Despite his material advantage it took me quite some time to find the right
plan:
A) 46...Ne5+ 47.Ke4 Nxg4 48.b6 a3 49.Kf5 Ne3+ 50.Ke4 Nd5 51.Bd4 looks like a win, but I have not been able to
find a decisive continuation;
B) 46...Kd5! 47.b6 Kc4 48.Bb2 Nxb6 49.Bxf6 a3 50.Ke4 (50.Bxg5 Na4! and the knight will block the bishop on c3)
50...Nd5 51.Ba1 Nf4! 52.b5 Kxb5 53.Ke3 (53.Bf6 Kc4 54.Bxg5 Ne2 loses. Also, 53.Kd4 Kb4 should win for Black)
53...Kc4 54.Kd2 Kb3 55.Bf6 a2 56.Be5 Nh3 57.Ke2 Kc2 and Black wins.
Instead of the game move 42.Bd4 or 42.Rc3, White could also consider 42.Re3+!? and exchanging on the queenside.
His position is really hard to defend, but it might not be lost by force: 42...Kxd6 43.Ba3+ Nc5 44.b4 (44.Bb2 Kd5)
44...axb4 45.Bxb4 Ra8 (45...Kd5 46.a5) 46.a5 bxa5 47.Ra3 Kd5 (47...a4? 48.Rd3+)

analysis diagram

48.Rxa5 and in this line it is difficult to prove either a draw or a win for Black.

42...Rc6 43.Bc3

If 43.Re3+ Kxd6 44.Rd3 Ke6 45.Re3+ Kd5 Black remains a pawn up.

43...Rc5!?

It is often tempting to go ahead in material. Taking the pawn with 43...Rxd6 was winning. Because of the text, Spassky
could have forced a study-like draw later on.

44.Kg2 Rc8 45.Kg3

If 45.Kf3 Rd8 46.Ke2 (46.Rd2 Ne5+ 47.Bxe5 fxe5 48.Ke4 Rxd6 transposes to a later possibility in the game) 46...Ne5
47.Bxe5 fxe5 48.Rf3 Rxd6 49.Rf5 Rd4 50.Rxg5 Rb4, Black probably wins.

45...Ne5 46.Bxe5 fxe5

47.b4?

This is the losing mistake.

Commentators at the time, including Karpov, thought that White could have saved himself with 47.Kf3!?. Karpov is
both right and wrong, as the draw can be achieved but not with this move: 47...Rd8 48.b4 (Alexander Kotov shows
what is wrong with entering the pawn ending: 48.Ke4? Rxd6 49.Rxd6+ Kxd6 50.Kf5 Kd5 51.Kxg5 e4 52.Kf4
(52.Kh4 Kd4 53.g5 e3 54.g6 e2 and Black promotes with check) 52...Kd4 53.g5 e3 54.g6 e2 55.g7 e1=Q 56.g8=Q –
White doesn’t lose the race, but the position gives him no reason to be happy: 56...Qf1+ 57.Kg4 Qg2+ winning the
queen)
analysis diagram

48...Rxd6?! is the only move the World Champion analyses. For some time I thought that Black is winning, but my final
evaluation is that White holds. The endgame is like a ‘delicatessen’ – it is surprisingly rich in ideas. It is really worth
going into in some detail, even though it is irrelevant to the evaluation of the position as the alternative given on page
121 wins outright: 49.Rb3 Rd4!. Karpov now gives an interesting variation: 50.bxa5 Rf4+ and now two king moves can
transpose with one another. It appears the position is lost in a study-like way, but White has a very sophisticated
defence:
A) 51.Kg3 (White drops the g4-pawn) 51...bxa5 52.Rb6+ Kd5 53.Rb5+ Ke4 54.Rxa5 Rf3+ 55.Kg2 Ra3 56.Ra8 Kf4
(Karpov says that White has to be very careful here) 57.a5 Kxg4 and now:
A1) Not 58.a6? Ra2+ 59.Kf1 Kg3:
A11) 60.a7 g4 61.Ke1 e4 62.Kf1 e3 63.Ke1 Kg2 64.Rg8 (or 64.Kd1 g3 65.Ke1
analysis diagram

65...e2 and Black wins) 64...Rxa7 65.Rxg4+ Kf3 and Black wins.
It seems that Black can win this interesting endgame, which occurred at move 51 in the analysis of this game. We are
not far from the time when this type of position will be dissected completely by computers. It is almost certain that
Karpov analysed the position after this match. It would be nice to know how far he got and also why he did not write
about it. In chess, knowing more than your opponents can be a key element of success. Karpov is a player who always
wanted to win very much. He succeeded like very, very few in the history of chess.
A12) 60.Rg8 g4 61.Rg6 e4 62.Rf6 (62.Ke1 Kf3 and 62.Re6 Kf4 63.Rf6+ Kg5 64.Rb6 e3 lose) 62...e3 63.Re6.

analysis diagram
Now Black wins with 63...e2+ (63...Ra1+ 64.Ke2 Kg2 65.Rg6 g3 66.Kxe3=) 64.Ke1 Kh3 65.Rh6+ Kg2 66.Rg6 g3
67.Re6 Kf3 68.Rf6+ Ke3 69.Re6+ Kd3 70.Rd6+ Ke4 71.Re6+ Kf5 72.Re8 g2 73.Rg8 Ra1+ and White loses the rook;
A2) 58.Kf2! e4 (58...Kf4 transposes to Harold van der Heijden’s analysis later on) 59.a6 e3+ 60.Ke2 Kg3 61.Rg8 g4
62.Rg6 will be discussed in the line with 51.Ke3 below;
B) 51.Ke3 Rxg4 52.Rxb6+ (52.axb6?? Rg3+) 52...Kf5 53.Rb8 Rxa4 and now:

analysis diagram

B1) 54.Rf8+ Kg4 55.Ra8 Kg3 looks really tough for White: 56.a6 g4 57.Kd3 (57.a7 e4 58.Ke2 Ra3 59.Kd2 Kg2
60.Rg8 Rxa7 61.Rxg4+ Kf3 62.Rh4 e3+ wins for Black) 57...Kg2 58.a7 (58.Rg8 Rxa6 59.Rxg4+ Kf3 60.Rh4 e4+
(60...Rd6+ wins as well) 61.Kc2 (61.Kd4 e3 wins) 61...Rd6 62.Rh3+ Kf2 63.Rh2+ Kg3 64.Re2 Kf3 wins for Black)
58...g3 59.Ke3 e4 and now:
B11) 60.Kf4 Kf2;
B12) 60.Kd2 Ra3 61.Ke2 e3 62.Ke1 Ra2 63.Kd1 Kg1 64.Ke1 g2 65.Kd1 Rd2+ (65...Kf1 66.Rf8+ Rf2 67.Rxf2+ Kxf2
68.a8=Q g1=Q+ or 68...e2+ 69.Kc2 e1=Q wins) 66.Ke1 Rd7 67.Ke2 Rh7 68.Kxe3 Kh2 wins;
B13) 60.Ke2 Ra3 61.Kd2 Kg1 62.Ke2 g2 63.Kd2 Rd3+ 64.Ke2 (64.Ke1 Rd7 65.Ke2 Rh7 66.Ke3 Kh2 loses)
64...Rd7 65.Ke3 Rf7 (or 65...Rh7) 66.Kxe4 Kf1 and Black wins.
B2) 54.Ra8! Ra3+;
B21) 55.Ke2 Kf4 56.a6
analysis diagram

B211) 56...e4? White can escape after this mistake: 57.a7 Ra2+ 58.Kf1 Ke3 (58...Kg3 59.Rg8 Rxa7 60.Rxg5+ draws)
59.Kg1 (59.Rg8 Rxa7 60.Rxg5 Ra1+ 61.Kg2 Ke2 wins, e.g. 62.Rb5 e3 (62...Ra2 63.Rb1 e3 puts White in zugzwang)
63.Rb2+ Kd3 64.Rb3+ Kd2 65.Rb2+ Kc3) 59...g4 60.Kf1 Rf2+ 61.Kg1 Rf7 62.Kg2 Kf4 (after 62...Rh7 63.Kg3 Rg7
64.Kh2 g3+ 65.Kg2 Re7 66.Kxg3 White escapes) 63.Kf2 e3+ 64.Kg2 g3 65.Rb8! (65.Kf1? Kf3 66.Kg1 e2; 65.Re8?
Rxa7 66.Rf8+ Ke4) 65...Rxa7 66.Rb4+ Ke5 (66...Kf5 67.Kxg3) 67.Kxg3 Rf7 68.Ra4 draws;
B212) 56...Ra2+. This is the simpler win: 57.Kd3 (57.Kf1 Kg3) 57...e4+ 58.Kd4 e3 59.Rf8+ (59.a7 e2 60.Rf8+ Kg3
and Black wins) 59...Kg3 60.Kxe3 Rxa6 61.Rf3+ Kg2 62.Rf2+ Kg1 63.Rf3 Rg6 64.Ke4 g4 65.Rg3+ Kf2;
B213) 56...Kg3. This serves to illustrate how Black wins if he avoids pushing the e-pawn: 57.Rg8 (57.a7 g4 58.Kd2 e4
59.Ke2 e3 60.Kf1 Ra2 61.Ke1 Kg2 and Black wins) 57...g4 58.Rg6 e4 59.Re6 (59.Kd2 Kf3 60.Rf6+ Kg2 wins)
59...Kh3 60.Rg6 g3 61.Rh6+ Kg2 62.Rg6 Kh2 63.Rh6+ Kg1 64.Re6 g2 65.Rh6 Ra2+ 66.Ke1 e3 67.Rg6 Kh2 68.Rh6+
Kg3 69.Rg6+ Kf3 70.Rf6+ Ke4 wins;
B22) 55.Kf2 Kf4 (White will be able to get the critical position against 55...Kg4 as well: 56.a6 Ra2+ 57.Ke3 Ra4
58.Kf2 (58.Re8 Rxa6 59.Rxe5 Rf6) 58...Ra3 (58...e4 59.Ke3 Kg3 60.Rg8 g4 61.Rg6 looks drawish) 59.Ra7 e4 60.Ra8
e3+ 61.Ke2 Kf4 62.Rf8+ Kg3 63.Rf6 g4 transposes to 55...Kf4) and we are at a crossroads again.
analysis diagram

My original analysis went 56.a6? Ra2+ 57.Kf1 g4 58.Rf8+ Kg3 59.Rf6 Ra1+ 60.Ke2 Kg2 and here I stopped without
daring to reveal my opinion as to whether this was a draw or not. Dutch endgame study specialist Harold van der
Heijden showed that there was a beautiful win for Black: 61.Rg6 g3 62.Rh6 Kg1 63.Rg6 g2 and now:
B221) 64.Rf6 Ra2+ 65.Ke3 Kh2 66.Rh6+ Kg3 67.Rg6+ Kh3 68.Rh6+ (68.Kf3 e4+ 69.Kxe4 Kh2 wins, but not
69...Ra1? 70.Kd3!! with a draw) 68...Kg4 69.Rg6+ Kf5 70.Rg8 Rxa6! 71.Rxg2 Ra3+ 72.Kf2 Ra2+ 73.Kf1 Rxg2
74.Kxg2 Ke4 and Black easily collects the full point in this pawn ending;
B222) 64.Rh6

analysis diagram
64...Ra2+! (Lasker would have enjoyed seeing his motif, the ladder, in practice) 65.Ke3 (65.Ke1 e4 66.Rg6 Kh2–+)
65...Kf1 66.Rf6+ Ke1 67.Rg6 Ra3+! 68.Ke4 Kf2 69.Rf6+ Ke2 70.Rg6 Ra4+ 71.Kxe5 Kf3 72.Rf6+ Ke3 73.Rg6 Ra5+
and White loses as in the famous Lasker study.
56.Rf8+!! instead of 56.a6 in the previous line B22 doesn’t look that exceptional, but White’s drawing plan is really a
particularly deep concept. This move stems from Antonio Palma, who had seen Harold’s line on the Internet. After
56...Kg4 57.Ra8

analysis diagram

the move 57...e4 is a sad necessity for Black as he is in zugzwang and has to play his pawn to e3, giving the white king
the f3-square; yet it gives the best winning chances. If 57...Ra2+ 58.Ke3 Kg3 59.Ke4 g4 60.Kxe5=; or 57...Rf3+
58.Kg2 Rf6 59.Rb8 Kf4 60.Rb4+ e4 61.Ra4 and White is safe.
Now, 58.a6 e3+ (quite incredible: once the e-pawn reaches the third rank and White’s king is not restricted to the first
rank, he can hold) 59.Ke2 Kg3 60.Rg8 g4 61.Ra8 (White can also occupy the sixth rank with 61.Rg6 as the pawn is
already on e3: 61...Kf4 (61...Kh3 62.Rh6+ Kg2 63.Rf6 (or 63.Rg6 g3 etc.) 63...g3 64.Rg6 transposes to 61.Ra8)
62.Rf6+ Kg5 (62...Ke4 63.Re6+ Kf5 64.Rxe3 Rxa6 65.Kf2=) 63.Rf8 Rxa6 64.Kxe3 Ra2 (64...Kh4 65.Kf2=) 65.Rh8!
(65.Rf2 Ra1 66.Rg2 Kh4 67.Kf2 g3+ loses) 65...Rg2 66.Rh7!=) 61...Kg2 62.Rg8 g3 63.Rg6 Kh2
analysis diagram

64.Kf3! The disadvantage of Black having the pawn on the third rank is that White can play this key king move.
64...Kg1!.
Black leaves his g3-pawn unprotected. Both white pieces can take it, but either capture would lose simply.
Here White has two drawing moves, both of which require precision:
I) 65.Ke2! g2
IA) 66.Rh6? Ra2+ 67.Kxe3 Kf1 loses;
IB) 66.Rb6! (the rook has a new function on the b-file) 66...Ra5 (66...Kh2 67.Rh6+ Kg3 68.Rg6+ Kf4 69.Rxg2 Rxa6
70.Rg8=)
analysis diagram

67.Kf3!! (67.Rh6 Ra2+ 68.Kxe3 Kf1 69.Rf6+ Ke1–+; 67.Rf6 Kh2 68.Rh6+ Kg3 69.Rg6+ Kh3 70.Rh6+ Kg4 71.Rg6+
Rg5 72.Rxg5+ Kxg5 73.a7 g1=Q 74.a8=Q Qf2+ 75.Kd3 Qd2+! 76.Kc4 e2 and Black wins) 67...Rf5+ (67...Rg5 68.a7
Ra5 69.Rg6 Rxa7 70.Rxg2+ Kf1 71.Rh2=) 68.Kxe3 Rh5 (68...Rg5 69.Kf3= Rg7 70.Rb7! and White draws) 69.a7 Kh2
70.Rg6 Rh8 71.Kd4 draws;
IC) 66.Rf6! (this is a difficult move to understand, but the rook has a special role on the f-file) 66...Ra5 67.Kxe3 Rg5
68.a7 Kh2 69.a8=Q g1=Q+ (Black promotes with check; this almost always brings the win with an open king and
heavy pieces on the board; almost always!) 70.Rf2+ and the check is blocked by a countercheck;
II) 65.Rb6! g2
analysis diagram

IIA) 66.Rc6 Ra5

IIA1) 67.Rd6? Rh5 68.a7 Kh2 69.a8=Q g1=Q wins;
IIA2) 67.Ke2? Kh2 68.Rh6+ Kg3 69.Rg6+ Kh3 70.Kf3 (70.Rh6+ Kg4 71.Rg6+ Rg5–+) 70...Rf5+ 71.Ke2 and now
71...Rf6 or 71...Rf2+ 72.Kd3 e2 73.Rh6+ Kg4 74.Re6 Kf5 75.Re8 g1=Q 76.Rf8+ Ke5 77.Re8+ Kd6 and Black wins,
or;
IIA3) 67.Kxe3? Rg5!! (67...Rh5? 68.a7 Rh8 69.Rg6 Kh2 70.Kd4=) 68.Kf3

analysis diagram

68...Rg7!!. Remarkably, if the White rook were on b6 the position would be a draw.
IIA4) 67.Rf6?! (objectively this might draw, but it’s hard) 67...Rh5 (67...Re5? 68.a7 e2 69.a8=Q e1=Q 70.Qa7++–)
68.a7 Kh2 69.a8=Q g1=Q 70.Ke2 Qg4+ 71.Qf3 Qc4+ 72.Ke1 (72.Kxe3 Re5+ loses) 72...Qc3+ and Black has
winning chances; however, I have not been able to find a forced win: 73.Ke2 Qd2+ 74.Kf1 Qc1+ 75.Ke2 Qb2+
76.Ke1 Qb4+ 77.Ke2 Rh4
analysis diagram

78.Rf7 and it is difficult to see how Black can force a win;

IIA5) 67.Rb6! Rf5+ (67...Ra1 68.Ke2; 67...Ra2 68.Kxe3=) 68.Kxe3 Rh5 69.a7 Kh2 70.Rg6 Rh8 71.Kd4=.
IIB) 66.Rf6! Ra5 (66...Ra4 67.Kxe3 draws, or 66...Kf1 67.Kg3+=) 67.Kxe3!!= Rg5 (67...Rh5 68.a7 Rh8 69.Rh6= or
67...Rh5 68.a7 Rh8 69.Rg6 Kh2 70.Kd4=) 68.a7 Kh2 69.a8=Q g1=Q+ 70.Rf2+ and White holds because of this
countercheck; this is the reason why 66.Rf6 is the saving move.
Let’s return to the analysis position after 47.Kf3 Rd8 48.b4.

analysis diagram
When I shared my concerns with Harold van der Heijden, that the evaluation of Karpov’s line might be in doubt, he
came up with a subtle win for Black: 48...axb4!!.
This move is in itself not spectacular, yet the win is quick.
49.Rb3 Rf8+!! 50.Ke3 (50.Kg3 Rf4 51.Rd3 Rd4 is the end for White) 50...Rf4 51.Rd3 Rxg4! wins as Black threatens to
exchange the rook.
We have learned from Kasparov’s My Great Predecessors that Spassky had also found this subtle tactical road to the
win.

analysis diagram

His study-like saving move is 47.Kf2!!. White approaches the centre and at the same time clears the way for his rook to
go after the g5-pawn. Kasparov writes that Spassky mentions this absolutely brilliant move in his analysis.
Unfortunately Spassky’s analysis was not available to us for examination. In case of 47...Rd8 48.Rf3 Rxd6 49.Rf5 Rd3
50.Rxg5 Rxb3 51.Rg6+ Kd5 52.g5 (thus far Spassky) 52...Rb4 53.Ke3 Rxa4 54.Rxb6 White draws without any
problem.
Spassky’s move triggered the idea to go after the g5-pawn at once: 47.Rf3!?. This gives chances, though not as clear-cut
as 47.Kf2!!. However, it helps us to appreciate Spassky’s moves even more, e.g. 47...e4 48.Rf5 Rc5 (Spassky has
driven the rook away from the eighth rank) 49.Rf8 Rc3+ 50.Kf2 Kxd6 51.Rg8 Rxb3 52.Rxg5 Rb4 53.Ke3 and White
still resists.
Let’s return to the game.
47...e4!

Black wins a big tempo by attacking the rook.

After 47...axb4 48.d7 Rd8 49.Rb3 Rxd7 50.Rxb4 Rd6 51.Kf3 White holds.

48.Rd4?

The long, fruitless defence takes its toll and with White’s next two moves he goes down without a fight.
A) After 48.Rb3 Kxd6 49.bxa5 bxa5 50.Kf2 (50.Rb5 Rc5 – Black keeps as many pawns as he can) 50...Kd5 51.Ke3
Rc5 52.Ra3 Ke5 53.Rb3 Rd5 54.Rc3 Rd3+ and Black wins as Karpov has pointed out;
B) 48.d7 would have obliged Karpov to calculate accurately. The move loses and Karpov would have overcome this
gesture as well. However, some of the ensuing positions are interesting and instructive: 48...Rd8 49.Rd4 Ke5 50.Rd2 e3
(50...axb4 51.Rd1 b3 brings White into zugzwang: 52.Rd2 e3 (52...Ke6 53.Rd4 b2 54.Rxe4+ (or 54.Rb4 Rxd7 55.Rxb2
Rd3+ 56.Kf2 Ke5 and the black king invades) 54...Kxd7 55.Rb4 Kc6 56.Rxb2 Rd3+ 57.Kf2 Rd4 and Black should be
winning with his extra pawn) 53.Rd3 Ke4 54.Rxb3 Rxd7 55.Rxb6 Rd1 56.Rb4+ (56.Re6+ Kd3 57.Rd6+ Kc2 58.Rc6+
Kd2 59.Rd6+ Ke1 60.Re6 e2 and Black wins) 56...Kd3 57.Rb3+ Kd2 58.Rb2+ Kc3 59.Ra2 Kd3 60.Ra3+ Ke4 and
Black wins – interestingly, the existence of the a4-pawn helps him here as White can give no check and his king will be
caught with ...Rg1+):
B1) 51.Rd3 Ke4 52.Rd6 axb4 53.Rxb6 Rxd7 54.Rxb4+ Rd4 55.Rb5
analysis diagram

55...Rxa4 (the position wins without the g-pawns, the method would be the same if they were on the board. With
55...Rd1 56.Rb4+ Kd3 57.Rb3+ Kd2 58.Rb2+ Kc3 59.Ra2 Kd3 Black wins as we have already seen) 56.Rb2 (56.Rxg5
Ra1 57.Rg8 e2 wins for Black) 56...Ra1 57.Rb4+ Kd3 58.Rb3+ Kd2 59.Rb2+ Kc3 60.Rh2 Kd3 and Black soon
promotes the pawn;
B2) 51.Rd1 e2 (maybe this is the simplest win. Not 51...Ke4?? 52.bxa5 e2 (52...bxa5 53.Rd6) 53.Re1 Ke3 54.axb6
Rxd7 55.a5=) 52.Re1 axb4 53.Rxe2+ Kd4 54.Re6 b3 55.Rxb6 Kc3 56.Rb7 b2 57.a5 Kc2 and Black wins. The existence
of the a-pawn again helps Black as his king can hide on a2.
C) About 48.Rd1 the same can be said as about 48.d7. In addition, White can hope to take enough pawns and sacrifice
his rook under the most favourable circumstances. Karpov would likely have avoided such minefields. 48...Rd8 49.Re1
Ke5 50.bxa5 bxa5 51.Rf1.
analysis diagram

If White can take the g5- and a5-pawns and then sacrifice the rook, Black will not necessarily win. Black has two
instructive winning moves:
C1) 51...e3 52.Rf5+ Ke4 (52...Kd4 53.Kf3 Rxd6 54.Rxg5 Kd3 55.Rxa5 Rf6+ 56.Kg3 e2 also wins) and now:
C11) 53.Rf6 Kd3 54.Re6 e2 wins;
C12) 53.Rxa5 Rxd6 (53...Rf8 54.Rf5) 54.Rb5 Rd1 55.Rb4+ Kd3 (55...Rd4) 56.Rb3+ Kd2 57.Rb2+

analysis diagram

57...Kc3 (interestingly, with this move the king goes after its opposite number!) 58.Ra2 Kd3 59.Ra3+ Ke4 60.Ra2 Rg1+
61.Kh2 Rxg4 62.a5 Kf3 63.a6 e2–+;
C13) 53.Rxg5 Rxd6 54.Rxa5 (54.Rg8 Rd1 55.Re8+ Kd3 56.Rd8+ Kc2 57.Re8 Kd2 58.Rd8+ Ke1 wins for Black)
54...Rd1 55.Ra8 e2 56.Re8+ Kd3 57.g5 (57.Rd8+ Kc4) 57...e1=Q+ 58.Rxe1 Rxe1 59.Kf4 Kd4 and Black wins.
C2) 51...Rxd6 52.Rf5+ Kd4 53.Rxa5 (53.Rxg5 e3 54.Rb5 (54.Kg2 Kd3 55.Rb5 e2 56.Re5 Kd2) 54...Rf6 55.Rb8 e2
56.Re8 Kd3) 53...e3 54.Rb5 (54.Rxg5 Re6–+; or 54.Kf3 Kd3 55.Rxg5 Rf6+ 56.Kg2 e2 57.Re5 Rf1 58.Kg3 e1=Q+
59.Rxe1 Rxe1 60.Kf4 Kd4–+) 54...Kd3! 55.Rb3+ Kd2 56.Rb2+ Kd1 57.Kf3 (57.Ra2 e2; 57.Rb1+ Kc2 58.Ra1 Rd1)
57...e2 58.Rb1+ Kd2 59.Rb2+ Kd3 60.Rb3+ Kc2 61.Re3 Kd1 and Black wins.
Let’s return to the game!

48...Ke5 49.Rd1?!

White is lost anyway, but after this Black doesn’t even have to calculate.
49.d7 Rd8 would transpose to 48.d7.

49...axb4 50.Rb1 Rc3+

This provides the opportunity for Black’s king to invade. It is all over now.

51.Kf2 Rd3 52.d7 Rxd7 53.Rxb4 Rd6 54.Ke3 Rd3+ 55.Ke2 Ra3

White resigned as he will soon be two pawns down. An e- and g- pawn with one rook on each side produces a draw only
if the king of the stronger side is cut off from the pawns.

GAME 41
Karpov, Anatoly
Pomar Salamanca, Arturo
Nice Olympiad Final-A 1974 (13)

1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.Nf3 Nd7 7.h4 h6 8.h5 Bh7 9.Bd3 Bxd3 10.Qxd3 Qc7 11.Bd2
e6 12.Qe2 Ngf6 13.c4 Bd6 14.Nf5 0-0-0 15.Nxd6+ Qxd6 16.Ba5 Rde8 17.Ne5 Qe7 18.Bc3 Rd8 19.f4 Nxe5 20.fxe5
Nh7 21.0-0-0 Ng5 22.a3 f5 23.exf6 gxf6 24.Rhf1 Rhe8 25.Rde1 Qf7 26.g4 Rf8 27.Qc2 Qg8 28.Bb4 Rf7 29.Qg6
Qxg6 30.hxg6 Rg7 31.Rxf6 Rdg8 32.Ref1 Rxg6 33.Rxg6 Rxg6

34.Rf8+ Kc7?!

As it turns out, this was careless, but it was hard to foresee the consequences.
If 34...Kd7:
A) 35.Kc2 Ne4 36.Rf7+ Kc8 37.Re7 Rxg4 (37...Nf2 38.Bd6 Rxg4 39.Be5 c5 40.Rc7+ Kd8 41.Rxc5 Rg2 42.Rb5 White
is somewhat better) 38.Rxe6 c5 39.dxc5 Nxc5 and Black holds;
B) 35.b3 Ne4 36.Rf7+ Kc8 37.Kc2 Rxg4 and Black probably holds;
C) 35.Rf4 Nh3 36.Rf7+ Kc8 37.Bd6 Rxg4 38.Be5 (after 38.Rc7+ Kd8 39.Rxb7 Rxd4 40.c5 Nf2 41.Rxa7 Ne4 Black’s
pieces arrive in time) 38...c5 (38...Rg2 39.Rc7+ Kd8 40.Rxb7 Nf2 41.Kc2 Ng4+ 42.Kc3 Nxe5 43.dxe5 and White is
pressing) 39.Rc7+ Kd8 40.Rxc5 and White is slightly better.
35.Ba5+!! b6

After 35...Kd7
A) 36.Rd8+ Ke7 37.Ra8 (37.Rb8 b6 38.Bb4+ Kf6 and Black’s king is better than in the game) 37...a6 38.Ra7 Nf3
39.Rxb7+ Ke8 40.Bc3 Rxg4 and the position is messy;
B) 36.b3 Ne4 37.Rf7+ Kc8 38.c5 Nf6 39.Rc7+ Kb8 40.Re7 Rg8 41.Bc7+ and Black’s position is uncomfortable.

36.Bd2!

Karpov increases the pressure on Black’s position.

36...Ne4

After 36...Kb7 37.Bf4 Nh3 38.Rf7+ Kc8 39.Be5 White maintains his grip.
37.Bf4+!

Here the point of the check on a5 becomes apparent – the king is trapped. Karpov creates surprisingly big trouble with
surprisingly little material.

37...Kb7 38.Rf7+ Ka8

Not 38...Ka6?? when 39.Bb8 wins.

39.Rf8+ Kb7 40.b4 Rxg4 41.Rf7+ Ka8?

Efim Geller has pointed out that 41...Kc8! was one way to stay in the game: 42.Be5 Rg1+! (interestingly, the great
player Geller, who contributed significantly to chess culture, misses this possibility in his Chess Informant analysis. He
only mentions 42...c5, but after 43.Rc7+ Kd8 44.Rxa7 Ng5 45.Bc7+ Kc8 46.Bxb6 White’s pawns are dangerous)
43.Kc2 Rg2+ 44.Kd3 Nf2+ 45.Kc3 (after 45.Ke3 Ng4+ 46.Kd3 Nxe5+ Black’s problems are behind him) 45...Ne4+
46.Kb3 Nd2+ and White cannot run away with the king.

42.Kc2! h5 43.a4!

Karpov draws the mating net tighter and tighter.

43...h4 44.Kd3!

The king diverts the knight from the centre.

44...Ng5

After 44...Nf2+ 45.Ke2 Nh3 46.Rf8+ Kb7 47.Rb8+ Ka6 48.Bd2 White wins.

45.Rf8+!

White catches the king.

45...Kb7 46.Rb8+ Ka6

47.Bd2! Rg3+ 48.Kc2 1-0

Black cannot avoid the checkmate.

Chapter 3
World Champion

Fischer forfeited his title when he refused to play the match against Karpov under the prevailing FIDE regulations. In
March 1975, the FIDE congress had agreed to Fischer’s demand to abolish a set number of games, but did not agree
that the Champion would retain his title if the score was tied at nine wins each.
Decades later Karpov was to declare that he considered the match that never took place to be a gap in his career. His
former second Razuvaev later told me that he had expected that Fischer might have beaten Karpov, but no way by more
than a two-point margin. I think that if the match had taken place, it would have forced Karpov to work very hard and
maybe raised his level just like his matches with Kasparov did with the latter.
Be that as it may, Karpov fully lived up to the expectations of a World Champion by winning the vast majority of his
tournaments, especially in the second half of the 1970s. All in all he competed in 34 tournaments during his reign from
1975 to 1985, winning 29 of them, four of which with a tie. Of his 617 games in this period, he won 227, drew 353 and
lost 37, a percentage of 65,4.
In 1978 he successfully defended his title against Kortchnoi in Baguio City. This match was highly political and close.
In Merano in 1981, Kortchnoi was beaten much more convincingly in a rather one-sided match.

First Reign (1975-1977)

In the USSR Team Championships (Spartakiad) in Riga Karpov scored +4 =3. The first two games are taken from this
event. In Milan, all the world’s top players except Fischer competed. Karpov won the event after a very close play-off
with Portisch (3½-2½). Karpov began his schedule in 1977 with a convincing victory in the German holiday resort of
Bad Lauterberg. His game against Timman, where the Dutch grandmaster tried bravely but Karpov took over and won,
is not included here. Semen Furman came third in what was to be his last tournament. Maybe Karpov’s trainer was
already ill at that time. Furman had been a very strong grandmaster who had defeated seven World Champions in his
career. He had made a major contribution to Karpov’s meteoric rise and his standing as a great champion, during ten
years.
Karpov’s opponent was an Armenian player who died tragically at a young age.

GAME 42
Grigorian, Levon
Karpov, Anatoly
Riga Spartakiad preliminaries 1975 (3)

1.c4 c5 2.Nf3 Nf6 3.Nc3 e6 4.g3 b6 5.Bg2 Bb7 6.0-0 Be7 7.d4 cxd4 8.Qxd4 d6 9.Bg5 a6 10.Bxf6 Bxf6 11.Qd3 Be7
12.Rfd1 Ra7 13.Ne4 0-0 14.Rac1 Ba8 15.Qe3 Nc6 16.Nc3 Qb8 17.Nd4 Ne5 18.b3 Rc8 19.Bxa8 Qxa8 20.Nf3
Nxf3+ 21.exf3 Qc6 22.a4 Rd7 23.Rd3 Qc5 24.Rcd1 Bf8 25.Ne4 Qc6 26.Nc3 Rcd8 27.Qd2 h6 28.h4 Qc5 29.Ne4
Qa3 30.Qc1 Qxc1 31.Rxc1 d5 32.Rcd1 Bb4
33.Nc3

Another idea is 33.f4!?, which prevents ...e5. After 33...f6 34.Kg2 Kf7 35.Nc3 d4 36.Ne2 or 33...f5 34.Nc3 g6 35.cxd5
Bxc3 36.Rxc3 exd5 37.Kf1 White is OK. And in case of 33...Kf8 34.Nc3 Bxc3 35.Rxc3 d4 36.Rcd3 Black cannot push
his central pawns.

33...Bxc3

After 33...d4 34.Ne2 e5 35.f4 f6 36.fxe5 fxe5 37.Kg2 White maintains his pawns in the centre.

34.Rxc3 d4 35.Rc2

Blocking the d-file with 35.Rcd3!? leads to 35...e5 36.a5!? Rd6 (after 36...bxa5 37.Ra1 Rd6 38.Rxa5 f6 39.h5 (39.Kf1!?
Kf7 40.Ke2) 39...Kf7 40.g4 Ke6 41.Kf1 White is somewhat worse but he should be able to hold) 37.axb6 Rxb6
analysis diagram

38.f4 (or 38.Re1 f6 39.f4 exf4 40.gxf4 and White is likely to remove all the queenside pawns, after which it is easier to
hold the position) 38...exf4 39.Rxd4 Rxd4 40.Rxd4 fxg3 41.fxg3 Rxb3 42.Rd8+ Kh7 43.Kf2 Rc3 44.Rd6 and White
manages to simplify to a rook endgame with 3 pawns against 2 on the same side, which is drawish.

35...e5 36.Re2

36.a5 is less effective here than with the rook on d3: 36...bxa5 37.Ra2 Rb8 38.Rd3 f5 39.Re2 Re8 40.Ra2 e4 and Black is
better.

36...Re8 37.Rde1 f6 38.Kg2

White can try to break up the pawn chain right away with 38.f4!?. It is very hard to judge the complications that result.
For example, 38...d3 (after 38...Kf7 39.fxe5 Rxe5 40.Rxe5 fxe5 41.Rxe5 d3 42.Re1 White holds) 39.Re3 Kf7 (if 39...e4
40.f3 (or 40.f5 Re5 41.f3 exf3 42.Kf2 Rxf5 43.Rd1 d2 44.Rxf3 Rxf3+ 45.Kxf3 Rd3+ 46.Kf2 a5 and Black has good
winning chances) 40...f5 41.h5 (41.g4 g6 42.h5 fxg4 43.fxe4 gxh5 and it is not easy to tell how much value Black’s
extra pawn has) 41...Rd4 42.g4 Kf7 43.gxf5 d2 44.Rd1 exf3 45.Kf2 and White holds) 40.f3 d2 (40...Rd4 41.fxe5 fxe5
42.Kf2 and White is safe) 41.Rd1 exf4 42.Rxe8 Kxe8 43.gxf4 Rd3 44.Kf2 a5 45.Ke2 Rxb3 46.Rxd2 Rb4 47.Kd3
analysis diagram

47...Rxa4. This endgame is troublesome for White: 48.Rb2 Rb4 49.Rxb4 axb4 50.Kc2 g6! 51.Kb3 h5 52.Kxb4 g5 and
Black wins.
Back to the game.

38...Kf7 39.f4

White decides to try and break the pawn chain after all.

39...d3 40.Re3
40...e4 41.f3

Otherwise the pressure of Black’s pawns will be too much to bear.

41...f5 42.h5

Not 42.g4? exf3+ 43.Kxf3 Rxe3+ 44.Kxe3 fxg4.

42...g6 43.hxg6+ Kxg6 44.Kf2

Black is a pawn up, but it is not so simple to create a winning position.

44...h5!

This stops g3-g4, but that is not all.

45.Rd1?!

The Armenian player wants to transfer the king to d2, but he simply will not be able to do it. Better practical chances for
White arise after 45.Rh1. Then 45...Rd4 46.Ke1 b5 47.axb5 (47.fxe4 loses to 47...fxe4 48.f5+ Kg5) 47...axb5 48.fxe4
fxe4 49.cxb5 (49.Kd2 bxc4 50.Kc3 Red8 51.bxc4 d2 52.Rd1 Kf5 53.c5 h4 54.c6 R4d6 wins for Black)
analysis diagram

49...Rc8! is the subtle move that catches White’s king (not 49...Ra8 50.Kd2!): 50.Kd1 Rc2 51.Rhe1 Rb4 52.Rxe4 Rxb3
and Black wins.

45...Kf6 46.Ke1 Rg8! 47.Kf2

47...h4!

White is too tied up defending against the central pawns to deal with threats on the kingside.
48.fxe4

The rooks invade decisively after 48.gxh4 Rdg7! 49.Ree1 Rg2+ 50.Ke3 R8g3 51.Rf1 a5 and White cannot release the
grip. Black wins in all lines:
A) 52.b4 axb4 53.Rb1 d2 54.Rbd1 Rh2 55.Rxd2 Rxf3+ 56.Rxf3 Rxd2;
B) 52.Kd4 Rg8! 53.fxe4 Rd8+;
C) 52.Rd2 Rxd2 53.Kxd2 Rg2+ 54.Kc3 d2;
D) 52.h5 Rh2 53.Kd4

analysis diagram

53...Rg8! 54.fxe4 Rd8+ 55.Ke3 Re2+.

48...fxe4 49.Rxe4 Rxg3

Black is threatening 50...Kf5.

50.f5

After 50.Re3 Rxe3 51.Kxe3 h3 Black wins.

50...Kg5 51.b4
51...Rf7!

Now Black is winning easily.

52.Re5 Rxf5+ 53.Rxf5+ Kxf5 54.c5 bxc5 55.bxc5 Ke6 56.Rh1 Rg4 57.Rd1?

There were chances to put up stiffer resistance with 57.Ke3 Rxa4 (or 57...Kd5 58.Kxd3 Kxc5 and Black takes the a-
pawn and wins here as well) 58.Kxd3 Kd5 59.Rc1 Ra3+ 60.Kd2 Kc6 61.Rg1 Kxc5 and according to the computer
program Black wins with the two rook pawns.
However, one should be careful playing with a- and h-pawns when there is only a single set of rooks left on the board.
Some positions in which the rook of the stronger side is passive are drawn.

57...Rd4!

This finishes off quickly. Karpov doesn’t even allow White to reach the two rook pawns’ ending.

58.Ke3 Rd5 59.Kd2 Rxc5 60.Rh1 Rh5 61.Kxd3 h3 0-1

Karpov beat Gurgenidze and Kupreichik and drew quickly with Tal and Petrosian. His game against Spassky was voted
the best game of 1975.

GAME 43
Karpov, Anatoly
Spassky, Boris
Riga Spartakiad preliminaries 1975 (2)

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Bb7 5.Bg2 Be7 6.Nc3 0-0 7.Qc2 d5 8.cxd5 Nxd5 9.0-0 Nd7 10.Nxd5 exd5 11.Rd1
Nf6 12.Ne5 c5 13.dxc5 Bxc5 14.Nd3 Bd6 15.Bf4 Re8 16.e3 Ne4 17.Bxd6 Qxd6 18.Nf4 Rac8 19.Qa4 Qe7 20.Qxa7
Nxf2 21.Nxd5 Bxd5 22.Qxe7 Nxd1 23.Rc1 Rb8 24.Qb4 Bxg2 25.Kxg2 Nxe3+ 26.Kg1 Re6 27.Qf4 Rd8 28.Qd4
Rde8 29.Qd7 Ng4 30.Rc8 Nf6 31.Rxe8+ Rxe8 32.Qb7 Re6 33.Qb8+ Ne8 34.a4 g6 35.b4 Kg7 36.Qb7
36...h5?

Karpov has played superbly in this game and now Spassky is lost. Interestingly, this move reveals that Spassky did not
know that the position with queen and two pawns against rook and three pawns is a draw with the pawn on h6. After
36...h5? White’s king can invade via g5. Therefore, stronger was 36...h6 and Black may well reach a draw if he can
sacrifice his knight for the queenside pawns.

37.h3 Kf6 38.Kg2 Rd6 39.a5 bxa5 40.bxa5 Re6 41.a6 Nc7!?

Maybe here very few players would take the knight, but many players, especially at a lower level, tend to forget about
the fortress motif. Kasparov was very good at this and spotted it in his My Great Predecessors book. In his first World
Championship game with Short he found it within a minute.
42.a7!

Karpov finds the fastest way to win. Taking the knight would prolong the game. After 42.Qxc7 Rxa6 43.Qc3+ Ke6
44.Qg7 White does win, and it would be a win even if Black’s king was on h7 as White can invade via g5.

42...Re7 43.Qc6+ Ke5

44.Kf3! 1-0

Here Karpov reveals the major problem associated with fortresses – zugzwang. It is possible to build a fortress but one
must have a spare move to be able to play. Fortresses and zugzwangs are similar to exchange sacrifices, they are often
not that difficult if the idea occurs to you, but they do not spring to mind easily!

GAME 44
Ljubojevic, Ljubomir
Karpov, Anatoly
Milan 1975 (3)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.d4 exd4 7.e5 Ne4 8.Nxd4 0-0 9.Nf5 d5 10.exd6 Bxf5 11.dxe7
Nxe7 12.Bb3 Qxd1 13.Rxd1 Rad8 14.Re1 Nc5 15.Nc3 Rd7 16.Be3 Nxb3 17.cxb3 Rfd8 18.Rad1

18...f6

Karpov prepares the centralization of his king.

19.Rxd7 Rxd7 20.Rd1?!

White chooses the wrong path to achieve the draw he wants. Maybe Ljubojevic felt uneasy because Karpov was
prepared to carry on playing a drawish endgame. One should never get upset just because the opponent insists on
playing on, as long as he does so within the rules. Needless to say, full concentration is required even if the position
appears to suggest a tedious draw. The Yugoslav number one player should have followed Karpov’s example and
brought his king to the centre with 20.f3.

20...Rxd1+ 21.Nxd1 Nd5 22.Bd2?!

The Serbian grandmaster gives up a pawn to reach an opposite-coloured bishop ending. His position will not be lost, but
it’s not easy to defend either.
A worthwhile alternative was 22.Nc3 Nb4 (22...Nxe3 23.fxe3), when Matanovic recommends forcing the a-pawn to the
4th rank: 23.Kf1 (23.a3 Nd3 24.Nd5 Be6 25.Nxc7 Bxb3 draws: 26.Bd4 Kf7) 23...Kf7 (with 23...Nd3 24.Nd5 Kf7?
Black can even overpress: 25.Nxc7! Nxb2 26.Ke2 and White wins a piece) 24.Ke2 Ke6 and White should hold
(24...Nd3? 25.g4 Bg6? 26.f4) 25.a3 Nc6 (25...Nd3? loses to 26.g4 Bg6 27.f4) 26.b4! Ne5 and Black is still a little
better.

22...Bb1 23.a3 Bc2 24.Ne3 Nxe3 25.Bxe3

25...Bxb3

The pawn is in the pocket, yet one plus pawn is often not enough to win an opposite-coloured bishop ending.

26.f3 Kf7 27.Bf4 c6 28.Bd6 Ke6 29.Bf8

Marin, who in his excellent book Learn from the Legends devotes a chapter to Karpov’s opposite-coloured bishop
ending, writes that it was more natural to defend b2 with the king after 29.Bc7!? Kd5 30.Kf2 Kc4 31.Ke2 Ba4 32.Kd2
Kb3 33.Kc1. This is a draw according to the Romanian grandmaster.

29...g6 30.Kf2 a5 31.Ke3

If 31.Bc5 Kd5 32.Bb6 a4, White probably holds after 33.Ke3.

31...b6!!

Karpov sets Ljubojevic problems just as he did Spassky. He places the pawns on the colour of the opponent’s bishop.
Unfortunately my pupils sometimes like to do the same thing. But they do not feel as comfortable as Karpov when they
have to assess whether the pawns are targets or whether they restrict the opponent’s bishop.

32.h4 c5 33.g4?!

If 33.Bh6 a4 and the black king can move to the queenside in time, indeed faster than in the game. But after 34.Bf4 Kd5
35.Bc7 b5 the black king is unable to shoulder away the white king. Matanovic recommended harassing the queenside
pawns with 36.Bd8. Then 36...f5 37.Be7 (37.Kd2; 37.g4) 37...Bc4 38.g4 and White should hold.

33...Bd1!

Karpov temporarily forces Ljubojevic’s king to keep defending the f3-pawn.

34.Ke4?

Losing a tempo. Better would have been 34.h5 gxh5 35.gxh5 Kd5 36.Bg7 f5, for instance 37.h6! Kc4 38.f4 Kb3
39.Kd2 Bf3 40.Kc1 and the white king gets to defend the queenside. The position is not clear.

34...a4 35.h5

Marin prefers to defend with 35.g5!. Then, if 35...f5+ 36.Ke3, White’s bishop defends the kingside while the king can
look after the b2-pawn if need be.

35...gxh5 36.gxh5 f5+ 37.Ke3 Kd5 38.h6 Kc4 39.f4 Kb3

40.Bg7?

This is a losing move as White’s king will be too far removed from the real action. Necessary was 40.Kd2!,
approaching with the king and attacking the bishop. Then the piece sacrifice with 40...Kxb2! (if 40...Bf3 41.Kc1!)
offers good practical chances, although White can hold the position with precise defence. He replies 41.Kxd1 and now:
A) 41...Kxa3 42.Kc2 Kb4 43.Kb2 b5! (the piece sacrifice causes White a headache; however, he can save the game)
44.Be7 Kc4 45.Ka3! (it is necessary to stop the pawn as soon as possible. If there were no kingside pawns White could
allow the pawns to advance further and still draw, but with these pawns on he does not have this possibility. If 45.Bd6
b4 and Black takes over) 45...Kd5 46.Bf8 and White is safe;
B) 41...c4!
analysis diagram

B1) Not 42.Be7? as he will be denied the opportunity to sacrifice the bishop to stop the pawns. If 42...b5 43.Bf8 (it is
too late to occupy the diagonal now with 43.Bf6+ because of 43...c3 44.Ke2 Kc2! (after 44...b4 45.axb4 (45.Kd3 Kb3-
+) 45...Kc2 46.b5 a3 47.b6 a2 48.Bxc3! Kxc3 49.b7 a1=Q 50.b8=Q the queen ending is a draw) 45.Be7 (45.Ke3 b4
46.Kd4 b3 and Black wins) 45...Kb1 wins. A triangulation with the king for this purpose is quite rare) 43...c3 44.Bg7

analysis diagram

If Black plays 44...Kxa3 45.Kc2 Kb4 (not 45...b4? 46.Bf8 and White even wins) 46.Bxc3+ Kc4 White is not in danger
of losing. Black can also hold the position as all he has to do is capture f4 and take his king back to f7.
But with 44...b4! Black can create a winning passed pawn. Matanovic misses this win in his Chess Informant analysis.
45.Ke2 Kc2 Black wins with another ‘barge’;
B2) 42.Bg7+!. White must play this at once. He cannot waste time improving the positions of his bishop and king.
Even so he achieves no more than a drawn position: 42...c3 (in case of 42...Kxa3 43.Kc2 b5 44.Bf8+ Ka2 45.Bd6,
Black can make no progress)
analysis diagram

43.Ke2! (White’s king must waste no time) 43...Kc2 (43...Kxa3 44.Bxc3 Kb3 (44...b5 45.Be5 b4 46.Bd6=) 45.Kd3 b5
46.Be5 a3 47.Bf6 b4 48.Be7=) 44.Ke3 b5 45.Kd4 and now:
B21) 45...Kb3 46.Kd3 Kxa3

analysis diagram

47.Kc2!!. The only move to save White. (47.Kxc3? b4+ 48.Kc2 Ka2, or 47.Bxc3 b4 48.Bf6 b3, and Black wins)
47...Kb4 48.Bxc3+ Kc4 (the position is equal) 49.Bd2 a3 (49...b4 50.Kb2 Kc5 51.Kc2 Kc4) 50.Be1 b4 51.Bh4 b3+
52.Kd2 Kd5 53.Kc3 Ke4 and again, Black soon takes the f4-pawn and returns to f7;
B22) 45...Kd2 46.Kc5 c2 47.Bb2 c1=Q+ 48.Bxc1+ Kxc1 49.Kxb5 Kd2 50.Kxa4 Ke3 51.Kb5 Kxf4 52.a4 Kg3 53.a5
f4 54.a6 f3 55.a7 f2 56.a8=Q f1=Q+. The queen ending is drawish as Black cannot win even if he takes the h-pawn. Of
course, Karpov would have played on for a long time.

40...Kc2!!

The Black king ‘shoulders away’ his counterpart and his domination suffices for the full point. The move itself is not so
spectacular that it deserves the double exclamation mark. It was the beauty and the effect that justified this award.

41.Be5 Bh5 42.Bf6

White decides to wait and see how his opponent will deliver the final blow. He could try to do something with 42.Bc7!?
but the result is the same – White is lost:
A) Not 42...b5? when there is no tactical win at Black’s disposal; there is one difference in comparison to the game:
43.Bb6 c4 44.Bd4 b4! 45.axb4 Be8 46.Be5 (46.Kf2? c3 47.bxc3 Kd3 48.Be3 (48.Be5 Kc4) 48...a3 49.Bc1 a2 50.Bb2
Kc2 and Black wins) 46...Bb5 47.Bf6 and now:
A1) Improving the bishop with 47...Ba6 doesn’t help: 48.Be5 c3 49.Bxc3 a3 and giving up two pawns doesn’t bring
Black a win:
analysis diagram

50.Kd4! a2 (in the event of 50...axb2 51.Bxb2 Kxb2 52.Ke5 Kc3 53.Kxf5 White easily reaches the h1-corner) 51.b3
Kxb3 52.Ba1 Kc2 53.Ke5 and White is safe.
A2) 47...c3 48.Bxc3! (48.bxc3 Bc4 49.b5 a3 50.b6 a2 51.b7 a1=Q 52.b8=Q Qg1+ and mate) 48...a3 49.Kd4! a2 50.b3
Be2 51.Ba1 and White won’t lose.
B) 42...Kxb2! (this wins prosaically) 43.Bxb6 c4:
B1) 44.Bd4+ (White loses if he is two pawns down) 44...Kxa3 45.Kd2 Kb3 46.Be5 a3 47.Bf6 a2 48.Be5 Bf3 49.Bf6
Be4 50.Be5 Kb4 51.Bf6 Kc5 52.Ke3 Kd6 53.Be5+ Ke6 54.Kd2 Kf7 55.Ke3 Kg6 56.Bg7 c3 57.Bxc3 Kxh6 58.Kf2
Kg6 59.Kg3 h5 60.Kh4 Bf3 61.Kg3 Bg4 and the black king returns to the queenside;
B2) 44.Bc5 c3! (I believe 44...Bf7!? is another way to win: 45.Ke2 Bd5 46.Ke3 Be4 47.Ke2 c3 48.Bd4 Kb3 49.Kd1
(in case of 49.Bc5 Black can win by tactical means: 49...c2) 49...c2+ 50.Kc1 Kxa3 and the black king goes to the
kingside, wins the h-pawn by diverting the bishop and returns to the queenside to win the bishop. But Marin’s method
is faster) 45.Bd4 Kb3 46.Kd3 c2 47.Kd2 Bd1! (if 47...Bf3? 48.Kc1 White maintains the one pawn deficit, which is
quite an achievement for him) 48.Kc1 Kxa3 and White is obliged to give up a kingside pawn for the a-pawn, so he
loses.

42...Bf7 43.Be5
43...Bb3!!

It is not immediately obvious why this is such an important move.

44.Bg7 b5 45.Bf8 c4 46.Bg7

After 46.Bb4 Kxb2 47.Kd2 c3+ 48.Bxc3+ Kxa3 wins, as Marin has pointed out.

46...b4!!

Such moves can be missed when one calculates opposite-coloured bishop endings over the board.
47.Kd4

Ljubo denies Karpov the opportunity to win nicely. This move signifies resignation. The game would have reached a
more attractive conclusion if he had played 47.axb4 c3!! (two exclamation marks for beauty) 48.Bxc3 (48.bxc3 Bc4
49.b5 a3 50.b6 a2 51.b7 a1=Q 52.b8=Q Qg1+ 53.Kf3 Bd5+ and mates in 3)

analysis diagram

48...a3!. The point of 43...Bb3 – White cannot play 49.Kd4 and 50.b3 as the bishop blocks that square. If 49.bxa3 Kxc3
50.b5 Bd5 51.a4 Be4 (capturing the queenside pawns immediately would allow unnecessary counterchances, as
White’s king would be free on the kingside) 52.b6 (52.a5 Bd3 53.b6 Ba6) 52...Kc4 53.a5 Bb7 54.Kf2 Kd4 and Black
wins the f4-pawn and the game.
47...c3 48.bxc3 bxa3 49.c4 a2 50.Kc5 Kb1 51.Kb4 a1=Q 52.Bxa1 Kxa1 53.c5 Kb2 54.c6 a3 55.c7 Be6 56.Kc5 a2
57.Kd6 Bc8 0-1

Ljubojevic played carelessly, but Karpov’s relentless endgame play contributed to his punishment.
Against Uhlmann in the last round, Karpov refused a draw offer. Even though runner-up Uhlmann was a point behind
and a draw sufficied to win the tournament, Karpov went for the win.

GAME 45
Uhlmann, Wolfgang
Karpov, Anatoly
Skopje 1976 (15)

1.c4 Nf6 2.Nc3 e6 3.Nf3 b6 4.e4 Bb7 5.Qe2 Bb4 6.e5 Ng8 7.d4 Ne7 8.Qd3 d5 9.exd6 cxd6 10.a3 Bxc3+ 11.Qxc3
Nd7 12.Be2 Rc8 13.0-0 0-0 14.Bg5 h6 15.Bh4 Qe8 16.Bg3 Nf5 17.Bf4 b5 18.Bd3 Qe7 19.Bxf5 exf5 20.Rfe1 Qf6
21.d5 Rxc4 22.Qxf6 Nxf6 23.Bxd6 Rfc8 24.Be5 Nxd5 25.Nd4

25...f6! 26.Nxf5

After 26.Nxb5 fxe5 27.Nd6 R8c7 28.Nxc4 Rxc4 29.Rxe5 Nf4 Black is better.

26...fxe5 27.Nd6
27...Nf4!

Three Black pieces are under attack, yet Karpov moves the fourth. This is the start of some remarkable play with the
knight.
An interesting alternative is 27...R8c7!? 28.Nxc4 bxc4 29.Rxe5 Nf4 30.f3 Rd7, which Uhlmann evaluates as equal and
Karpov says is insufficient. However, it looks as if this gives Black a very clear advantage.

28.Nxb7

After 28.Nxc8 Rxc8 29.Rxe5 Nxg2 (or 29...Bxg2 30.Rf5 Ng6 – this little tactical move keeps the advantage) 30.Rxb5
Bf3 Black wins.

28...Nd3! 29.Nd6 Nxe1 30.Nxc4 Nc2 31.Nd6 Rd8!

The only move to keep the advantage. 31...Rc6? allows 32.Nf5.

32.Nb7 Rd2!

Classical endgame play; the rook goes to the second rank. It will cause serious damage there.

33.Rc1

After 33.Rb1 e4 34.Nc5 e3 35.Ne4 exf2+ 36.Nxf2 Ne3 Black has a clear advantage. Actually White loses a pawn, and
probably the game.

33...Nd4

Karpov makes a fifth move with the knight. The reward for this attractive sequence is a pawn.

34.Rc8+ Kh7 35.h4 Rxb2 36.Ra8

36...Ne2+

Karpov goes after White’s king. Also winning was 36...Ra2 37.Rxa7 b4. However, Black is not able to win more
material immediately in that case.

37.Kh2 Nf4 38.Kg1 Nd3 39.Nd6 Nxf2 40.Rxa7 Ng4

The knight is doing a great job in this game.

41.g3

Stopping the checkmate but losing another pawn.

After 41.Rf7 Karpov shows one of the many wins in the position; probably he had planned to follow up with 41...Ne3
42.Nf5 Rxg2+ 43.Kh1 Rf2 44.Nxh6 Rf6 45.Rxf6 gxf6 46.Nf7 Nc4 and Black wins.

41...Ne3 42.Kh1 Kg6 43.Rb7 Rd2 44.Nxb5

After 44.Rb6 Kh5 45.Nxb5 Kg4 Black is about to checkmate.

44...Nf5 45.g4 Nxh4 46.Kg1

White stops the well-known mate threat with ...Nf3.

46...Rg2+ 47.Kf1 Rxg4 48.Ra7

48...Nf5

The knight has made a glorious series of leaps in this game and now it clears the way for the h-pawn.

49.a4 h5

Karpov has last moved a pawn on his 25th move. His remarkable handling of the knight in this game (14 moves from
the first diagram – 17 moves in the entire game) shows he is not only extremely strong at endgames, but also versatile.
Amusingly, all his remaining moves in this game are with pawns...

50.a5 h4 51.Ra8 h3 0-1

 GAME 46
Pfleger, Helmut
Karpov, Anatoly
Montilla 1976 (6)

1.c4 Nf6 2.Nc3 e6 3.Nf3 c5 4.g3 b6 5.Bg2 Bb7 6.0-0 a6 7.b3 d6 8.Bb2 Be7 9.d4 cxd4 10.Nxd4 Bxg2 11.Kxg2 0-0
12.Qd3 Qc7 13.Qf3 Ra7 14.Rfd1 Rc8 15.Rac1 Qb7 16.Qxb7 Rxb7 17.f3 Nc6 18.Ba3 Na7 19.Nc2 d5 20.Bxe7 Rxe7
21.Ne3 Rec7 22.Na4 Rc6 23.Rb1 dxc4 24.bxc4 b5 25.cxb5

25...Nxb5!

It is hard to imagine that Black could possibly win from such a seemingly equal position. The queenside pawn is much
less prone to exchange on a6 than on b5. Black desperately has to keep it to have winning chances. In the event of
25...axb5 26.Nb2 Ra6 27.Ra1 Kf8 (after 27...b4 28.Nd3 (28.Rd3 Nc6 29.a4) 28...Nc6 29.Rdb1 Ra4 30.Rb2 White
exerts pressure on b4) 28.Nd3 (28.a4 b4 29.Nd3) 28...Nc6 29.Kf2 White holds.

26.Kf2

This is one way to play the position. It was also possible to handle it with 26.f4! Ne4 (after 26...h5 27.Kf3 Kf8 28.h3
(or 28.Rd3 Ke7 29.Rbd1 and the rooks have a strong grip on the d-file) 28...Ke7 29.g4 hxg4+ 30.hxg4 Rh8 31.Rb3 Nd6
32.Rd4 Rh3+ 33.Kg2 White is not worse) 27.Rb4 (27.Rd7 Kf8 28.Kf3 Ke8; 27.Rd3 f5 28.Rbd1 h5 29.h3 and White is
in the game) 27...f5 28.g4 g6 29.Kf3 and White has not killed all the play in the position, but he is not worse.

26...g5!

Karpov makes a luft and gains space.

27.Rb4 Kg7!

Karpov likes to improve his king. This is only a small addition to the position, but every addition counts.
28.h4 h6 29.Rh1?

A waste of a tempo. Better was 29.hxg5 hxg5 30.Rd2 Rc1 (in case of 30...R8c7 31.Nb2 Nc3 32.Nd3 Black has no edge)
31.g4! Rh8 32.Rc4! Rch1 33.Rc6. Now White has a target as well, and Black seems to have no more than a perpetual on
the kingside.

29...R8c7!

Here is another small improvement.

30.hxg5 hxg5 31.Ra1

31.g4 Rd6 32.Rc4 is equal.

31...Rd6

Karpov looks to the d-file and the second rank. An interesting move was 31...Nd5!. By simplification he could already
have tried to invade: 32.Nxd5 exd5 33.Rd1 (33.Nb2? Rc2; or 33.f4 g4 34.Rh1 Rc2 and White is worse) 33...d4!
(White’s position is fixed and not easy to defend) 34.Rd2 (in case of 34.Nb2 Rc2, Black’s rook is menacing) 34...Rh6
35.g4 Rc1 36.Nb2 Rch1 Black has a very dangerous attack. For example, the knight has good chances to help the rooks
against the king: 37.Nd1 (or 37.Nd3 Nc3! 38.Rxd4 Nd1+ and White must give up the exchange) 37...f5! 38.gxf5 Nd6
and White loses an exchange.

32.Nb2

White should have eased the pressure on his position with 32.Rc4!. This would clear the way for the a1-rook:
A) 32...Nd5 33.Rac1;
B) 32...Rd2 33.Rxc7 Nxc7 34.Ke1 Rd4 35.Nc5 – White soon activates his rook and holds;
C) 32...Rxc4 33.Nxc4 Rc6 34.Nab2 g4 35.a4 Nc3 36.a5 – White is safe as the a-pawn and the knights are protected and
his rook will arrive on the scene soon.
32...Nd5 33.Nxd5?!

This keeps the rook on a1 passive.

Again, with 33.Rc4! there was the possibility of improving the a1-rook fairly quickly: 33...Rdc6 (33...Nbc3 34.Rc1!)
34.Rac1 Rxc4 35.Rxc4 Nbc3 36.a3 f5 37.Rxc7+ Nxc7 38.Nd3 Kf6 39.Nc4 and White has solved his problems.

33...Rxd5

34.Rc4

White still had a chance to improve the a1-rook, this time with 34.a4!. Then 34...Nc3 (after 34...Nd4 35.Rd1 e5 36.Rc4
White is freed) 35.Nd3 a5 (35...e5 36.Rc1) 36.Rb8 and now the other rook is freed: 36...Rc4 (other moves fail to keep
the tension: 36...Rd4 37.Ra8 or 36...Kf6 37.Ke3 Rc4 38.Rc1=) 37.Rc1 Rdd4 38.Ra8 and White is holding easily.

34...Rxc4 35.Nxc4 Nc3! 36.Ke1 Rd4 37.Ne3

White would have done better to cover the a4-square with 37.Nb2!. Then, after 37...f5 (or 37...Rb4 38.Nd3 (with
38.Nd1 Nb5 39.Rc1 f5 40.Nc3 Nd4 41.Rb1 White holds) 38...Ra4 39.Kd2! and White wittily keeps the balance:
39...Rxa2+ 40.Rxa2 Nxa2 41.Kc2 a5 42.Kb3 Nb4 43.Nxb4 axb4 44.Kxb4 and his king arrives back in time) 38.Nd3
(after 38.Nd1 Nd5 39.Rb1 Ra4 40.Rb2 g4 Black has similar pressure as in the game) 38...Kf6 (38...Ra4 39.Nc5 and
White takes a pawn), improving the rook with 39.Rc1! forces a draw: 39...Nxa2 40.Ra1 (or 40.Rc6 a5 41.Nc5 and
White is safe) 40...Ra4 41.Nb2 Ra3 42.Nc4 and White holds the position.

37...Ra4! 38.a3 Nb5

Tying up White’s pieces.

39.Nc2
39...g4 40.fxg4?

This reduces the number of pawns but isolates two of them.

A) With 40.Kf2 White could prevent Black from using the ensuing motif. If the knight takes on a3, he cannot follow up
by taking on c2 with a check. Therefore the rook is allowed to move: 40...Kf6 (or 40...e5 41.Rb1 gxf3 42.exf3 Rc4
43.Ne3 and White is just a little worse) 41.Rd1 e5 42.Rd3 Ke6 and Black is slightly more active;
B) The king can also come to the rescue of the rook with 40.Kd2! e5 41.Kc1! (waiting passively with 41.Ke3 brings no
joy: 41...f5 42.Kf2 Kf6 43.Rb1 Ke6 and Black has some chances) 41...Rc4 (41...f5 42.Kb2 Nd4 (not 42...Kf6?? as
43.Kb3 traps the rook) 43.Nxd4 Rxd4 44.Kc3 and White holds) 42.Kb2 gxf3 43.exf3 Nd4 44.Nxd4 exd4 45.Rc1
(45.g4 Kf6 46.Re1 Rc3 47.Re4=) 45...Rxc1 46.Kxc1 Kf6 47.Kd2 Ke5 48.Kd3 Kd5 49.g4 Ke5 50.g5 and White draws.

40...Rxg4 41.Kf2 f5 42.Kf3?!

The king will be unfortunately placed on the third rank, as the g3-pawn cannot be protected by the white rook and
checks on the third rank may cause problems.
Freeing the rook with 42.Rd1 was still an option: 42...Rc4 (42...Ra4 43.Rd3 Rc4 44.Ne3 is not that bad for White; or
42...Nc3 43.Rd3 Ne4+ 44.Kg2 Kf6 and White is not exactly having fun, but he is not completely out of it either.
Indeed, he has real chances to hold) 43.Ne3 Ra4 44.Nc2 Kf6 45.Rd3 and it is hard to tell whether or not Black can win
this.

42...Kf6 43.Rd1 Rc4 44.Ne3 Ra4 45.Nc2 e5!

Even with limited material White is pushed back.

46.Kf2
46...Ke6!

Karpov keeps improving his pieces.

47.Rd3 Rc4 48.Ne3 Rc3!

As the black king is superior, exchanging the rooks wins for him.

49.g4?

After some inaccuracies the German grandmaster commits his final clear, possibly decisive, mistake. The pawn
becomes weak on g4 and Black’s central pawns cannot be attacked by the white king.
A better option was 49.Rxc3 Nxc3 50.Kf3 a5 51.Nc4 (in case of 51.g4 e4+ 52.Kg3 (or 52.Kf2 f4 53.Nc4 a4 54.Nb2
Kf6 55.e3 f3 56.Kg3 Kg5 57.Nc4 and White is still alive) 52...Nxe2+ 53.Kf2 Nd4 54.gxf5+ Nxf5 55.Nc4 White
holds) 51...a4 52.Ke3 and White still seems to be able to hold.

49...f4 50.Rxc3 Nxc3 51.Nc2

51...e4!

Cutting the enemy king out of the game.

52.a4

After 52.Nb4 a5 53.Nc6 (or 53.Nc2 Kf6 54.Ke1 e3 55.Kf1 Kg5 and Black invades) 53...a4 54.Na5 Kd5 Black wins.
He can also go after the g5-pawn.

52...Nxa4

Finally the weak a-pawn falls.

53.e3 f3 54.Nd4+ Ke5 55.Nc6+ 0-1

In the Soviet Championship Karpov suffered a loss against Geller in the third round, but nevertheless managed to take
the lead and eventually win the tournament. The fact that he saved the following endgame was important.

GAME 47
Petrosian, Tigran
Karpov, Anatoly
Moscow USSR Championship 1976 (5)

1.Nf3 Nf6 2.c4 b6 3.g3 Bb7 4.Bg2 e6 5.d4 Be7 6.0-0 0-0 7.Nc3 Ne4 8.Qc2 Nxc3 9.Qxc3 c5 10.Rd1 d6 11.b3 Nd7
12.Bb2 Nf6 13.d5 e5 14.Nxe5 dxe5 15.d6 Bxg2 16.Qxe5 Re8 17.dxe7 Qxe7 18.Qxe7 Rxe7 19.Kxg2 Rxe2 20.Bxf6
gxf6 21.Rd7 Rb2 22.Re1 Rxa2 23.Ree7 Rf8 24.Rxa7 Rxa7 25.Rxa7 Rd8 26.Rb7 Rd6 27.g4 Kg7 28.f3 Kg6 29.Kg3
29...h5!

Black gives up a pawn but, in return, devours all of White’s pawns on the queenside. It was already long known that an
ending of rook and h- and f-pawns versus rook is drawn if the weaker side’s king is not cut off and the stronger side’s
pawns are not too close to promotion.

30.gxh5+ Kxh5 31.Rxf7

Petrosian chooses the best practical chance. After 31.h4 Kg6 32.Kf4 Rd4+ 33.Ke3 Rd6 White can’t make progress.

31...Rd3!

Black must get rid of the queenside pawns.

32.Rxf6 Rxb3 33.Kf4 Rb4

Karpov decides to retain the c-pawn. It was possible to simplify into the h- and f-pawn rook ending with 33...b5.

34.Kf5 Rxc4 35.Rxb6

The position would be a draw without Black’s c-pawn. The black king is cut off on the sixth rank but it can’t be kept
there. Computer programs give a definitive evaluation of this kind of position.
35...Rc1

Karpov insists on keeping his last pawn, thus experiencing the difficulties of the ensuing endgame. It is recommended,
in an over-the-board game, to free the rook quickly and defend from the a-file early. Two options were 35...Rc2!? and
35...Ra4!?.

36.Rc6 c4

At this moment Black could again have liberated his rook with 36...Rd1. If White takes the pawn, 37.Rxc5 Kh6 draws.
One of Karpov’s best qualities, if not his best, in the endgame, is that he picks up a plan and sticks to it very
consistently. If one overdoes this, one’s best quality may temporarily become a bad quality. Here Karpov nearly pays
for persisting too long on the chosen path of keeping the c-pawn.

37.Rc8 Kh6 38.Kf6 Kh7 39.f4 Rc2

After 39...Ra1 40.Rxc4 Ra6+! (other black moves lose) 41.Kg5 Rg6+! (only move) 42.Kf5 Ra6 Black holds.

40.f5

After 40.h4 Black can defend as in the game, but he can also free his rook: 40...Ra2 41.Rxc4 (on 41.Rc7+ Kh6 42.Rxc4
Ra6+! draws) 41...Ra6+! 42.Kg5
analysis diagram

Now, with 42...Rg6+! Black draws. His last two moves were ‘only’ moves.

40...c3

Karpov could have drawn with 40...Rxh2! .

41.h3 Rc1!

This was a not the right moment to give up the c-pawn by playing 41...Ra2? since after 42.Rxc3 Ra6+ 43.Kg5 Ra7 44.h4
Rb7 45.h5 Ra7 46.f6 Ra5+ 47.Kg4 Ra4+ 48.Kg3 Ra7 49.Rc4 White wins.
42.h4

After 42.Kf7 Rf1 43.f6 Ra1 Black can successfully defend with side checks. 44.Rxc3 Ra7+! (the only move to draw)
45.Ke6 Ra6+ 46.Ke7 Ra7+ 47.Kd6 Ra6+ 48.Rc6 Rxc6+. This is the trivial and, interestingly, also the only way to hold
the position. After 49.Kxc6 Kg6 Black draws as Matanovic has pointed out.

42...Rc2

After 42...c2, 43.h5 wins according to Matanovic’s Informant analysis. However, computer analysis proves he was
incorrect: 43...Ra1 44.Rxc2 (44.Rc7+ Kh6) 44...Kh6! with a draw, while 44...Ra7? 45.Rc8 Rb7 46.Re8 loses.

43.h5 Rc1

Karpov sealed this move. He and his seconds must have analysed the endgame very deeply, maybe even to perfection.
However, Karpov has kept the knowledge of that consultation to himself – he never published the analysis. The
precision with which he continues suggests that his team did a good job.
With 43...Ra2 44.Rxc3 Kh6! Black would also draw.

44.Kf7

44...Rc2!

Not 44...Ra1? when 45.Rxc3 wins: 45...Ra7+ (45...Kh6? 46.Re3) 46.Ke6 (after 46.Kf6?? Kh6 Black survives)
46...Ra6+ 47.Kd7! Kg7 48.Rc6!.

45.f6 Rc1 46.Ke7 c2

This makes the rook passive; but Black still has a king move. Again not 46...Ra1? as 47.Rxc3 wins for White: 47...Ra7+
48.Kd6 Ra6+ 49.Rc6.

47.Kf7 Kh6 48.Rc5 Kh7 49.Rc6 Kh6 50.Kf8 Kh7!

As before, 50...Ra1? loses to 51.Rxc2.

51.Rc7+

In the event of 51.f7 Ra1 52.Rxc2 Ra8+ 53.Ke7 Ra7+ Black holds as his king can go to g7. White can interpose his rook
and win if the black king can be held off from the f-pawn after that. If not, Black draws.

51...Kh8!!

Karpov defends with the required precision. He has built a stalemate motif into his strategy. Even if Razuvaev did not
mention it, one suspects that Anatoly and his team had probably gotten this far in their home analysis.
With 51...Kh6? Black could still resist, but he cannot save the game after some remarkable moves by White: 52.f7 Kh7
53.h6 (after 53.Rc6 Ra1 54.Rxc2 Ra8+ Black holds) and now, remarkably, with the pawn on h6, White wins – with the
pawn on h5 it is a draw:
A) 53...Ra1 54.Rxc2 and White wins;
B) 53...Kxh6 54.Kg8 Rg1+ 55.Kh8 – both Dvoretsky and Matanovic show that here the famous motif of the Lasker
ladder wins again: 55...Rf1 56.Rc6+ Kh5 57.Kg7 Rg1+ 58.Kh7 Rf1 59.Rc5+ Kh4 60.Kg7 Rg1+ 61.Kh6 Rf1 62.Rc4+
Kh3 63.Kg6 Rg1+ 64.Kh5 Rf1 65.Rc3+ Kg2 66.Rxc2+ (finally, with check);
C) 53...Kh8(!). This move is not mentioned in any available analysis. It loses, but only to a very subtle idea. In
adjournment analysis at this level such a line would offer little hope, but over the board it would set up a real obstacle:
C1) 54.h7 Kxh7 55.Rc6 Ra1 and Black draws;
C2) 54.Ke7 Re1+ 55.Kd7 (55.Kf6 Rf1+) 55...Rf1 56.Ke8 Re1+ – White can make no progress this way;
C3) 54.Rc8! Kh7 (54...Ra1 55.Rxc2)
55.Rc6!! (this thrusts Black into a zugzwang. After 55.Ke7 Re1+ 56.Kd6 Rf1 Black would be safe) 55...Kh8 (this
allows White to promote with check, but 55...Ra1 56.Rxc2 loses as well) 56.Ke7 Re1+ 57.Re6! Rf1 58.Rf6 Re1+
59.Kd7 Rd1+ 60.Ke6 Re1+ 61.Kd5 Rd1+ 62.Ke4 Re1+ 63.Kd3 and White wins.

52.f7 Ra1! 53.Rxc2

There is a stunning possibility after 53.Ke7 Re1+ 54.Kf6 Rf1+ 55.Kg6 Rg1+ 56.Kh6:

analysis diagram

56...c1=Q+!! 57.Rxc1 Rg6+!! and the stalemate saves Black.

53...Ra8+! 54.Ke7 Ra7+ 55.Kf6 Ra6+ 56.Kg5 Ra5+

Here 56...Kg7 was a draw, too.

57.Kg4 Ra4+ 58.Kg3 Ra3+ 59.Kg2 Kg7! 60.Rf2 Kf8!

60...Ra8?? 61.h6+ would be disastrous.

61.Rf5

61...Ra6!

Black has to take a longer route via the h-file to collect the f7-pawn. Not 61...Ra7? 62.h6.

62.Kg3 Rh6

The most clinically precise way to draw.

63.Kg4 Rh7 ½-½

Black wins the f7 pawn, so there is no point playing on.

Razuvaev told me that the Karpov team had worked very hard on the adjourned position. At that time it had long been
known that rook endings with f- and h-pawn are drawn. However, the closer the pawns are to promotion the better the
chances for the stronger side. According to the Russian grandmaster, what made the adjourned position difficult was
that in the examples they knew, when a passive rook defended the bishop pawn as in the game, the stronger side won.
Then they found the saving stalemate idea.
Even modern analytical tools cannot find fault with Karpov’s play. During the game he never had a lost position, even
though he was stepping through a minefield. This display shows the superb level of Karpov’s analysis team.
Tigran Gyozalyan told me that he was studying in Moscow at that time and that he has warm memories of Petrosian.
The Armenian community in Moscow sensed that their hero would score a win against the World Champion and quite
a number of Armenians went to see the adjournment hoping for a celebration afterwards. Petrosian himself was
optimistic about the outcome. It was said that even a news report on Soviet television was predicting a Petrosian win.
Many have compared Karpov’s style to that of the Armenian ex-World Champion. And not without reason: both were
equally profound in their understanding of the game and both of them were certainly not flashy in style. One must also
venture deeply to see the beauty of their play.
Interestingly, Karpov later found himself on the stronger side of a c- and a-pawn ending against Bareev. After 14 further
moves the following position arose.

Karpov - Bareev
Tilburg (1), 1991

82...Rg6!

Evgeny draws with an unusual motif, one which can occur with a g-pawn as well.

83.Rc8 Rh6 ½-½

If the pawn moves Black keeps giving checks on the sixth rank.
This game was also played with an adjournment.
The following ending is remarkable even by Karpov’s standards, yet there seems to be no published analysis of it.

GAME 48
Zakharov, Alexander
Karpov, Anatoly
Moscow USSR Championship 1976 (11)

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bxc3+ 5.bxc3 c5 6.f3 d6 7.e4 Nc6 8.Ne2 b6 9.Ng3 0-0 10.d5 Na5 11.Bd3 Ba6
12.Qe2 Nd7 13.f4 exd5 14.cxd5 Bxd3 15.Qxd3 c4 16.Qf3 Nb3 17.Rb1 Re8 18.0-0 Ndc5 19.Be3 Nxe4 20.Nxe4 Qe7
21.Nd2 Qxe3+ 22.Qxe3 Rxe3 23.Nxc4 Rxc3 24.Nxd6
24...Nd2 25.Rbc1 Rd3!?

Karpov calculates correctly that he can subject himself to the pin and even take the pawn. If 25...Rxa3 26.Rfe1 Rd3
27.Nc8 White has compensation for the pawn.

26.Rfd1 Rxd5 27.Ne4 Nb3 28.Rxd5 Nxc1

White has lost a pawn, but his pieces are much better developed. The position is dynamically balanced.

29.Kf2 Nb3

It is safer to bring the knight back – it might get hunted down far from camp.

30.Rd7 f5

With 30...Nc5 Black could have forced a draw: 31.Nxc5 bxc5 32.Rc7 g6 33.Rxc5 Re8 34.Rc7 (34.Ra5 Re7) 34...Re4
35.Rxa7 Rxf4+ 36.Ke3 Rf1 and Black holds this endgame.

31.Nd6 Nc5
32.Re7?!

To all appearances it makes no difference whether White plays this or 32.Rc7. The text move indeed stops 32...Ne6, but
it has a drawback. Remarkably, this slight inaccuracy allows Karpov to create problems, which in turn cause Zakharov
to lose the game.
White should have kept the rook away from the king with 32.Rc7!. Then, 32...Ne6 (after 32...g6 33.Rc8+ Rxc8 34.Nxc8
Black loses his extra pawn) 33.Re7 Nxf4 34.Nxf5 Rf8 (after 34...g6 35.Kf3 Nd3 36.Nh6+ Kh8 37.Ke4 (37.g4 – Black
cannot even think of winning the game with a king position like this) 37...Nc5+ 38.Ke5 White’s activity balances the
pawn deficit) 35.g4 Rf7 36.Re8+ Rf8 37.Re7 is a draw.

32...g6 33.Nf7

It is difficult to judge the alternative 33.Nc4 as it leads to a slightly worse position: 33...b5 (if 33...Ne4+ 34.Kf3 Rc8
35.Ne3 a5 36.Rb7 (with 36.g4 Rc3 37.gxf5 gxf5 38.a4 White can also hold) 36...Rc3 37.Rxb6 Rxa3 38.g4 Nd2+
39.Ke2 Rxe3+ 40.Kxd2 White holds) 34.Ne5 a5 35.Rc7 Ne4+ 36.Ke3 Nf6 37.Rb7 (37.Kd4 Rd8+ 38.Kc5 b4 39.axb4
axb4 40.Rb7 Rd2 is unpleasant for White) 37...Nd5+ 38.Kd4 Nxf4
analysis diagram

39.Nd7! Nh5 40.Rxb5 Rd8 41.Rd5 and White has enough resources. But these tactics were not easy to calculate during
the game.

33...Kf8!

Exploiting the drawback of White’s 32nd move. The king gets one square closer to the centre.

34.Rc7 Ne6 35.Rb7

If 35.Rd7 Re8.
35...Re8!!

Karpov prefers to return the pawn in order to remove the enemy rook from the seventh rank. The game is reminiscent of
the legendary Muhammad Ali-George Foreman heavyweight boxing match. Ali was only shielding his body and face in
the first couple of rounds, then suddenly knocked out Foreman who was tired from delivering so many punches.
Here the rook makes one active move and decides the outcome of the game.
A) With 35...Nxf4 Black could have won another pawn, but he would have remained disorganized. After 36.Ng5 h6
37.Nh7+ Ke8 38.Nf6+ Kd8 39.Rg7 White is very active, therefore he should not lose this position;
B) Black could have freed the rook with 35...Rc8. Then, after 36.Rxa7 Rc2+ 37.Kf3 Rc3+ 38.Kf2 Nxf4 39.Ne5 Rc2+
(39...h5 40.g3) 40.Kf3 Nxg2 41.Rxh7 there is little material left on the board, giving White good drawing chances.

36.Nd6?

White was probably in time-trouble and did not dare to enter a pawn ending which looks so difficult. The following
variation must have upset Zakharov afterwards, as he had indeed missed a very narrow way to escape: 36.Rxa7! Re7
37.Rxe7 Kxe7 38.Ng5

analysis diagram

Out of three tempting moves, 38...Nxg5! looks like the simplest win, but endings can be very tricky. Actually none of
the moves provide Black with a clear win.
Here is the second attempt first. Black could give White a chance to go wrong with 38...h6, but the draw is less hidden
than in the other line: 39.Nxe6! (after 39.Nf3? Nxf4 the extra pawn is enough to win; not 39.Nh3? Nd4 40.Ng1 (or
40.a4 Ne6) 40...Nb5 when Black wins the pawn and the game) 39...Kxe6 40.Ke3 Kd5 41.Kd3 g5 (41...b5 42.h4) and
now:
A) 42.g3? would be a losing mistake. The win is instructive, so we will examine it: 42...g4! 43.a4 h5 44.Kc3 Ke4
45.Kb4 Kf3.
analysis diagram

If 45...h4 46.gxh4 (Black wins the queen ending after 46.Kb5 hxg3 47.hxg3 Kf3 48.Kxb6 Kxg3 49.a5 Kxf4 50.a6 g3
51.a7 g2 52.a8=Q g1=Q+) 46...Kxf4:

analysis diagram

A1) 47.Kc3 (keep in mind that the king can also move backwards and try to stop the enemy pawns. In this case it does
not help, but sometimes it does) 47...Ke5 48.Kd3 f4 49.h5 Kf5 50.h6 Kg6 51.Ke4 f3 52.Ke3 Kxh6 53.h3 Kg5
54.hxg4 Kxg4 and Black wins;
A2) 47.Kb5 Ke5 48.Kxb6 f4 49.a5 g3 50.hxg3 fxg3 51.a6 g2 52.a7 g1=Q+ (promoting with check is a key element
now) 53.Kb7

analysis diagram

53...Qg2+ and Black wins as he can checkmate or win the a-pawn. He can also win by stalemating the white king and
then checkmating him, as White still has a pawn on h4. Please note that if White had no h-pawn and Black’s king stood
on e4, it would be a draw.
Back to the position after 45...Kf3.

analysis diagram

46.Kb5 Kg2 47.Kxb6 Kxh2 48.a5 h4 49.a6 (or 49.gxh4 g3 50.a6 g2 51.a7 g1=Q+ (the promotion with check helps)
52.Kb7 Qb1+ 53.Ka8 Qb6 54.h5 Qc7 55.h6 Qc8 mate) 49...hxg3 50.a7 g2 51.a8=Q g1=Q+ – Black wins a second
pawn and the game;
B) Instead of 42.g3?, the simple 42.fxg5 leads to a draw. After 42...hxg5 43.a4 the position is equal.
Here is the second attempt to get something out of this position (analysis diagram after 38.Ng5).

analysis diagram

By not aiming for a direct win with 38...Nxf4!? Black can keep the upper hand. White should be able to hold the
position. He can exchange the pawns and rescue the knight if Black tries to trap it.
The lines presented are just illustrations. Probably Black can try to win this knight ending in different ways. Karpov
could possibly have pressed as follows: 39.Nxh7 and now:
A) 39...Nd5 40.Ng5 b5 (or 40...Nc3 41.Nf3 Nb1 42.Ne5 g5 43.Nc4 b5 44.Ne3 f4 45.Nc2 and White holds) 41.Nf3
Kd6 42.Nh4 Ne7 43.Ke3 and White is not worse;
B) 39...b5 40.g4 Ne6 41.h4 – again White holds, see the line with 39...Ne6;
C) 39...Ne6!. Suddenly the h7-knight is under pressure. White can rescue it in two ways:
analysis diagram

C1) 40.g4 Kf7!? (with the frightening threat of trapping the knight) 41.g5 (in the event of 41.h4? Kg7 42.Ng5 Nxg5
43.hxg5 Kf7 Black wins) 41...Kg7 42.Nf6 Nxg5 43.Nd5 Ne4+ 44.Ke3 b5 45.Nc7 Nd6 46.Kd4 Kf6 47.Nd5+ Kg5
48.Kc5 Ne4+ 49.Kxb5 f4 50.Kc4 and White holds;
C2) After 40.Kg3 comes 40...Kf7 (if 40...g5 41.h4 gxh4+ (or 41...f4+ 42.Kg4) 42.Kxh4 b5 43.Ng5 Nd4 44.Nf3 Nc2
45.Kg5 Ke6 46.Kf4 White sails home safely) 41.Kh4 and now:
C21) 41...Nd4 42.Ng5+ Kf6 43.Nh7+ Kg7 44.Ng5 Nb5 45.Ne6+ Kf6

analysis diagram
46.Nf8 and White is safe;
C22) 41...Kg7 42.Ng5 Nc7 43.Nf3 Nb5 44.Kg5 Nxa3 45.Ne5 b5 46.Nxg6 b4 47.Nf4 and Black’s b-pawn is no
longer dangerous;
C23) 41...Nc7 42.g4 Kg7 (or 42...fxg4 43.Ng5+ Kf6 44.Kxg4 Nb5 45.Ne4+ Ke5 46.a4 and White is home and dry)
43.Ng5 Nb5 44.gxf5 gxf5 45.Kg3 Nxa3 46.Kf4 Kg6 47.Nf3 and Black can do nothing to keep the f-pawn.
Let’s return to Black’s winning attempt, which is the hardest to judge properly in advance (position after 38.Ng5).

analysis diagram

38...Nxg5, 39.fxg5 Kd6 40.Ke3 Kc5 41.Kd3 Kb5 42.Kc3 Ka4 43.Kb2 b5!
analysis diagram

44.g3! (otherwise 44...b4 wins) 44...b4 (by exchanging, Black clears the way for the king. Kasparov beat Sveshnikov
with this motif once. Of course the most classical example is Cohn-Rubinstein, St Petersburg 1909) 45.axb4 Kxb4
46.Kc2 Kc4 47.Kd2 Kd4 48.Ke2 Ke4 49.Kf2

analysis diagram

49...f4 and Black wins, or so I thought. But when I showed this fascinating endgame to my trainer colleague in
Singapore Tigran Gyozalyan, he suggested I should check it again as White can give up a pawn, and the ending of two
pawns against one could be a draw.
His suspicion turned out to be fully justified:
A) Not 50.h3? Kf5 51.gxf4 Kxf4 52.Kg2 Kxg5 and Black wins;
B) 50.Ke2? fxg3 (50...Kf5 51.Kf3 fxg3 52.Kxg3 Kxg5 and Black wins on account of mutual zugzwang) 51.hxg3 Kf5
52.Kf3 Kxg5 and Black wins;
C) 50.Kg2 Ke3 51.gxf4 Kxf4 52.Kf2 Kxg5 53.Kg3 and White holds;
D) 50.gxf4! Kxf4 51.Kg2 Kxg5 and now the computer can provide a definitive evaluation.
analysis diagram

52.Kg3! – reaching the crucial position with Black to move. If the black pawn were on h6 it would also be a draw, as
later when the king gets to h3 there would no longer be the choice between pushing the pawn one or two squares. Of
course it is extremely difficult to get this far in calculation over the board and even harder to evaluate it properly.
Back to the game.

36...Re7 37.Rb8+ Kg7 38.g3

If 38.Kg3 Nc5.

38...Nd4 39.Ne8+?!

Accelerating the end, but Black is winning anyway.

39...Kf7 40.Nd6+ Ke6 41.Nc4

41...Kd5

Karpov centralizes his king, but it does not look as though he is going to cross the seventh rank.

42.Ne5

This move was sealed.

42...Nc6!

White’s king is cut off, and so Black wins the rook endgame easily. The rest is not so interesting. Black probably did not
want to resign right after the resumption.

43.Nxc6 Kxc6 44.Rg8 Kb5 45.h4 Rf7 46.Ke3 Ka4 47.h5 gxh5 48.Kd4 Kxa3 49.Ke5 b5 50.Rb8 b4 51.Ke6 Rg7
52.Rb5 Ka4

And White finally resigned.

I remember how international soccer has been caricatured by the English goalgetter Gary Lineker: 22 men run around
after the ball and the Germans win in the end. It’s like that in chess: whoever occupies the second rank with a rook, it
makes no difference, Karpov wins the endgame anyway. With some exaggeration we can draw a general conclusion:
two chess players move a limited number of pieces around, Karpov wins.
 GAME 49
Karpov, Anatoly
Gheorghiu, Florin
Moscow European Team Ch. 1977 (3)

1.c4 c5 2.Nf3 Nf6 3.Nc3 e6 4.g3 b6 5.Bg2 Bb7 6.0-0 Be7 7.d4 cxd4 8.Qxd4 d6 9.b3 0-0 10.Rd1 Nbd7 11.Bb2 a6
12.Qe3 Qb8 13.Nd4 Bxg2 14.Kxg2 Qb7+ 15.Qf3 Qxf3+ 16.Nxf3 Rfc8 17.Nd4 Rab8 18.Rac1 h6 19.e4 Ne8 20.f4
Bf6

21.Kf3!

The king stands well on f3.

21...Rb7

Boris Gulko recommended bringing the king to the centre with an immediate 21...Kf8. Now after 22.e5 Gulko’s move
is tactically justified: 22...dxe5 23.Nxe6+? Ke7 is good for Black.

22.Ba3 Rbc7 23.Nce2 Nc5 24.Rd2!?

The usual little improving move by Karpov!

24...g6 25.Nc2!?

Karpov finds a better place for the knight.

25...Bg7

With 25...b5 Black could have tried to become active. Then, after 26.cxb5 axb5 27.Ned4 Bxd4 28.Nxd4 b4 (28...Nxe4
29.Kxe4 Rxc1 30.Bxc1 Rxc1 31.Nxb5 gives White a small edge) 29.Bxb4 Nxe4 30.Rdc2 Rxc2 31.Rxc2 Rxc2 32.Nxc2
f5 the endgame is balanced.
26.Ne3

If 26.g4 g5.

26...f5?!

It turns out that Black will lose the battle on this part of the board. 26...b5!? 27.cxb5 axb5 28.Rdc2 Na6 looks playable.

27.exf5 gxf5 28.h3 h5

29.Rg1!

This is an unpleasant move to face. It is hard for Black to appreciate exactly how much power White has on the
kingside.

29...Rf7

After 29...Kf7, 30.g4 hxg4+ 31.hxg4 fxg4+ 32.Rxg4 followed by f4-f5 gives White a good game.
More productive might, again, be 29...b5 and Black has time to ease White’s grip: 30.g4 (30.Bxc5 Rxc5 31.g4 (after
31.Nd4 Kf7 32.cxb5 Bxd4 33.Rxd4 axb5 34.Rb4 Nc7 Black is not worse) 31...bxc4 32.Nxc4 fxg4+ 33.hxg4 hxg4+
34.Rxg4 Rd8 35.Nd4 Kf7 and Black is in the game) 30...hxg4+ 31.hxg4 fxg4+ 32.Rxg4 bxc4 33.bxc4 Kf8 (or 33...Kf7
34.f5 Be5 35.Nf4 and if White has an edge it is just a small one) 34.Ng3 (in case of 34.f5 Rf7 Black can exchange
enough pawns to equalize) 34...Bc3 35.Rg2 Bd4 36.f5 Bxe3 37.Kxe3 Re7 38.fxe6 Rxe6+ 39.Kd4. Here, White still has
some pressure, but with so few pawns left on the board Black has good drawing chances.

30.g4 hxg4+ 31.hxg4 fxg4+ 32.Rxg4 Kf8

33.Ng3!

Karpov makes a move the point of which is easy to miss.

33...a5?

The former Romanian number one wants to stabilize the queenside, but the real danger is on the other side.
Moving the king away by 33...Ke7 would have been safe for Black: 34.f5 (34.Rg6 Rf6) 34...Be5 35.Kg2 (not 35.Ke2
Nf6 36.Rg6 Bxg3 37.fxe6 Rh7 and Black suddenly becomes active) 35...Nf6 36.Rg6 Bxg3 37.fxe6 Kxe6 38.Bxc5
Rxc5 39.Kxg3 Kd7 and Black holds easily;
By counterattacking with 33...Bc3!? Black could also defend himself: 34.Rh2 (34.Rd1 Ke7 35.f5 Ng7 36.Rg6 Rcf8 and
times will not be so hard for Black) 34...Nd3! 35.Ke4
analysis diagram

35...Ne1! 36.Rh6 Nf6+ and Black is kicking. The position after 37.Rxf6 Rxf6 38.Ne2 Bd2 39.Bxd6+ Kf7 is extremely
unclear.

34.Rg6 Ke7

If the rook leaves the seventh rank with 34...Rf6, there follows 35.Rxf6+ Bxf6 36.f5 Be5 37.fxe6 Bxg3 38.Bxc5 dxc5
39.Kxg3 Rc6 40.Rf2+ Ke7 41.Nd5+ Kxe6 42.Re2+ Kd7 43.Re7+ Kd8 44.Rb7 Nf6 45.Rxb6 Rxb6 46.Nxb6 and
although Black has chances to hold, he will certainly suffer.

35.f5

The point of 33.Ng3 was to weaken the d5-square.

35...Rf6 36.Rxf6 Nxf6

37.Re2

Karpov targets the king on principle. Alternatively, 37.fxe6 Kxe6 38.Ngf5 Bf8 39.Bb2 would have given a good
endgame with decent chances.

37...Rf8??

After 37...e5 38.Bxc5 dxc5 39.Ne4 Black’s position is unenviable, yet this would have avoided immediate collapse.

38.Bxc5! bxc5 39.fxe6 Kxe6 40.Nef5+ 1-0

GAME 50
Karpov, Anatoly
Debarnot, Roberto Luis
Las Palmas 1977 (3)

1.e4 e6 2.d4 d5 3.Nd2 c5 4.exd5 exd5 5.Ngf3 Nc6 6.Bb5 Bd6 7.dxc5 Bxc5 8.0-0 Ne7 9.Nb3 Bd6 10.Re1 0-0 11.Bg5
Bg4 12.Be2 Qc7 13.h3 Bd7 14.Nfd4 Nxd4 15.Nxd4 h6 16.Be3 Bh2+ 17.Kh1 Bf4 18.Bf3 Rfe8 19.c3 Rad8 20.Qb3
Bxe3 21.Rxe3 Qb6 22.Re5 Bc6 23.Rae1 Qxb3 24.axb3 Kf8 25.b4 Ng6 26.Nxc6 bxc6 27.Rxe8+ Rxe8 28.Ra1 Re7
29.g3 Ne5 30.Be2 Rc7 31.Kg2 Ke7 32.f4 Nd7 33.Rd1 Nb6 34.Kf3 c5 35.bxc5 Rxc5 36.Ke3 Ra5 37.Rd4 Ra1 38.Rb4
Kd6 39.h4 Rg1 40.Kf3 Ra1 41.Bd3 Rd1 42.Bb5 Rb1 43.Rb3 Ra1
44.Kg4

The players had reached this ending – with this particular pawn structure – roughly ten moves earlier. After some
manoeuvring, the king now starts to create problems for Black on the kingside. Any particular threat is hard to
distinguish; however, the text provokes a pawn move.

44...g6

On 44...Rb1, 45.Kf5!? may be troublesome.

45.f5!

Karpov forces an enemy weakness on the kingside.

45...Ke5

It is not easy to find the best way for Black to defend the position. There are other reasonable options as well. Black has
to be careful in all cases, but his chances to hold are not bad at all.
After 45...h5+ 46.Kf4 (in the event of 46.Kg5 Rg1 47.fxg6 Rxg3+ 48.Kxh5 fxg6+ (or 48...Rxg6 49.Ra3) 49.Kh6 Rg4
50.Rb4 Nc4 51.b3 Rxh4+ 52.Kg5 Rh3 53.bxc4 Rxc3 Black can take all the white pawns and draw) 46...Na4 47.fxg6
fxg6 48.Be8 Nc5 49.Ra3 Nd3+ (after 49...g5+ 50.Kxg5 Ne4+ 51.Kxh5 Rxa3 52.bxa3 Nf6+ 53.Kg6 Black is in
trouble, or 49...Ne6+ 50.Ke3 Rxa3 51.bxa3 Nf8 52.Bf7 and White is a little better) 50.Ke3 Rxa3 51.bxa3 Ne5 52.Kf4
and White has the advantage as the bishop is now stronger now than the knight.
Alternatively, the line 45...gxf5+ 46.Kxf5 (46.Kh5 Rg1) 46...Kc5 (or 46...Rg1 47.Kf4 Rg2 and the rook ties White up)
47.g4 (47.Kf6 Rg1) 47...Rh1 48.h5 Rh2 (48...a5 49.Ba6 a4 50.Rb5+ Kc6 51.Rb4 and White is somewhat better) 49.Kf6
(49.Rb4 Rf2+) 49...Rg2 50.Kg7 Rxg4+
51.Kxh6. This position is hard to judge, but perhaps a ‘human’ would prefer to play with the white pieces.

46.fxg6 fxg6 47.Rb4!? Re1?!

This is the first step that Black allows White to push him back. Better was the forcing 47...a5! and now:
A) 48.Rf4 Rb1 (48...Ra2 49.Rf2 a4 50.h5 a3 51.Re2+ Kf6 52.bxa3 gxh5+ 53.Kxh5 Rxa3 also holds) 49.Rf2 Nc4
50.Bxc4 dxc4 51.h5 gxh5+ 52.Kxh5 Rh1+ 53.Kg6 Rg1 with equality;
B) 48.Rb3 a4 49.Ra3 Rxa3 50.bxa3 Nc4 51.Bxa4 Nxa3 52.Be8 Kf6 53.Bc6 Nb1 with nothing left to play for.

48.Bd3 Kf6?

It was hard to foresee the consequences of this retreat. Black should have played the brave 48...g5! which looks as
though it holds: 49.hxg5 hxg5:
A) 50.Kxg5? Re3;
B) 50.Rd4 Re3 51.Bb5 Kf6 and the feeling is that Black should not lose (with 51...a5 52.Rd1 Kf6 Black should also
hold);
C) 50.Kf3 Nd7 51.Rb7 Kd6 52.Kf2 Ra1 – White is better but it seems much closer to a draw than to a win.
49.Rf4+!

Now the black king gets pushed back.

49...Kg7 50.Kf3!

White gets closer by centralizing his king.

50...Re5 51.Rb4 Re7 52.Rb5!

Karpov once again makes the shortest possible move. It has a big effect on the outcome of the game. White intends to
improve the rook via c5.
52...Rc7?

Black prevents the rook move, but allows another force to fight in the centre, and this turns out to be decisive. Black
should have kept the white king away from the centre with 52...Kf6. Although after 53.Rc5 the rook will cause a
headache, it seems unable to force the win on its own:
A) with 53...g5 Black can wait. His position is very hard to crack: 54.b4 Ke5 55.hxg5 (55.Kg4 Nd7) 55...hxg5 56.Kg4
(56.Ke3?! Na4 57.Rc6 Nxc3) 56...Kd6 57.Bb5 (in the event of 57.Kxg5 Black gets rid of the last kingside pawn with
57...Re3, which is a relief for him)

57...Re3 and Black is safe as the rook keeps the enemy king and rook busy defending their pawns;
B) 53...Na4 54.Rxd5 (in case of 54.Rb5 Re6 or 54.Rc6+ Re6 55.Rxe6+ Kxe6 White has no advantage) 54...Nxb2
55.Be4 (55.Rd6+ Ke5) 55...Rc7 56.Rd6+ Ke5 57.Rd5+ Kf6 and Black cannot be pushed aside.

53.Ke3!

The block is removed and so the white king gets to the centre, where it will be highly influential.

53...Kf6
54.Kd4

This is an effective square for the king.

54...g5

Creating a weakness, but he has no better option. After 54...Re7 55.Rc5 g5 56.b4 Black is struggling. With the passive
54...Rc8 he achieves nothing, e.g. 55.b4 Rc7 56.Rc5.

55.hxg5+ hxg5

If 55...Kxg5 56.c4 Kg4 57.c5 (or 57.cxd5 Kxg3 58.d6 Rd7 59.Ke5 and White is about to win) 57...Nc8 58.Kxd5 Rd7+
59.Kc4 Kxg3 60.Bf5 Rc7 61.Kd5 White wins.

56.Ra5

Karpov keeps improving the potential of his rook and again, it just requires a short move.

56...Ke6 57.b3 Kf6 58.Ra1 Nd7 59.Ra5 Nb6 60.g4!

White has a threat, and moreover Black is in zugzwang.

60...Ke6

Moving the king causes less problems than moving the rook. After 60...Rd7 61.Rc5!? Re7 62.c4 dxc4 63.Rf5+ Kg6
64.bxc4 Black is losing.
61.c4!

This creates a strong passed c-pawn and also lays bare Black’s weakness on g5. Despite the limited amount of material
on the board White is winning, as Karpov points out.

61...dxc4 62.bxc4 Rd7+ 63.Kc3 Rg7 64.Bf5+ Kf6 65.Kd4 Re7

If 65...Rc7 66.c5.

66.c5 Re5 67.Be4! Nd7 68.Ra6+ Re6 69.Rxe6+ Kxe6 70.Bf5+ Ke7 71.c6 1-0

Another endgame from Las Palmas, where Karpov scored a fabulous +12 =3.

GAME 51
Karpov, Anatoly
Visier Segovia, Fernando
Las Palmas 1977 (14)

1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 g6 5.c4 Bg7 6.Be3 Nf6 7.Nc3 0-0 8.Be2 d6 9.0-0 Nd7 10.Qd2 Nc5 11.Rfd1
Qa5 12.Nb3 Qb4 13.Qc1 Be6 14.Nd5 Nxb3 15.axb3 Qxb3 16.Rd2 Bxd5 17.exd5 Ne5 18.Ra3 Qb4 19.Qc2 b5
20.Rb3 Qa4 21.Rxb5 Qxc2 22.Rxc2 a5 23.Rb7 Rfb8 24.Rxe7 Bf6 25.Rc7 Bd8 26.Ra7 Rxa7 27.Bxa7 Rb7 28.Bd4
Nd7
29.Bf3!?

Black is playing to exchange the dark-squared bishops and then to try and build a fortress on the dark squares with the
knight. Karpov finds a very interesting and original way to fight against this imaginative plan. In case of 29.Kf1 Bf6
30.Bxf6 Nxf6 31.Ke1 Nd7 32.Kd2 Nc5 Black gains control of the queenside. In the event of 29.Bc3 Bf6 30.Bxa5
Bxb2 the position is not easy to break open.

29...Bf6?!

Black doesn’t adjust to White’s move and continues as planned. 29...f5 was one way to cover the g4-square. After
30.g4!? Bf6 31.Bxf6 Nxf6 32.gxf5 gxf5 White would have decent winning chances.
If 29...h5!?:
A) 30.Bc3 Bf6;
B) 30.h3 h4 – Black has a strong grip here;
C) 30.Kf1 Bf6 31.Bxf6 Nxf6 32.c5 (32.Ke2 Nd7) 32...dxc5 33.Rxc5 Rxb2 34.Rxa5 and Black also has real drawing
chances.

30.Bxf6 Nxf6 31.c5! dxc5

Now the c5- and a5-pawns are not to be taken because of the back rank mate.
32.d6!

This move requires some exact calculation.

32...Rb6

Alternatively, 32...Rd7?! would place Black in some difficulty:

A) 33.Rd2 Ne8 34.Rd5 Rxd6 35.Rxc5 Rd2 36.b3 Ra2 37.g4 Kg7 and Black holds;
B) 33.g4 Rxd6 34.Rxc5 a4 35.Rc4 Ra6 and Black is safe;
C) But 33.Rxc5! Rxd6 34.Rxa5 Rd2 35.Ra2 (35.b4 Rb2, or 35.Rb5 Ne4 36.b4 Nxf2 and White has winning chances)
35...Ne4! (this had to be foreseen, as 35...Kg7 gives White time to chase the rook away from the second rank: 36.Kf1!
Rc2 (36...h5 37.Ke1 Rc2 38.Bd1) 37.Bd1)
36.Ra1! Nxf2 (36...Ng5 37.Rb1) 37.Rb1! Nd3 38.b4 and White has reasonable winning chances.

33.Rd2 Nd7

34.g4!

This is the subtle continuation of Karpov’s deep plan. He gains space and fixes Black’s pawns on the kingside at the
same time.

34...h6

Blocking the g-pawn with 34...g5? is not recommended as it gives up control of the f5-square after 35.Be4.
35.h4 Kf8 36.g5 hxg5

36...h5 covers the g4-square, but in vain: 37.Bg2 Rb4 38.Bh3 Ke8 39.Re2+ Kd8 40.Re7 is very promising for White.

37.hxg5 Rb4 38.Kf1 Rd4 39.Ke2 Ke8

It is hard to tell whether the text move or 39...f5! presents White with more problems:
A) 40.gxf6 Kf7!;
B) 40.Bc6 Ne5 41.Bb5

41...Nf3!! (a very pretty move) 42.Rc2 (42.Rd3 Nxg5 43.Rxd4 cxd4 44.Kd3 Ne4 45.d7 Ke7 and the d-pawn will soon
fall) 42...Nxg5 43.d7 Ne6 and Black soon gets the d7-pawn;
C) After 40.Bd5! Nb6 (in case of 40...Rxd2+ 41.Kxd2 Ke8 42.Kc3 Nb6 43.Be6 it is not easy for Black to make a
move; or 40...Rg4 41.Bc6 Ne5 42.Bb5 Re4+ 43.Kf1 Rd4 44.Re2 Re4 45.Rc2 c4 46.Rd2 and White has a dangerous
initiative) 41.Rxd4 cxd4
42.Ba2! (after 42.Be6 Na4 43.d7 Ke7 44.Bf7 Kxd7 Black is safe) 42...Na4 (42...Ke8 43.Kd3) 43.Kd3 Nxb2+
44.Kxd4 Black is in dire straits.

40.Rxd4 cxd4 41.Bc6 Kd8 42.Bd5 f6 43.Kd3 fxg5

44.Kxd4

Black’s position is extremely hard to defend. Even if objectively he might be able to draw, White can play on forever.

44...Nf6 45.Be6 Nh5 46.f3 Nf4 47.Bc4 Ng2 48.Bd5 Kd7 49.Ke5 Kd8 50.Be4 Ne3 51.Kd4 Nf5+ 52.Kc5 Kd7
53.Bc6+ Kd8 54.Bb5 Nh4 55.Be2 Ng2 56.Kd4 Kd7 57.Ke5 Ne3 58.Bb5+ Kd8 59.b3 Nf5 60.Bc4 Nh4 61.Bd5
Ng2 62.Be6 Nf4 63.Bc4 Ng2 64.Bb5 Ne1 65.Bc6 Ng2
66.Kf6

Finally, after very cautious preparations and avoiding all the knight tricks, Karpov’s king approaches the g-pawns.

66...Nf4 67.Bb5 Kc8 68.Ke7

Now the d-pawn becomes unbearably strong.

68...Nd5+ 69.Ke8 1-0

Black resigned: after 69...Nf6+ 70.Kf7 Nd5 71.Bc4 Nb4 (if 71...Nb6 72.Ke8) 72.Ke8 Nc6 73.Be6+ Kb8 74.Bd5
White wins.

GAME 52
Hübner, Robert
Karpov, Anatoly
Tilburg 1977 (7)

1.c4 c5 2.Nf3 Nf6 3.Nc3 d5 4.cxd5 Nxd5 5.g3 g6 6.d3 Bg7 7.Bd2 b6 8.Qa4+ Bd7 9.Qh4 Bc6 10.Bg2 e6 11.Qxd8+
Kxd8 12.Rc1 Na6 13.Nxd5 Bxd5 14.Bc3 f6 15.a3 Ke7 16.0-0 Rhc8 17.Nd2

According to Karpov his positional advantage was gradually crystallizing, while Hübner, in reply, indicated that he
could not find any advantage for Black.
17...Nc7 18.b4?

He opens up the queenside, but Black simply has more pieces in that area. Interestingly, Hübner gives the text a double
question mark, while Karpov gives none. Timman wrote in his article on Kasparov’s book that stronger players analyse
better. It seems to me that the present example contradicts the Dutch grandmaster’s statement. White’s position may
well be lost after this serious mistake.
A) After 18.b3 Nb5 (18...e5 19.Nc4) 19.Bb2 Nd4 20.Rfe1 (in case of 20.Bxd4 cxd4 21.Bxd5 exd5 Black will put his
rook on c3) 20...Bxg2 21.Kxg2 f5 Black is a little better, but White should not panic;
B) With 18.Nc4! the former World Championship candidate shows the way to full equality: 18...Nb5 19.a4! Nxc3
20.bxc3 f5 21.Ne3 (there is nothing serious for White to worry about according to Hübner) 21.Bxd5 exd5 22.Ne3 Ke6
23.c4 dxc4 24.dxc4 is balanced.

18...Bxg2!

According to Karpov this coaxes the king to g2. Another, maybe more important factor of the position is that Black
opens up the queenside where he is stronger and has a good chance to invade.

19.Kxg2
19...cxb4! 20.Bxb4+ Kd7

Hübner stops analysing and commenting on the game from this point, I think this says enough about his opinion on the
rest of the game. He believes it is an easy win for Black, and at this level it is. Karpov, on the other hand, keeps
analysing the details as though it is still a fight. One even senses that Hübner was somewhat upset about the additional
attention. In our opinion the truth lies somewhere in the middle. The game deserves some additional comments, though
maybe fewer than were provided by the winner.

21.Bc3

According to Karpov 21.Nc4 was stronger, but Black’s advantage is beyond doubt here as well: 21...Nd5 22.Bd2 b5
23.Na5 Bf8 24.e4 Rxc1 25.Rxc1 (25.Bxc1 Nc3) 25...Bxa3 26.exd5 Bxc1 27.dxe6+ Kxe6 28.Bxc1 Rc8 and Black is
clearly better because of the c-file and his better king.

21...Nd5 22.Bb2 Bh6! 23.e3

After 23.Rfd1 Rxc1 24.Bxc1 Nc3 (24...Rc8 wins as well) 25.Re1 Na2 26.Nb3 Nxc1 27.Nxc1 Rc8 White loses.
23...Bxe3!

Thus, White gains a clear advantage as his rook controls the c-file and he has the distant passed pawn as well.

24.fxe3 Nxe3+ 25.Kf3 Nxf1 26.Nxf1 Rxc1 27.Bxc1 Rc8 28.Bb2 Rc2!

Karpov would also have found this 15 years earlier as a teenager.

29.Bxf6 Ra2 30.Ke3 Rxa3 31.Nd2 b5 32.Ne4 b4?!

A small error which cedes control of the c4-square. According to Karpov 32...a5 was more accurate.

33.Kd4 a5 34.Kc4 Ra2 35.h4

White misses 35.Kb5! which would force Black to find good moves to keep his advantage: 35...Rxh2! 36.Kxa5 b3
37.Nc5+ Kd6 38.Nxb3
38...Rf2! (collecting the g-pawn in an indirect manner; 38...Rg2 39.d4!) 39.Bc3 Rf3 40.Kb4 Kd5 and Black’s advantage
should be enough for the win.

35...Kc6 36.Bd4

White has slightly stabilized his position.

36...Re2!

Karpov improves his rook nicely by transferring it to the first rank.

37.Be5 Re1! 38.Bf6 Rb1 39.Be7!

After 39.Ng5 a4 (or 39...Rc1+ 40.Kb3 Kb5) 40.Nxh7 a3 41.Ng5 b3 42.Kb4 a2 43.Ka3 Kd5 Black also wins.

39...e5!

Karpov prepares a check on c1, when the king can only go back to the b-file.

40.g4 Rc1+ 41.Kb3 Kd5 42.Bg5

In case of 42.Bd8 Kd4 43.Bxa5 Kxd3 44.Nf6 Rb1+ 45.Ka2 Kc2 wins.

42...Rb1+ 43.Kc2 Rh1 44.Kb3 Rh3! 45.Nf6+ Kd4 46.Nxh7 Rxd3+ 47.Kc2 a4 48.Be7 Rc3+ 49.Kb1 Rc7 0-1
Second Reign (1978-1980)

In the year 1978, before his title match with Kortchnoi, Karpov played only one tournament, in Bugojno, sharing first
place with Spassky. In the Baguio match, Karpov missed Furman (his highly succesful trainer who had beaten seven
World Champions – his passing away didn’t do Karpov’s chess any good) and Razuvaev, but he had the excellent
grandmasters Balashov, Zaitsev and Tal helping him. Karpov started well and took a 5-2 lead, but then he ran out of
energy, losing a lot of weight as well, whereas Kortchnoi started to play very strongly. Interestingly, Karpov broke
down in endgames especially. Finally he managed to win this match, which could have gone either way, by 6-5. We
have selected three endgames, in two of which Karpov had to defend for a long time.
In 1980, Karpov had to withdraw after five rounds in the Bad Kissingen tournament because his father had passed away.
Later on, Karpov shared first prize in the Montreal tournament of stars with Tal, who experienced one of the most
inspired periods of his career; he had won the Riga Interzonal with the highest ever percentage in the previous year.
Karpov’s tournament record in 1980 was less impressive, but his endgame remained very strong, as several games in
this section show. And he did attain an Elo rating of 2725.

GAME 53
Karpov, Anatoly
Larsen, Bent
Bugojno 1978 (1)

1.e4 c6 2.d4 d5 3.Nd2 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.Nf3 Nd7 7.h4 h5 8.Bd3 Bxd3 9.Qxd3 e6 10.Ne4 Qa5+
11.Bd2 Qf5 12.0-0-0 0-0-0 13.Be3 Nh6 14.Neg5 Qxd3 15.Rxd3 Be7 16.Re1 Rhf8 17.Nh3 Ng4 18.Bg5 Rfe8 19.Bxe7
Rxe7 20.Nfg5 Ndf6 21.Rd2 Red7 22.Ree2 g6 23.c3 b6 24.Nf3 c5 25.dxc5 bxc5 26.Nhg5 Kc7 27.Rxd7+ Rxd7
28.Nd2 Nd5 29.g3 Re7 30.Nge4 Kc6 31.Nb3 c4 32.Nd4+ Kb6

33.Nc2!

It is hard to appreciate exactly what Karpov is aiming to achieve. Larsen actually misses that there already is a threat.
33...f5?

This not only helps White to find the target at c4, it also weakens the e5-square. Black had no time to do anything else
than defend the pawn with 33...Kc6!. After 34.Na3 Nb6 35.Nd2 (in case of 35.f3 Nh6 36.Nd2 Nf5 Black also has his
target) 35...Kd5 White will sooner or later play b2-b3 and he is just a fraction better.

34.Nd2 Kc5 35.Na3

It has taken six knight moves to put maximum pressure on the c-pawn.

35...Nb6 36.f3 Nf6 37.Re5+!

White could not hide his intentions with 37.Kc2 because of 37...e5.

37...Nfd5

38.Kc2!

Here is another superb king move in the endgame by Karpov. It looks like just another little improving move, but
actually it has an instant function – it protects the c-pawn.
After 38.Ndxc4 Nxc4 39.b4+ Kc6 40.Nxc4 Nxc3 41.a3 White would also have some advantage.

38...Nd7

In case of 38...a5 39.f4 (or 39.b3 cxb3+ 40.Nxb3+ Kd6 41.f4 a4 42.Nd4 and Black is also slightly worse – it is easy for
him to go wrong) 39...a4 (39...Nd7 40.Re2 N7b6 41.b3 nets a pawn) 40.b4+ axb3+ 41.axb3 cxb3+ 42.Nxb3+ Kc6
43.c4 Nf6 44.Nd4+ Kd7 45.Rb5 the endgame is rather unpleasant for Black.

39.Re1 N5b6?!

Moving the other knight with 39...N7b6 40.f4 Re8 41.Nf3 (41.b4+ cxb3+ 42.axb3 would be tough for Black, too)
41...Nd7 42.b4+ cxb3+ 43.axb3 a5 44.Ne5 finds White not much better than in the game.

40.Ndxc4!

Karpov started to play against the c-pawn on move 30, now thanks to this neat tactical shot the ripened fruit falls from
the tree. If 40...Nxc4, 41.b4+ and Black has to return the knight.

40...e5 41.Rd1

41.b4+ also wins: 41...Kc6 (41...Kd5 42.Na5) 42.Na5+ Kc7 43.Nb5+.

41...Nxc4 42.b4+ Kc6 43.Nxc4

The rest requires no comment. White easily converts his winning advantage.

43...Rg7 44.Rd6+ Kc7 45.Ra6 g5 46.hxg5 Rxg5 47.Rxa7+ Kd8 48.f4! exf4 49.gxf4 Rg2+ 50.Kb3 Rf2 51.Ne3 Nf6
52.Nxf5 Rxf4 53.Nd4 Rf1 54.Ra8+ Ke7 55.a4 Kf7 56.a5 Kg7 57.a6 Nd5 58.Rd8 1-0

GAME 54
Karpov, Anatoly
Miles, Anthony
Bugojno 1978 (3)

1.c4 b6 2.d4 e6 3.d5 Qh4 4.Nc3 Bb4 5.Bd2 Nf6 6.e3 Bxc3 7.Bxc3 Ne4 8.Qc2 Nxc3 9.Qxc3 0-0 10.g3 Qe4 11.f3 Qg6
12.Ne2 Bb7 13.0-0-0 d6 14.g4 Nd7 15.h4 Qf6 16.Qxf6 Nxf6 17.e4 Nd7 18.Nc3 Ne5 19.Be2 Ng6 20.Kd2 Nf4
21.Bd3 Ba6 22.Ke3 Ng2+ 23.Kd2 Nf4 24.Rh2 Rae8 25.b3 Re7 26.Ke3 e5 27.Ne2 Nxd3 28.Rxd3 Ra8 29.Ng3 Bc8
30.b4 a5 31.a3 axb4 32.axb4 Ra4 33.Rb2 Re8 34.Rc3 Bd7 35.c5 Rea8 36.cxd6 cxd6
37.Nf5!

This forces Black to give up the bishop, and it opens the way for the rook to reach the c6- or the c8-square.

37...Bxf5 38.gxf5 Ra3 39.Rc2 Kf8 40.Rxa3 Rxa3+ 41.Kf2

White’s rook gets to the key d6-pawn.

41...Rb3 42.Rc6 Rxb4 43.Rxd6 Ke7 44.Rc6 Kd7

45.h5!
It is not hard to see it now, but Karpov had seen this nice break well in advance. It is possible that he had even seen the
key position with the zugzwang in the possible pawn ending in advance.

45...Rb2+ 46.Kg3 Rb1

After 46...h6 47.f6 g6 48.hxg6 fxg6 49.Rc8 White promotes his f-pawn.

47.h6 g6 48.fxg6 hxg6

Better than 48...Rg1+!? which can lose in an instructive way: 49.Kf2 Rxg6 50.Rxg6 fxg6 51.Ke3 Kd6 52.Kd3 and now:
A) 52...g5? 53.Kc4;
B) 52...b5 53.Kc2 (for 53.Kc3 Kc5 see 52...Kc5) 53...Kc7 (53...Kc5 54.Kc3 b4+ (once Black plays 54...g5, he is
doomed as he will not have a spare move: 55.Kb3 b4 56.d6 Kxd6 57.Kxb4 Kc6 58.Kc4 and White gets the opposition)
loses to 55.Kb3 Kb5 56.d6 Kc6 57.Kxb4 Kxd6 58.Kb5) 54.Kc3 Kb6 (54...Kb7 55.Kb4 Kb6 56.d6 Kc6 57.d7 Kxd7
58.Kxb5 wins) 55.Kb4 g5 56.d6 Kc6 57.d7 Kxd7 58.Kxb5+–;
C) 52...Kc5 53.Kc3 – White now wants Black to move in this position:
C1) 53...Kb5 54.Kb3 Kc5 55.Ka4 g5 (55...b5+ 56.Ka3 Kb6 57.Kb4 wins) 56.Kb3 Kb5 57.Ka3 Kc5 58.Ka4 b5+
59.Ka3 b4+ 60.Kb3 Kb5 61.d6 Kc6 62.Kxb4;
C2) 53...b5 and now:

analysis diagram

C21) 54.Kb3 b4
C211) 55.d6? Kxd6 56.Kxb4 Kc6 57.Kc4 g5 – Black holds as it is a mutual zugzwang with White to move;
C212) 55.Kc2 Kb6 (55...Kb5 56.Kb2 Kb6) 56.Kb2 Kb5 – White will have to settle for the queen ending. 57.Kb3 Kc5
etcetera;
C213) 55.Ka4 Kc4 (55...g5? 56.Kb3 Kb5 57.d6) 56.d6 b3 57.d7 b2 58.d8=Q b1=Q 59.Qd5+ Kc3 60.Qxe5+ – White
probably wins this queen ending but it will take time;
C22) 54.Kb2! (this is the best time to make the triangulation as the b5-square is occupied) 54...Kb6 55.Kb3 Kb7
(55...Kc5 56.Kc3 wins) 56.Kb4 Kb6 57.d6 Kc6 58.d7 and White wins.

49.Rc2!

The rook moves to support the h-pawn from behind.

49...Rg1+ 50.Rg2 Ra1 51.Rh2 Ra8 52.Kg4 f6 53.Rb2 Kc7

After 53...Rh8 54.Rxb6 Rxh6 (54...Ke7 55.h7) 55.Rxf6 wins.

54.Rc2+ Kb7 55.d6 b5

If 55...Rd8 56.h7.

56.d7 1-0

A convincing win against the British grandmaster, who was also capable of producing ‘squeezing’ endgames. In
Torquay 1998 for instance, he beat Nigel Short from a totally drawn position.

GAME 55
Kortchnoi, Viktor
Karpov, Anatoly
Baguio City World Championship 1978 (5)

1.c4 Nf6 2.d4 e6 3.Nc3 Bb4 4.e3 c5 5.Ne2 d5 6.a3 Bxc3+ 7.Nxc3 cxd4 8.exd4 dxc4 9.Bxc4 Nc6 10.Be3 0-0 11.0-0
b6 12.Qd3 Bb7 13.Rad1 h6 14.f3 Ne7 15.Bf2 Nfd5 16.Ba2 Nf4 17.Qd2 Nfg6 18.Bb1 Qd7 19.h4 Rfd8 20.h5 Nf8
21.Bh4 f6 22.Ne4 Nd5 23.g4 Rac8 24.Bg3 Ba6 25.Rfe1 Rc6 26.Rc1 Ne7 27.Rxc6 Qxc6 28.Ba2 Qd7 29.Nd6 Bb7
30.Nxb7 Qxb7 31.Qe3 Kh8 32.Rc1 Nd5 33.Qe4 Qd7 34.Bb1 Qb5 35.b4 Qd7 36.Qd3 Qe7 37.Kf2 f5 38.gxf5 exf5
39.Re1 Qf6 40.Be5 Qh4+ 41.Bg3 Qf6 42.Rh1 Nh7 43.Be5 Qg5 44.Qxf5 Qd2+ 45.Kg3 Nhf6 46.Rg1 Re8 47.Be4
Ne7 48.Qh3 Rc8 49.Kh4 Rc1 50.Qg3 Rxg1 51.Qxg1 Kg8 52.Qg3 Kf7 53.Bg6+ Ke6 54.Qh3+ Kd5 55.Be4+ Nxe4
56.fxe4+ Kxe4 57.Qg4+ Kd3 58.Qf3+ Qe3 59.Kg4 Qxf3+ 60.Kxf3

60...g6

Black has a difficult endgame, but Karpov finds a way to rescue himself.

61.Bd6

White decides to win the piece, but he will be left with a solitary a-pawn and his bishop has the wrong colour. However,
the game is not an automatic draw.
As there is little material on the board, it is difficult to force matters in the bishop versus knight ending. After 61.d5
gxh5 62.d6 Nc6 63.Bf6 (or 63.Kf4 Nb8 64.Kf5 Kc4 65.Ke6 Kb3 and Black takes the pawns) 63...Nb8 64.Kf4 Kc4
65.Ke5 a5 66.bxa5 (66.Ke6 axb4 67.axb4 Kxb4 68.Bd8 Kc5 69.Bc7 Nc6 and Black is safe) 66...bxa5 67.Ke6 Kb5
68.Bd8 Kc6 69.Bxa5 h4 White cannot win.

61...Nf5

Filip mentions 61...gxh5! which he evaluates as equal. In fact, this would have attained the draw much earlier than the
game continuation: 62.Bxe7 Kxd4 63.Ke2 Kc4 64.Kd2 Kb3 65.Kd3 (65.b5 Ka4 66.Kc2 Kxb5) 65...b5 66.Kd4
(66.Bc5 Kxa3) 66...Kxa3 67.Kc5 Ka4 68.Bd8 h4 and Black exchanges all the white pawns.

62.Kf4

After 62.hxg6 Nh4+ 63.Kg4 Nxg6 64.Kh5 Kxd4 65.Kxg6 Kc4 66.Kxh6 (66.Bb8 Kb3) 66...Kb3 Black also draws.

62...Nh4 63.Kg4
63...gxh5+ 64.Kxh4 Kxd4!

Karpov correctly judged this position as a draw. Anyway, he had little option but to play it.

65.Bb8!

Otherwise Black gets rid of all the white pawns.

65...a5 66.Bd6 Kc4 67.Kxh5 a4 68.Kxh6 Kb3

Not 68...b5? which would be going for too much: 69.Kg5 Kb3 70.Kf5 Kxa3 71.Ke5 Kb3 72.Kd5 a3 73.Kc5 and the
white king arrives on the queenside in time to force a win.

69.b5 Kc4 70.Kg5

70...Kxb5

An interesting twist and one of those miracles of chess. Without both his pawns Black would be able to achieve an
elementary draw – without only the b-pawn as well.

71.Kf5 Ka6

After 71...Kc6 72.Ke6 b5 73.Bb4 Kc7 74.Be7 Kc6 75.Bd6 it looks as if Black is in zugzwang, for if he goes to the
corner he becomes stalemated and the b-pawn is forced to move, and White wins. Dramatically, 75...b4!! saves the
game.

72.Ke6 Ka7 73.Kd7 Kb7 74.Be7 Ka7!

Karpov doesn’t move the b-pawn.

75.Kc7 Ka8 76.Bd6 Ka7

Not 76...b5?? when 77.Bc5 wins.

77.Kc8 Ka6

After 77...Ka8? 78.Bb8 b5 79.Kc7 b4 80.axb4 a3 White cannot hold the a-pawn but he can checkmate with the b-pawn.

78.Kb8 b5 79.Bb4!

Kortchnoi makes Karpov suffer, stopping the advance of the b-pawn. He can try winning the position for another 50
moves without taking the b5-pawn.
In the event of 79.Bc5 b4 80.Bxb4 Kb6! Black draws. The other possible king move loses as Black loses the a-pawn
without getting back to the corner.

79...Kb6 80.Kc8
80...Kc6!

The king cannot stay in the corner, as 80...Ka6? leads to 81.Kc7 Ka7 82.Bc5+ Ka6 83.Kc6 Ka5 84.Be3 Ka6 85.Bb6
b4 86.axb4 a3 87.b5 mate.

81.Kd8 Kd5

Black has reached a theoretically drawn position. It is, however, not that easy to hold as he cannot allow the king to be
driven away.

82.Ke7 Ke5 83.Kf7 Kd5 84.Kf6 Kd4 85.Ke6 Ke4 86.Bf8 Kd4 87.Kd6 Ke4 88.Bg7 Kf4 89.Ke6 Kf3 90.Ke5 Kg4
91.Bf6 Kh5

At this point the game was adjourned for the second time. The Soviets could relax as Averbakh had investigated this
endgame back in the mid-1950s.
Current computer programs, by the way, have already learnt how to handle this position as well.

92.Kf5 Kh6 93.Bd4 Kh7 94.Kf6 Kh6 95.Be3+ Kh5 96.Kf5 Kh4 97.Bd2 Kg3 98.Bg5 Kf3 99.Bf4
99...Kg2!

Averbakh had already pointed out in the above-mentioned article that this move leads to a draw.

100.Bd6 Kf3 101.Bh2 Kg2 102.Bc7 Kf3 103.Bd6 Ke3 104.Ke5 Kf3 105.Kd5 Kg4 106.Kc5 Kf5

107.Kxb5

Kortchnoi decides not to waste any more time and takes the pawn. From now on things are easier for Black. Yet it is
better to know the ‘drawing area’. According to Rauzer the king should stay in the a7-b6-c5-d4-e5-f4-g5-h6 zone. He
published his analysis back in 1928.
Incidentally Timman got a similar type of position against Mestel in London, 1982. Mestel’s king was outside the zone
and the Dutch grandmaster went on to win. In the beginning his rook pawn was on the second rank, which provides a
win if the king doesn’t reach the corner. However, matters take a turn – just underlining how interesting chess can be –
if White has pawns on both a2 and a3. Once again this helps the weaker side. That is not the end of the twists: if the
weaker side has two pawns then it serves the opponent again.

107...Ke6 108.Kc6 Kf6 109.Kd7 Kg7 110.Be7 Kg8 111.Ke6 Kg7 112.Bc5 Kg8 113.Kf6 Kh7 114.Kf7 Kh8
115.Bd4+ Kh7 116.Bb2 Kh6 117.Kg8 Kg6 118.Bg7 Kf5 119.Kf7 Kg5 120.Bb2 Kh6 121.Bc1+ Kh7 122.Bd2 Kh8
123.Bc3+ Kh7 124.Bg7 ½-½

Kortchnoi decided to draw with a stalemate. Karpov had saved a very important game. Larsen joked that Kortchnoi has
a problem pronouncing ‘nichya’ (it means ‘nobody’s’ as well as ‘draw’), that is why he drew in this way. The two
adversaries set a record for the longest game ever in a World Championship final.
The motif of the wrongly-coloured bishop and the rook pawn has attracted many composers. Let me show you my own
favourite. This example shows where one’s own pawn can stop the opponent reaching the corner. The composer is also
from the Soviet Union.

Oleg Pervakov, 1989

1.Kf2! Kc7 2.Bh7!! Kd6 3.Kxf3 Ke6 4.Kg4

4...Kf7! 5.Kg5 g6 6.Kh6 g5 7.h3

And White can soon freely take the pawn.

For a long time I had been wondering about the result of a position with the wrongly-coloured bishop and rook pawn
with additional queens. Computer programs gave me the answer: with queens on it’s a win unless the stronger side’s
king is very far away on the other side. With additional rooks it was known to be won, and all endings are winning with
an additional knight against bishop of any colour, as well as with additional rightly-coloured bishops.
 GAME 56
Karpov, Anatoly
Kortchnoi, Viktor
Baguio City World Championship 1978 (14)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Nxe4 6.d4 b5 7.Bb3 d5 8.dxe5 Be6 9.c3 Bc5 10.Nbd2 0-0 11.Bc2 Bf5
12.Nb3 Bg4 13.h3 Bh5 14.g4 Bg6 15.Bxe4 dxe4 16.Nxc5 exf3 17.Bf4 Qxd1 18.Raxd1 Nd8 19.Rd7 Ne6 20.Nxe6
fxe6 21.Be3 Rac8 22.Rfd1 Be4 23.Bc5 Rfe8

24.R7d4!?

After 24.Kh2 Bd5 25.b3 a5 26.Rd4 Bc6 White would have to exchange one of his rooks.

24...Bd5?!

Grandmaster Larsen thinks it is quite possible that Kortchnoi had already chosen a losing plan. The recommendation of
the great Danish player was to play 24...Bc6!. Then, after 25.h4 a5 26.b3 (in case of 26.Kh2 a4 27.Kg3 Bd5 Black has
nothing to worry about) 26...Ra8 27.Kh2 a4, Black probably holds.

25.b3 a5 26.Kh2 Ra8

If 26...a4 27.c4.

27.Kg3 Ra6?!

This is awkward.
Larsen thinks Black’s endgame is tenable after 27...Bc6!. Then, after 28.c4 (or 28.h4 a4 29.b4 Bd5 30.a3 c6 and White
brings a rook to the seventh rank) 28...bxc4 29.bxc4 Rab8 30.Ba3 h6, of course Black has to be careful, but he probably
holds on.
Kasparov mentions 27...a4, but the former World Champion holds back his opinion on whether Black can hold the
position or not. There can follow: 28.c4 bxc4 29.bxc4 Bc6 and now there are three main options:
A) On 30.a3?, 30...Ra5! wins the e5-pawn;
B) 30.Ba3!?. This is Marin’s suggestion. He shows that Black cannot release the grip but he is also not sure whether
White can win if Black defends passively: 30...Ra5 (30...Ra7!? 31.Kf4 Kf7 – it is better to be a live chicken than a dead
lion, that is the motto of this move) 31.Kf4 h6 (31...Ra7 32.h4 Kf7 – it is still possible to remain passive) 32.h4 g5+?
(32...Ra7!? is ugly but it offers a chance to hold; 32...Kf7 was also possible) 33.hxg5 hxg5+ 34.Kxg5 Rxe5+ 35.Kf6
Re2 36.Rd8 and this wins according to Marin;
C) 30.Bb4. White’s idea is to play 31.a3, then h4 h5 and, if Black allows, he pushes the g-pawn all the way to g6.
White can follow up with Rf4 and Rf7. If Black plays ...g6 himself, then White switches both rooks to the h-file. Black
plays 30...h6 with the idea ...Kh7 to avoid these difficulties, and now 31.a3 Kh7 32.Kh4 is Bent Larsen’s
recommendation, in order to mount the pressure. He doesn’t guarantee that it will bring about Black’s downfall.

28.h4 Rc6?

This is a bad mistake. As White’s response is the only move it seems that Viktor had missed Karpov’s 33rd move.

29.Rxd5!

Nice, and it’s not too difficult to see that it’s winning.

29...exd5 30.Rxd5 Rce6 31.Bd4 c6 32.Rc5

The rook is caught in a cage, but White can blow it apart.

32...Rf8?

Black defends the f3-pawn instead of exchanging White’s very strong rook by 32...Rd8! which promises a tougher fight.
It might even have saved the game against other opponents. Against Karpov the result is less in doubt. According to
Kasparov’s analysis Black still holds.
The defence of this position is a tremendous task. The position really caught Kasparov’s attention and he delves deeply
into it. In his analysis one can see the hands of a genius, he demonstrates many remarkable ideas. However, in the end
he made a clearly decisive mistake.
An attempt is made here to improve on Garry’s critical line at certain points. The lines are so long that a definitive
evaluation is not possible. However, it seems that one of the suggestions is more than just hopeful. If all lines are
correct it would have been a nearly impossible task to hold the position over the board without an adjournment. Even if
we consider the very special abilities of Karpov’s analysis team, it would have been difficult to come with an answer
within anything less than 24 hours.
33.Kxf3 Rd5! 34.Rxd5 (34.Ke4? Rexe5+) 34...cxd5

analysis diagram

A) 35.Ke3 h5 (Kasparov gives this move an exclamation mark. Other commentators mention 35...Rh6 followed by
36.h5 g6 and Black is stirring up trouble, e.g. 37.Kf4 gxh5 38.Kg5?? h4 and Black wins) 36.gxh5 Rh6 37.Kf4 Rxh5
38.Kg4 g6 39.a4 bxa4 40.bxa4 Kf7 41.f4 Ke6 42.Bb6 Rh8 and now:
A1) 43.Bxa5 Ra8 44.Bb6 Rxa4 45.Kg5 Ra2 46.Kxg6
analysis diagram

46...Rg2+ and Black draws because White’s king is pushed back to the edge of the board, as Kasparov has pointed out.
Without the king, his pawns cannot march;
A2) 43.Kg5 (White cannot push his pawns in a constructive way) 43...Rh5+ 44.Kxg6 Rxh4 45.f5+ Kxe5 46.Bc7+ Ke4
47.f6 Rg4+ 48.Kh5 Rg1 49.f7 Rf1 50.Kg6 d4 and Black holds as White’s bishop is of the wrong colour;
B) 35.Kg3. Kasparov shows with an impressive line how Black draws here as well: 35...b4! 36.f4 Rc6 37.cxb4 axb4
38.f5 Rc2 39.e6 Rxa2 (Black’s counterplay comes in time to save the game) 40.Kf3! Ra3 41.Kf4 Rxb3 42.g5 Rh3
43.Bc5 Rxh4+ 44.Ke5 b3 45.f6

analysis diagram
45...Re4+ (45...gxf6+? 46.gxf6 Re4+ loses to 47.Kf5) 46.Kxd5 gxf6 47.gxf6 Rf4. As a reward for his accurate play
Black can give up his rook for the pawns and draw;
C) 35.h5 b4 36.Ke3 (White brings on the king to stop the rook operations on the queenside) 36...Rc6 37.Kd3

analysis diagram

37...g6! (as the White king has left the area, Black acts before the white pawns start to roll) 38.hxg6 Rxg6 39.f3 Kf7
40.cxb4 axb4 41.Bc5 Ra6 42.Kd4 (42.Bxb4 Rxa2 43.Bd6 Rf2 44.Ke3 Rf1 45.b4 d4+ 46.Kxd4 Rxf3 47.b5 Ke6 48.b6
Rb3 and Black holds again) 42...Rxa2 43.Kxd5 Rd2+ 44.Bd4 Rd3 45.f4 Rxb3 46.e6+ Ke8 47.f5 Rd3 48.Ke4 Rg3
49.Kf4 Rd3. The rook denies White the time to push the f-pawn. Kasparov has found a great role for this rook to play;
D) 35.a3!. This subtle move is Larsen’s suggestion. He published his book on the Baguio World Championship match
just a few weeks after it had ended. Naturally it is quite possible that Karpov saw this possibility during the game.
Karpov’s main tactics during the match were to play fast and bring the creative Kortchnoi into time-trouble. He won a
few games because of Viktor’s time-trouble, but he spoiled some wins as well.
analysis diagram

D1) If 35...g6 36.Kg3 (with 36.h5 Kf7 37.Kf4 White advances his pawn roller and wins according to the Dane. But
with 37...h6 it looks as if Black can hold on the kingside) 36...h5 (36...Re8 37.f4 Rf8 38.h5 wins for White according to
Kasparov) 37.f4 hxg4 38.Kxg4 Kf7 39.Kg5 (after 39.h5 gxh5+ 40.Kxh5, cutting off the king with 40...Rg6 saves
Black) 39...Re8 40.f5 gxf5 41.Kxf5 Re6 42.Be3 Rc6 43.Bd2 b4 44.axb4 axb4 45.cxb4 Rc2 46.e6+ wins as Kasparov
has pointed out;
D2) 35...Rc6!!. Only Kasparov mentions this subtle rook move, with which Black looks for counterplay at once. He
first shows three winning attempts as well as the remedy to each of them:
D21) Firstly, with 36.b4 White can stop Black’s pressure on the c-pawn, but this does allow pressure on the a-pawn:
36...axb4 37.axb4 Kf7.
White would now like to push his pawns. With 38.Kf4 he could try to invade the kingside with his king. 38...h6! is the
way to stop this invasion (38...Ra6 39.Kf5 g6+ 40.Kg5 Ra1 41.f4 Rf1 42.f5 gxf5 43.gxf5 h6+ 44.Kg4 is winning for
White, according to Kasparov): 39.h5 g6 40.hxg6+ Rxg6 41.Kf5 Rg5+ and Black holds.
Therefore the right continuation is 38.Kg3 Ke6 and now:
D211) 39.f4 (White would like to play f4-f5, but Black can prevent this and hold) 39...g6 40.h5 Ra6 41.Kh4 Ra2
42.Kg5 Rg2. Kasparov shows how the active rook can save Black with this pin;
D212) 39.Kf4, combining these two plans, creates a different situation.
analysis diagram

39...Rc7! (alternatives are: 39...Rc8 40.Kg5 (40.Bc5!?) 40...Rf8 41.h5 Rf3 and Black prevents the pawn march, or
39...h6 40.h5 Ra6 41.Kg3 Ra2 42.f4 Kf7 43.f5 Re2 44.e6+ and Black still probably draws; however, White has created
some chances. He can contemplate playing g4-g5 and Kg4) 40.h5 Rf7+! (in case of 40...h6 41.Ke3, the white pawns
are free to go – bad news for Black) 41.Kg3 g5 and Black successfully blocks the kingside.
D22) Secondly, after 36.Kg3 Kasparov shows that White has no time to carry out f2-f4 and e5-e6: 36...b4 37.axb4 axb4
38.cxb4 Rc1! 39.e6 Re1 40.Be3 Kf8 41.Kf4 Ke7 42.Ke5

analysis diagram
42...d4 43.Kxd4 Kxe6 and Black is in no danger of losing despite the fact that White has three pawns for the exchange.
Black’s rook activity is adequate compensation;
D23) Thirdly, 36.Kf4. Garry does not run out of winning attempts, but his active employment of the rook saves Black
again: 36...Kf7 (36...h6? 37.h5!) 37.Kf5 g6+ 38.Kg5 b4! 39.cxb4 axb4 40.axb4 Rc1 41.b5 (41.f4 Rd1! 42.Bc5 d4 43.f5
gxf5 44.gxf5 h6+! 45.Kf4 d3 46.e6+ Ke8 47.Ke3 d2 and Black captures one of the connected passed pawns, which is
enough) 41...Rb1 42.f4 Rxb3 43.f5 Rxb5 44.e6+ Ke7 45.h5 (45.Kh6 gxf5 46.gxf5 Rb3 47.Kxh7 Rg3=) 45...gxf5
46.gxf5 Ra5 47.Bf6+ Ke8 48.Bg7 d4. This is the most elegant draw amongst Black’s options.
Next, Kasparov gives the fourth, and winning, alternative:
D24) 36.h5!!

analysis diagram

It is worth checking if Black can follow the same course as in the second option above (line D22 with 36.Kg3). So
36...b4 37.axb4 axb4 38.cxb4 Rc1 39.b5! (in case of 39.e6 Black draws like in the 36.Kg3 line, as it makes no
difference whether the king stands on g3 or f3) 39...Rb1 40.b6 Rxb3+
analysis diagram

41.Ke2! (this move changes the picture completely compared with the 36.Kg3 line) 41...g5 42.hxg6 hxg6 43.f4 Kf7
(Black misses just one tempo) 44.f5 with decent winning chances.
Therefore Black’s main defence would be 36...Kf7 37.Kg3 and now in the event of 37...g5, after 38.hxg6+ Rxg6 (in
case of 38...hxg6 39.f4 Ke6 40.Kh4 Rc8 41.Kg5 Rg8 Black is again short of one move: 42.a4! fixes the a5-pawn in
time: 42...bxa4 43.bxa4 Ke7 44.Bb6 Kf7 45.Bxa5 Ra8 46.Bb6 Rxa4 47.f5 gxf5 48.gxf5 Rc4 49.Bd4 and White wins)
39.Kh4 (39.f4 h5) 39...Rh6+ 40.Kg5

analysis diagram
A) 40...Rg6+ 41.Kh5 Rg8 (41...h6 42.f4) 42.f4 b4 43.axb4 axb4 44.cxb4 Rb8 45.f5 Rxb4 46.e6+ Ke8 (if 46...Kg8,
47.Bc5 Rxb3 48.f6 wins) 47.Bc5 Re4 48.Kh6 d4 49.f6! Rxe6 50.Kg7 spells the end for Black;
B) 40...Rh1 41.f4 a4 42.bxa4 bxa4 43.f5 h6+ 44.Kf4 Re1 45.e6+ Kf8 46.Kg3 Re2 (46...Re4 loses to 47.Kf3 Re1
48.Be3 Kg7 49.Kf2 Rb1 50.Bd4+) 47.Kf3 Re4 48.Bc5+ Kg7

analysis diagram

49.Be7 Re1 and White’s f-pawn grows too strong;

C) 40...Rh3 (an interesting moment. This time the activation comes too late) 41.f4 b4 42.cxb4!! (this does not give
Black time to activate the rook. The prosaic 42.axb4 may work as well: 42...axb4 43.cxb4 Rd3 (43...Rxb3 loses to
44.Kf5 Rxb4 45.e6+ Ke8 46.Ke5) 44.Bc5 Ke6 45.b5! (45.Bd6 d4 46.f5+ Kd5 47.Kf6 and in this extremely unusual
situation the position may well be balanced) 45...Rxb3 46.b6 d4 47.Bxd4 Kd5 48.Bf2 Rb2 (the rook harasses the
bishop) 49.f5! (this still looks winning) 49...Kxe5 50.Bg3+ Kd5 51.Bc7 Kc6 52.Kh6 and White wins) and now
42...h6+ 43.Kf5 Rxb3 44.e6+ Ke8 45.Ke5 wins for White.
Alternatively, 42...Rd3
analysis diagram

43.bxa5!! Rxd4 44.a6 – it is difficult to recall another game where pawns dominate a rook like this.
Finally, 42...axb4 43.f5! Rd3 (43...bxa3 44.e6+ Ke8 45.f6 Rf3 46.Bc5) 44.e6+ Kg8 45.Bc5 bxa3 46.f6 spells the end.
Black’s alternative to 37...g5 (36...Kf7 37.Kg3) is 37...b4.

analysis diagram

38.axb4 axb4 39.cxb4 Rc1 40.f4 Rb1? (this is the line Kasparov has examined thoroughly) 41.f5 Rxb3+ 42.Kf4 Rxb4
43.e6+ Kg8 44.Ke5 h6 45.Kxd5
analysis diagram

A) 45...Rb8 46.Bc5 (46.g5!? does the trick as well) 46...Rd8+ 47.Kc6 Rd1 48.Bd6 Ra1 49.Kd7 Ra7+ 50.Bc7 Ra8 51.g5
hxg5 52.h6 g4 53.hxg7 Kxg7 54.Be5+ wins for White, as Kasparov has pointed out;
B) 45...Rb3 (this is the only square that Black can use to go after the g-pawn) 46.g5! (46.e7 Kf7 47.Bxg7 Kxe7
48.Bxh6 Rb4 49.g5 Rb5+ 50.Ke4 Rb4+ 51.Ke3 Rb5 52.f6+ Ke6 may be a draw because of White’s very passive
bishop) 46...hxg5 47.h6 gxh6 48.f6 Rb7 49.Bc5 (this is the most accurate win. The white king has to go to g6) 49...g4
(49...Rh7 50.Ke4) 50.Ke4! g3 (50...Rc7 51.Bd6) 51.Kf5 Rc7 52.Be3! and the king reaches g6 safely;
C) 45...Ra4. Black tries to pin the bishop, but... 46.g5! (this breakthrough finally subdues Black) 46...hxg5 47.h6 gxh6
48.f6
analysis diagram

48...Rxd4+. Was it possible that Kasparov had found a new motif when he wrote that this sacrifice saves Black? When I
checked with the computer it turned out he had not: 49.Kxd4 g4 (after 49...h5 50.Kd5 White is faster)

analysis diagram

C1) Kasparov analyses only 50.Ke5?. This move is a mistake: 50...g3 51.f7+ Kg7 52.Kd6 g2 53.Ke7 g1=Q 54.f8=Q+
Kh7 and Black should be able to hold this unpleasant ending;
C2) 50.Ke3 (the least convincing winning move; Black reaches a lost queen ending) 50...h5

analysis diagram
51.e7! (now this is the only move; Black must be prevented from reaching f8) 51...Kf7 52.Kf4! Ke8! 53.Ke5 Kd7
54.Kf5 g3 (54...Ke8 55.Ke6) 55.Kg6 g2 56.Kf7 g1=Q 57.e8=Q+ and the queen ending is lost;
C3) 50.Ke4! is the simplest solution. The white king runs to help the pawn without hesitation: 50...h5 51.Kf5 Kf8
(51...g3 52.Kg6 g2 53.f7+) 52.Kg6 g3 53.f7 and it is all over;
C4) 50.f7+! (this is the nicest and the most instructive win) 50...Kf8 (Black’s best try. After 50...Kg7 51.Ke4 h5
52.Kf4 Kf8 53.Kg5 Kg7 54.Kh4 Kf8

analysis diagram

55.Kxh5! g3 56.Kg6 g2 57.Kf6 g1=Q 58.e7 he gets checkmated) 51.Ke5! Ke7

analysis diagram

52.Ke4! (the final move of the triangulation brings Black into zugzwang. White has no other win) 52...h5 53.Kf4 (this
is a mutual zugzwang) 53...Kf8 54.Kg5 Ke7 55.Kh4 Kf8 56.Kxh5 wins.
Instead of 40...Rb1, 40...Rd1!? is a last attempt to save Black. This position may well be extremely important for the
assessment of Black’s entire defensive plan starting on move 32 in our analysis.
According to my analyses Black can hold with this move: 41.Bc5 Rd3+ 42.Kh4 (in case of 42.Kf2 Ke6 43.Bd6 g6
44.Ke2 Rxb3 Black is safe) 42...Ke6 43.b5 (43.Kg5 d4 44.b5 Rxb3 45.Bxd4 Rxb5 46.h6 g6 47.f5+ gxf5 48.gxf5+ Kd5
and Black keeps attacking the bishop; or 43.Bf8 Rxb3 44.Bxg7 Kf7 and the position is probably equal) 43...Rxb3 44.b6
d4 (44...h6 45.Bd4 Kf7 46.f5 Kg8 47.e6 Rb4 48.Bc5 Rb5 49.Bf2 Rb4 and Black resists) 45.Bxd4 Kd5 46.Bg1 Rb1
47.Bf2 Rb3 48.Kg5

analysis diagram

48...Rb2. It looks as though Black draws despite the considerable material deficit.
I expect there will be further developments in the understanding of this fascinating endgame. The foundation was laid
by Bent Larsen, it was enriched by Karpov and Marin and topped by a huge contribution from Kasparov. To wager on
an objective evaluation would be difficult, but it appears to be a draw.
Finally, let me show you my favourite study with this motif. The build-up to the climax is equally attractive. The name
of the author as well as the source suggests he was a Hungarian.

I. Denes
Magyar Sakkvilág 1930
1.h4

White creates a passed pawn.

1...Ke4 2.h5 gxh5 3.gxh5 Kf5 4.Kd2!

White wants to arrive at e3 at exactly the right time.

4...c5

Black cannot break the grip of White’s pawns with 4...f6 5.h6 Kg6 because of the breakthrough 6.f5+! Kxh6 7.fxe6
Kg6 8.e7 Kf7 9.exf6 Ke8
10.Kc2! (White intentionally loses the right to move) 10...c5 11.Kd3 Kd7 12.Kc4 (Black is in zugzwang) 12...Ke8
13.Kxc5 d3 14.Kd6 and White wins.

5.Kd3

The c-pawn cannot be stopped on the fifth rank.

5...f6

Black has no other move.

6.h6 Kg6

7.f5+! Kxh6 8.fxe6 Kg6 9.e7 Kf7

Black seems to stabilize his position.

10.exf6 Ke8 11.Kc4 Kd7

Now it looks as if White is in zugzwang.

12.a6!

However, he loses the right of move by sacrificing a pawn.

12...bxa6 13.a5

Thanks to the twofold tempo loss, it is Black who is in zugzwang.

13...Ke8 14.Kxc5 d3 15.Kd6

White wins.
Returning to the game.
33.a4!

White wins the second pawn and keeps all the positive elements of his position.

33...bxa4 34.bxa4 g6

If 34...Ra8 35.Kxf3.

35.Rxa5 Ree8 36.Ra7 Rf7 37.Ra6!

With the help of a small finesse White improves his bishop. Not 37.Rxf7?, which would help Black’s king: 37...Kxf7
38.a5 Ke6 and Black has become active.

37...Rc7
38.Bc5! Rcc8 39.Bd6 Ra8 40.Rxc6 Rxa4 41.Kxf3 h5

At this point the game was adjourned. Kortchnoi had no real hope of survival.

42.gxh5 gxh5 43.c4 Ra2 44.Rb6 Kf7 45.c5 Ra4 46.c6 Ke6 47.c7 Kd7 48.Rb8 Rc8 49.Ke3 Rxh4

50.e6+

Kortchnoi resigned on account of 50...Kxe6 51.Bg3! or 50...Kxd6 51.Rxc8 Rc4 52.Rd8+ Kxc7 53.e7.

GAME 57
 Kortchnoi, Viktor
Karpov, Anatoly
Baguio City World Championship 1978 (17)

1.c4 Nf6 2.Nc3 e6 3.d4 Bb4 4.e3 0-0 5.Bd3 c5 6.d5 b5 7.dxe6 fxe6 8.cxb5 a6 9.Ne2 d5 10.0-0 e5 11.a3 axb5
12.Bxb5 Bxc3 13.bxc3 Ba6 14.Rb1 Qd6 15.c4 d4 16.Ng3 Nc6 17.a4 Na5 18.Qd3 Qe6 19.exd4 cxd4 20.c5 Rfc8
21.f4 Rxc5 22.Bxa6 Qxa6 23.Qxa6 Rxa6 24.Ba3 Rd5 25.Nf5 Kf7 26.fxe5 Rxe5 27.Rb5 Nc4 28.Rb7+ Ke6 29.Nxd4+
Kd5 30.Nf3 Nxa3 31.Nxe5 Kxe5 32.Re7+ Kd4

33.Rxg7?!

The king could have been driven far away with 33.Rd1+ Kc3 34.Rc7+ Kb3 35.Rxg7 Rxa4 (with 35...Kxa4 Black
probably draws, but it is quite a task) 36.Rf7 Ra6 37.Rf1 Nd5 38.Rxh7 and Black’s pieces should be able to get back in
time to the kingside to draw.
Kasparov shows that the best move is 33.Rf4+! (holding on to the a-pawn for a while is more important than driving the
king away) 33...Kc5 34.Rxg7 and according to Kasparov, White has chances to win.

33...Nc4 34.Rf4+ Ne4

After 34...Kd3 35.h3! Ne3 36.a5 Nfd5 37.Ra4, again according to Kasparov, the draw is still a long way off.

35.Rd7+

Not 35.Rxh7!? Rxa4 36.Re7 Ra1+ 37.Rf1 Ra2 and Black probably draws.
35...Ke3!

Karpov has attacking ideas on his mind.

After 35...Ke5 36.g3 (in case of 36.Rh4 Ncd6 Black is safe) 36...Ncd6 37.Rxh7 Rxa4 38.Re7+ Kd5 39.g4 Black
probably holds.

36.Rf3+ Ke2 37.Rxh7

Kasparov shows that Kortchnoi still had a reasonable chance to play for a win with 37.Re7! Nd2 38.Ra3

38...Ra5! (Black will have more problems if White can push the pawn to a5) 39.Ra2 Kd1 40.Rxh7 Nc3! – Black will
escape, but only with the aid of consistently good moves.
37...Ncd2

After 37...Rxa4 Black draws as well.

38.Ra3

Kasparov suggests that 38.Rff7 would really have made Black work for the draw: 38...Rxa4 (38...Rc6 39.Rc7) 39.h4.

38...Rc6!

With one move Karpov sets up two different checkmates.

39.Ra1??

Kortchnoi prevents the obvious one, but misses the other. Creating a luft with 39.h4?? does not help: 39...Rc1+ 40.Kh2
Nf1+ 41.Kh1 Nfg3+ 42.Kh2 Rh1 mate. White has to free the g2-square for the king to avoid checkmate:
A) 39.g4 Nf3+ 40.Kg2 Ne1+ 41.Kg1 Rc1 42.Ra2+! and White escapes;
B) With 39.g3! Nf3+ 40.Kg2 Ne1+ 41.Kh1 White can still try; 41...Rf6! 42.h4 (White has a perpetual with 42.Ra2+)
42...Nf2+ 43.Kg1 Nf3+ 44.Kg2 (44.Rxf3 Kxf3 gives White no real winning chances) 44...Ne1+ and this time it is
Black who holds the perpetual.

39...Nf3+!!

White resigned. Black traps the king in the corner with this beautiful tactical shot. Kortchnoi paid a high price for
getting himself caught into zugzwang.
There could have followed: 40.gxf3 (40.Kh1 Nf2 mate would hasten the end) 40...Rg6+ 41.Kh1 Nf2 mate.

GAME 58
Spassky, Boris
Karpov, Anatoly
 Montreal 1979 (13)

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.Bf4 Bb7 5.e3 Be7 6.Nc3 Nh5 7.Bg3 d6 8.Bd3 Nd7 9.0-0 g6 10.h3 Nxg3 11.fxg3 0-0
12.Rc1 Bf6 13.Rc2 Bg7 14.Rcf2 Qe7 15.Kh2 a6 16.Qe2 Rae8 17.Bb1 c6 18.a3 f5 19.e4 c5 20.exf5 exf5 21.Qxe7
Rxe7 22.dxc5

22...bxc5!

A surprising recapture. But it has its logic. It fixes c4 and opens the b-file. As the b1-bishop is restricted, Black hopes to
get some play. On 22...Nxc5 White would play 23.Nd5 Bxd5 24.cxd5.

23.Rd1

In case of 23.Nd5 Bxd5 24.cxd5 Rb8 Black’s play on the b-file wins the b2-pawn.
23...Bxc3!

This is another subtle move that gives up the strong bishop. Why? Karpov transfers from one advantage into a different
one. Black removes an enemy piece so that he can invade. In addition he saddles White up with doubled pawns.

24.bxc3 Rf6 25.Rfd2 Re3! 26.Ng1

White’s minor pieces are impotent. Not 26.Rxd6? Rxd6 27.Rxd6 Bxf3 28.gxf3 (28.Rxd7 Re2 wins) 28...Ne5 and Black
is winning. After all these exchanges the pitiful role of the b1-bishop can be felt.

26...Kf7!
Karpov chooses the best moment to centralize his king.

27.Rxd6 Rxd6 28.Rxd6 Ke7 29.Rd3 Re1 30.Ba2 Rc1!

This pins the white rook down to the defence of the c-pawn and threatens to kill White on the second rank.

31.Nf3

31...Bxf3!

The fewer pieces on the board, the greater the difference between the knight and bishop.

32.Rxf3 Ne5 33.Re3 Kf6 34.Bb3 a5

Spassky is in zugzwang.

35.Ba4

Giving up a pawn, but it doesn’t help. After 35.a4 h6 36.Ba2 Rc2 37.Bb3 Rb2 (37...Rd2 paralyses White as well)
38.Bd1 Rb1 39.Bc2 Rc1 40.Bb3 g5 White is in deep trouble.

35...Nxc4

Karpov starts collecting the pawns. Black is lost.

36.Re8 Rxc3 37.Rc8 Ne3

A perfect location for the knight to attack g2 and to escort the c-pawn on its march.

38.Bb5 c4 39.Kg1 Rc2 40.Bc6 c3 41.Bf3 g5 42.g4 f4 0-1

 GAME 59
Sosonko, Genna
Karpov, Anatoly
Waddinxveen 1979 (1)

1.d4 Nf6 2.c4 e6 3.g3 d5 4.Bg2 Be7 5.Nf3 0-0 6.0-0 dxc4 7.Qc2 a6 8.Qxc4 b5 9.Qc2 Bb7 10.Bd2 Be4 11.Qc1 b4
12.Bg5 h6 13.Bxf6 Bxf6 14.Nbd2 Bd5 15.Qc2 Nd7 16.e4 Bb7 17.Rfd1 Be7 18.Nc4 c5 19.dxc5 Qc7 20.Rxd7 Qxd7
21.Nb6 Qb5 22.Nxa8 Rxa8 23.Rc1 Rc8 24.Nd4 Qxc5 25.Qxc5 Bxc5 26.Nb3 Bb6 27.e5

27...Rxc1+!

Karpov saves the bishop pair.

28.Nxc1 Bc8! 29.Be4

If 29.Kf1 a5.

29...a5 30.Kg2 Bd4 31.Nd3

31...f5!

Karpov now opens up the position, reducing the chances of having to face a solid fortress. Interestingly, with 31...Ba6?!
32.b3 Bxd3? 33.Bxd3 Bxe5 Black would win a pawn but not the game.

32.exf6

White cannot give up a pawn with 32.Bf3 to reach an opposite-coloured bishop ending because 32...Ba6 33.Be2 Bxb2!
wins.

32...gxf6 33.g4 Kf8

Karpov probably intentionally removes his king from the colour of the white bishop. In this case it makes no difference.

34.b3 Ke7
35.Kg3?!

The king aims for f4, hoping to tie Black’s hands, but fails to achieve this.
A) With 35.Kf3 White could have tried to transfer the king to e2 as Matanovic has recommended. 35...f5!? denies him
this opportunity. 36.gxf5 exf5 37.Ba8 Ba6 38.Nf4 (38.Ke2 Bxf2!) 38...Kd6 and Black is still much better;
B) With 35.f4!? White could have removed the f-pawn first and only then moved to e2 with the king. This was his best
chance. White may survive 35...Kd6 36.Kf3 e5 37.h4 Be6 38.fxe5+ fxe5 39.g5 hxg5 40.hxg5 Bc3 41.g6, and in the
event of 35...f5 36.gxf5 exf5 37.Bf3 Ba6 (after 37...Kd6 38.Kf1 Ba6 39.Ke2 White has a kind of fortress) 38.Nf2 Kd6
39.Bd1 Be3 (in case of 39...Bb2 40.Bc2 Bc8 41.Kf3 it requires an all-out effort to squeeze a win out of it. In fact,
White just might hold) 40.Kf3 Bc1 41.Bc2 White still resists.

35...Kd6 36.Kf4 Bd7 37.h3?!

After 37.h4 Bc3 38.h5 e5+ 39.Kg3 Be6 40.Nc1 (40.Kf3 a4) 40...Bd2 41.Nd3 a4 42.bxa4 Bxa2 43.a5 Bc4 White is
struggling on.

37...Bb5 38.f3?!

White doesn’t know what to do. It is very hard to defend such passive positions. Matanovic prefers 38.Kf3.

38...Bd7 39.Kg3?

Sosonko probably anticipated ...e5, ...Be6, ...Bc3, with ...a4 to come, and voluntarily steps out of the check.
39.h4 would have been useful in case a pawn race should occur. The direct attempt to obtain an advantage with 39...e5+
is not convincing: 40.Kg3 Be6 41.f4 Bc3 (41...exf4+ 42.Nxf4 Bg8 43.Bb1 and White’s position is solid) 42.fxe5+ fxe5
43.g5 (it is better to push the pawn quickly, keeping Black occupied) 43...hxg5 44.hxg5 a4 45.bxa4 Bxa2 46.g6 (46.a5
Bc4) 46...Ke7 47.a5 Bc4. Here, Black has an advantage; however, there are very few pawns left on the board. That
provides White with a realistic chance to hold.
With 39...Bc3 Black could postpone his action and play for ...f5, for example 40.Bh7 (40.Bb7 Bd2+ 41.Kg3 f5 and
Black is somewhat better) 40...Bd2+ 41.Kg3 Kd5 42.Kf2 Bb5 (the direct invasion with 42...Kd4 can be met by
43.Ke2 Kc3 (43...Bc3 44.Nf4) 44.Nc5 Bb5+ 45.Bd3 Bxd3+ 46.Nxd3 e5 47.Nc5 Bf4 48.Nb7 and White is still in the
game) 43.Ke2 Bf4 and White has to suffer.

39...f5!

Karpov invades from another angle.

40.gxf5 exf5 41.Bb7 Bb5 42.Ne1

After 42.Nf4 Be5 the pin wins.

42...Bc3 43.Nc2 0-1

43...Bd3 may have been Karpov’s sealed move. After 44.Ne3 Bb1 45.Nc4+ Kc5 wins.
 GAME 60
Karpov, Anatoly
Hort, Vlastimil
Waddinxveen 1979 (6)

1.c4 Nf6 2.Nf3 e6 3.Nc3 c5 4.g3 b6 5.Bg2 Bb7 6.0-0 d5 7.cxd5 Nxd5 8.d4 Nxc3 9.bxc3 Nd7 10.Re1 cxd4 11.cxd4
Bb4 12.Bg5 f6 13.Bd2 Bxd2 14.Qxd2 Rc8 15.Qd3 Qe7 16.Rac1 0-0 17.Ng5 fxg5 18.Bxb7 Nc5 19.dxc5 Qxb7
20.Qe3 Rxc5 21.Rxc5 bxc5

22.Rc1! Qd5

Understandably, the Czech-born grandmaster wants to remove Karpov’s only queenside pawn.
A) The most natural defence 22...Rc8? would have failed because of the weak back rank: 23.Qxe6+ Kh8 24.Rxc5! and
White chops off the pawn;
B) Protecting the pawn from the side with 22...Rf5 would have led to 23.Qxe6+ Qf7 24.Qc8+ Qf8 25.Qxf8+ Kxf8
26.e4 Re5 27.Rc4 (27.f3 Ke7) 27...Ke7 (it seems that 27...g4!? is the simplest) 28.f4 gxf4 29.gxf4 Rh5 30.Kg2 Kd6
31.Ra4 Kc6 32.Rxa7 g6 and according to Timman Black has reasonable drawing chances;
C) A third option, 22...Qd7!?, has not been considered before. It is slightly passive, but keeping both heavy pieces on
the board makes it very hard for White to use his king. That will become a key factor in the game. After 23.Rxc5 h6
24.Ra5 Rf7 White is better, but it looks really hard to use the king even for a ‘wizard with monarchs’ like Karpov.
Black should hold.
In the event of 23.Qxg5 Black has a choice:
C1) 23...Rc8?! 24.Rxc5 Rxc5 25.Qxc5 Qd1+ 26.Kg2 Qxe2 27.Qxa7 Qe4+ 28.f3 Qe2+ 29.Qf2 Qc4 and it is not an
easy task to convert the extra pawn into a win. It looks no worse for Hort than the game;
C2) 23...c4 24.Qg4 (in case of 24.Rxc4? Qd1+ 25.Kg2 Qxe2 Black has no problems at all) 24...c3 25.Rxc3 Qd1+
26.Kg2 Qd5+ 27.f3 Qxa2 28.Qe4 and this position is very similar to the game;
C3) 23...Rf5!? 24.Qe3 Qd6. This is a pleasant endgame for White. It is easy to believe that Karpov would have been
able to squeeze an advantage and he could have worn down many an opponent in such a position.

23.Rxc5 Qxa2

It is usually a relief to reduce the opponent’s pawns to one side when trying to hold a position. However, Black still has
a problem pawn on e6.

24.Rxg5 Qb1+?!

Better was 24...Rf5! 25.Rg4 (25.Rxf5 Qb1+ 26.Kg2 Qxf5 27.Qxa7 Qe4+=) 25...a5 26.Rd4 Qb1+ 27.Kg2 Qb6 and
according to Timman the a-pawn provides Black with sufficient counterplay.

25.Kg2 Qb6

Again, many annotators have not been interested to state an opinion about the ending with queens on the board after
25...Qb7+. It would have given Black less problems than in the game. After 26.f3 Qd7 27.Ra5 Rf7 he is passive, but as
White’s king will be vulnerable to checks when he pushes the pawns, it may well be tenable.

26.Re5

It is probably better for Black to have the queenside pawn on b6 (after 26.Qxb6!? axb6) than on a7, as the rook has a
spare move (...Rb7) at its disposal and the king has more chances to assist as the pawn is closer: 27.Rb5 Rb8 28.Kf3
Kf7 29.Ke4 Ke7 30.Ke5 Rb7 and Black is passive but he has drawing chances. Black can try to keep moving the rook
to and fro on the b-file. Also the plan of ...Kd7-c6 followed by ...Re7 looks better with the rook on the seventh rank.
If 30...Kd7 31.f4 Kc6 32.Rb2!, the rook defends e2 and now e6 is hanging (32.Rb1 b5! 33.e4 Kd7 is a better version for
Black): 32...Re8 33.Rc2+ Kb7 (33...Kd7 34.Ra2 and the endgame is really hard for Black) 34.Kd6 and White’s king
domination is more important than the passed pawn, which is why White has a clear advantage.

26...Qxe3

Commentators have been so excited about the ensuing endgame that they have neglected to consider the alternative
26...Kf7. White should not settle for the ending with the pawn on b6, as 26...Kf7 has even provided Black with an extra
tempo compared to the above.
So White should keep the queens on the board now with 27.Qd3! as Black’s king is not safe: 27...Kg8 (after 27...Qc7
28.Qf3+ Kg8 29.Qe4 White’s heavy pieces have improved) 28.f3 Rf6 (28...Rd8 29.Qc4) 29.Rb5 Qc7 30.Qe4 and
White has an advantage.

27.Rxe3
27...Re8?!

This may be a losing mistake. If it is, what other possibilities could Black consider?
27...a5 is an interesting idea. Black has chances to create a weakness or obtain enough counterplay while the white king
marches to win the a-pawn: 28.Rxe6 Ra8 29.Re3 a4 30.Ra3 Kf7 and Black has chances to draw.
Many commentators felt that 27...Kf7!? should have led to a draw. It is known that this position is drawn if Black has
the f-pawn. Karpov drew such a position against Kortchnoi at their World Championship match in Baguio in 1978. He
also had a similar position with the extra pawn on his side against Olafsson in Bad Lauterberg in 1977. Even Karpov
was not able to grind down the Icelandic grandmaster.
The commentators have different opinions on how hard it is for Black to attain a draw with a weak pawn on e6. Timman
goes impressively deep in his analysis of this endgame of 4 pawns against 3 in his book The Art of Chess Analysis. He
fully lives up to the high expectations he sets with the book’s title.
28.Ra3 h5!? (hanging on to the a-pawn with 28...Ra8 looks insufficient, e.g. 29.Ra6 Ke7 30.Kf3 Kd7 31.Ke4 Kc7
32.Rxe6 a5
analysis diagram

33.Kd3! (the king is a more efficient piece for blocking the a-pawn. 33.Re7+ Kb6 34.Rxg7 a4 35.Rd7 a3 36.Rd1 a2
37.Ra1 Kc5 is not convincing) 33...a4 34.Kc2 a3 35.Kb1 and Black is in trouble) 29.Rxa7+.
From now on we follow Timman’s superb analysis: 29...Kf6 30.Ra5

analysis diagram

Now Polugaevsky suggested the active move 30...e5! with which Black increases his space, not relinquishing any to
White: 31.f4 (after 31.Ra6+ Kf7 32.Kf3 Rb8 33.h4 Rb2 Black’s rook is active enough on the second rank to hold)
31...exf4 32.gxf4 g6 33.e4 Kg7! (Timman explains that in this situation the king stands much better on the eighth rank
than on the sixth. Who said that endgames are boring? 33...Rb8 34.Ra7!) 34.Kg3 Rb8 35.e5 Rb4 36.Ra7+ Kg8 and
Black holds as the Dutch grandmaster has pointed out.
30...g6 (this leads to a troublesome defence, but some of Timman’s lines are so nice that they are worth a look) 31.f4
(31.h4 Rb8 32.Kf3 Rb2 and Black gets counterplay; 31.Kf3 Rb8 32.Kf4 Rb2! (Black must play actively) 33.Re5

analysis diagram

Now, waiting passively with 33...Ra2? would lead to disaster: 34.h4 Rb2 35.f3 Ra2 36.e4 Rb2 37.Ra5 Rb4 38.g4! hxg4
39.fxg4 Rc4 40.g5+ Kf7 41.Ra7+ Kf8 42.Ke5 wins.
Better is 33...Rb4+! 34.e4 Rb2 and now White misses the pawn on h4 to set up Timman’s winning position) 31...Rb8
32.e4 Rb2+ 33.Kh3 (threatening e4-e5)
analysis diagram

A) First, a look at 33...Rf2!?. This seems to neutralize White’s e4-e5 threat... (not 34.e5+? Kf5 35.Ra8 g5 36.fxg5
(36.Rf8+ Ke4=) 36...Kxg5 37.Rg8+ Kh6 38.Re8 Re2 39.Rxe6+ Kg5 40.Re8 Kg6 41.e6 Kg7=)... but it does so only
temporarily as after 34.Rb5!! Black is in zugzwang: 34...h4 (Black can still play 34...e5) 35.Kxh4 Rxh2+ 36.Kg4 Rg2
37.e5+ Kf7 38.Rb7+ Kf8 39.Rb3 and White wins;
B) Black can save the game with the magical 33...e5!!. My evaluation of the move is different from Timman’s.

analysis diagram

Quite remarkably, the Dutch grandmaster had almost the same position against Meulders in the 1978 Amsterdam zonal
tournament after 43 moves.
B1) 34.Rxe5! In the game with Meulders, Jan did not have this option as his rook was not on a5.
analysis diagram

34...Re2 and quite amazingly we have a case of reciprocal zugzwang here! 35.Re8 (35.g4 hxg4+ 36.Kg3 Kg7 37.Kxg4
Rxh2 leads to a well-known drawish position) 35...Kf7 36.Ra8 Rxe4 37.Kh4 and now:
B11) 37...Rb4? 38.Kg5 Rb5+ 39.Kh6 loses;
B12) 37...Kf6? 38.Ra6+ Kf5 39.h3 Re1 (39...g5+ 40.Kxh5 gxf4 41.g4+ Ke5 42.Kg5+–) 40.g4+ hxg4 41.hxg4+ Kxf4
42.Rf6+ also loses;
B13) 37...Re2! 38.Kg5 (38.h3 Rh2=) 38...Rxh2 39.Ra7+ Ke8 40.Kxg6 Rg2 41.Rh7 Rxg3+ 42.Kf6

analysis diagram
Here the Dutch grandmaster commits a clear mistake in his analysis by only looking at 42...Rg4? 43.f5 h4 44.Rh8+ Kd7
45.Kf7 and Black loses as his pawn only reaches the 4th rank;
42...Rh3! is the only move that draws, for instance 43.f5 h4 44.Rh8+ Kd7 45.Kf7 Rf3 46.f6 h3 and Black holds.
B2) 34.Ra6+

analysis diagram

B21) 34...Kg7? actually results in a lost position: 35.f5 gxf5 36.exf5 Rf2 37.Rg6+! (this check makes the difference)
37...Kh7 (37...Kf7 38.Rg5 e4 39.Rxh5 e3 40.Rh4+–). White can now catch Black’s king with Timman’s remarkable
line 38.Kh4!! Rxh2+ (38...Rxf5 39.Rg5 Rf2 40.h3+–) 39.Kg5 Rh3
analysis diagram

B211) 40.Kf6!. For a long time it seemed as if this variation, proposed by Timman, contained a hole, but it does win. It
would have been sad if it did not, as Timman has presented such compelling analysis and found a great idea. However,
White has to find some difficult moves: 40...e4 41.Rg7+ Kh6! (on 41...Kh8 42.g4 Ra3 (in case of 42...hxg4 43.Kg6
White wins) 43.Re7 wins)

analysis diagram

Now with 42.Kf7 White sets up a nice checkmate with 43.Kg8 and 44.Rg6. Timman stops his analysis here, but the
position merits further attention.
I) The advanced f-pawn allows White another winning attempt, leaving his rook unprotected: 42.Ke7!? and now:
IA) 42...Kxg7? 43.f6+ Kg6 44.f7 Rxg3 45.f8=Q and White wins;
IB) 42...Rh1! still draws but, interestingly, there are fewer options to hold: 43.f6 (43.Rg6+ Kh7 44.g4 hxg4 is a draw)
43...Ra1 (43...e3? 44.f7 e2 45.f8=Q e1=Q+ (if Black’s rook was on h2 it would be a draw) 46.Kd7 Qd2+ 47.Kc8 and
White wins) 44.Rg8 Ra7+ 45.Ke6 Ra6+ 46.Kf5 Ra5+ 47.Kxe4 Ra4+=;
IC) 42...Rh2!! 43.Rg6+ (after 43.f6 e3 44.f7 (in case of 44.g4 hxg4 45.Rxg4 e2 46.Re4 Rf2 Black holds) 44...e2
45.f8=Q e1=Q+ 46.Kd8 Qa5+ 47.Rc7+ Kg5 White cannot win) 43...Kh7 and now:
IC1) 44.g4 hxg4 draws;
IC2) 44.Re6 Rf2 45.Kf6 (45.f6 Kg6=) 45...Rf3 46.Rxe4 Rxg3 and Black obtains the draw;
IC3) 44.Kf7!? (White can still go after the king) 44...Rf2 (not 44...e3?? 45.Rg7+ Kh6
analysis diagram

46.Kg8!! h4 47.Rg6+ Kh5 48.g4 mate) 45.f6 e3 46.Rg7+ Kh6! 47.g4 hxg4 48.Rxg4 e2 and Black finally finds
redemption.
II) Back to the position after 42.Kf7 (instead of 42.Ke7).

analysis diagram

42...h4!. This is the moment to jettison the pawn. It forces White to re-orientate to find the win:
IIA) 43.Kg8 Rxg3 44.f6 Rf3 45.f7 Kh5 46.f8=Q Rxf8+ 47.Kxf8 and Black can draw by pushing either of his pawns;
IIB) 43.gxh4 Rxh4 44.Rg6+ (in case of 44.Rg3 e3 45.Rxe3 Ra4 Black draws; or 44.f6 e3 45.Kg8 e2 and Black is safe)
44...Kh5!. Black draws by moving his king forward;
IIC) 43.g4 Ra3! (Black must rely on his rook checks, so inadequate would be 43...Rg3?? 44.Rg6+ Kh7 45.Rg5, which
produces an attractive checkmate) 44.Rg6+ Kh7 45.Re6 h3 46.Rxe4 Kh6 (or 46...h2 47.Re1 Kh6 48.Rh1 Kg5 49.Rxh2
Ra7+ and White cannot hide from the checks) 47.Re6+ Kg5 48.Rg6+ Kf4 49.f6 Ra7+ 50.Kg8 h2 51.Rh6 Kg5 and
Black escapes;
IID) 43.Rg6+!! (at first, I thought the position was a draw, but during the proof-reading this move suddenly came to my
mind – it wins! Did Timman miss 42...h4, or did he have an answer to it? Whatever the answer, this subtle check makes
a difference as, compared to line IIB above (43.gxh4), the Black king now cannot escape via h5) 43...Kh7 44.gxh4 Rf3
(in case of 44...Rxh4 45.Rg3 the king is unfortunately positioned on h7. Alternatively, 44...Ra3 45.Re6 e3 46.Re8!? (the
white king goes to g5) 46...Ra4 47.Rxe3 wins) 45.f6 Ra3 (45...e3 46.Rg5 Kh6 47.Re5 Rf4 48.Rxe3 Rxh4 49.Kf8 and
White wins) 46.Rg7+ Kh6

analysis diagram

47.Rg8! and White wins because the black e4-pawn shields the white king – without it Black would draw: 47...Ra7+
(47...Kh7 48.Re8 Ra7+ 49.Re7 Ra4 50.Kf8+ Kh8 51.f7 Kh7 52.Re8 Ra7 53.Rxe4+–) 48.Ke6 Ra6+ 49.Kf5 Ra5+
50.Kg4! Ra7 51.h5 Kh7 52.Re8.
B212) 40.g4!!. This is an even more efficient way to hunt down Black’s king than Timman’s 40.Kf6!.
40...hxg4 41.f6! g3 (41...Rf3 42.Rg7+ Kh8 43.Kg6 – remarkably, if Black had no e5-pawn he could save the game. One
of those wonderful paradoxes that illuminate the game of chess!) 42.Rg7+ Kh8 43.Kf5! g2 44.Rxg2 and the f-pawn is
unstoppable: 44...Rf3+ 45.Ke6!. White hurries with the king. This position would be winning even if Black had no e-
pawn (but if now 45.Kxe5? Kh7 Black draws).
B22) 34...Kf7!. This is the move which draws. It transposes to the Timman-Meulders position, in which nine more
moves had been played (hence the altered move numbering below).

After 33...e5!, White had given a few checks and in the diagram position he played 35.fxe5 (35.f5 gxf5 36.exf5 Rf2)
35...Re2? (it is not known to me whether this mistake occurred before or after the adjournment) 36.Ra7+ Kf8 37.Kh4
Rxe4+ 38.Kg5 Rxe5+ 39.Kxg6 Re2 40.Rf7+ Ke8 41.h3 Rg2 (41...Re3 42.g4 Rxh3 43.g5 h4 44.Rf4 and White wins)
42.Rf3 h4 43.g4 Rg3 44.Rf4 Rxh3 45.g5 Ke7 46.Kg7 Ke6 47.g6 Ke5 48.Rf1 Rg3 49.Kf7 and here the Belgian player
resigned.
As Timman has indicated, Black should have gone 35...g5! 36.Rh6 g4+ 37.Kh4 Rxh2+ 38.Kg5 Rg2 39.e6+ Ke7 40.Kf4
Rf2+ 41.Ke5 Ra2 42.Rh7+ Ke8 and White cannot make any progress.
After this wonderful excursion, let’s return to the original game after 27...Re8?!.

28.Ra3 Re7 29.Ra5!

Karpov plays more prosaically, and still it is very powerful chess – he stops ...h7-h5.

29...Kf7 30.h4 h6 31.g4 Kf6 32.f4!

With his last three moves Karpov has gained space and prevented any pawn exchanges on the kingside.

32...Rb7 33.Kf3 Rc7

34.Ra6!?

Karpov forces Hort to play 34...g6. This is subtle indeed.

34...g6

If 34...Kf7, White plays 35.f5 and his king will invade via h5.
After 34...Rb7, 35.e3!? keeps up the pressure. In case of 35.f5 Rb6 Black escapes, as Cvetkovic has pointed out.

35.Ra5 Rd7 36.e3!

This is a typical Karpov preparation move.

36...Rb7
37.h5!

As Black has played ...g7-g6, Karpov now threatens 38.hxg6 and White will soon have two connected passed pawns, as
the Dutch grandmaster explains.

37...g5 38.Ra6 gxf4

In case of 38...Rb3 39.fxg5+ (39.f5 Rb6 40.Rxb6 axb6 41.e4 Ke5 42.Ke3 exf5 43.exf5 Kd5 44.Kd3 b5 45.Kc3 Kc5
46.Kb3 Kd5 47.Kb4 Kc6 48.f6 Kd6 49.Kxb5 Ke6 50.Kc5 Kxf6 51.Kd6 Kf7 52.Ke5 Ke7=) 39...hxg5 40.Rxa7 Rb4
41.Rh7 e5 42.Rh6+ Kf7 43.Rg6 e4+ 44.Ke2 Rb5 45.Kd1 Rc5 46.Kd2 (zugzwang) 46...Rd5+ 47.Kc2 Rd3 48.Rxg5
Rxe3 49.h6 White wins.

39.exf4 Rb3+ 40.Kg2 Rb7 41.Kg3 Kf7

42.Ra4! Kg7 43.g5

Now the point of Karpov’s 42nd move can be seen. There are no checks to cause problems on the fourth rank.

43...Rc7 44.Ra5

This was the sealed move.

44...Kg8 45.Rb5 Kf7 46.Kg4 a6 47.Rb8! Rc1

Another Dutchman, Cor van Wijgerden, has shown how White wins if Black inserts the exchange on g5. His lines are
instructive. Black’s pawns are far from their promotion squares and that gives White time to improve his king
decisively: 47...hxg5 48.fxg5 Rc4+ 49.Kf3 Rc3+ 50.Ke4 Rc4+ 51.Ke3 Rg4 (51...Rh4 52.g6+ Kg7 53.Rb7+ Kg8
54.Rh7 a5 55.Kf3 a4 56.Kg3 Rh1 57.Kg4 a3 58.Ra7 Ra1 59.Kg5 and Black’s king is caught) 52.g6+ Kg7 53.Rb7+
Kg8 54.Kf3 Rg5 55.Rh7 e5 56.Ke4 a5 57.Kd5 a4 58.Ke6 and White wins.
48.g6+!

White’s win becomes clear after this pawn advance, the rest is simple.

48...Kg7 49.Rb7+ Kf8 50.Rb6 Rg1+ 51.Kf3 Rf1+ 52.Ke4 Re1+ 53.Kd4 Ke7 54.Rxa6 Kf6 55.Ra7 e5+ 56.fxe5+ Rxe5
57.Ra6+ 1-0

Hort resigned. This was a grand endgame by the champion.

White does not check with 57.Rf7+? Ke6 58.Re7+ Kxe7 59.Kxe5 as this is a theoretical draw.
After the text, if 57...Re6 (57...Kf5 58.g7) 58.g7, Black has no chance to reach the corner with his king and rook. Even
if he managed this, he would lose after the white queen is brought to e6 and the king to d6. This puts Black into
zugzwang and the white king would approach safely to ensure the win.
Next, a game from the Interpolis tournament in Tilburg, where Karpov drew more games than usual, but still came first
by beating Vasily Smyslov in the final round.

GAME 61
Karpov, Anatoly
Hort, Vlastimil
Tilburg 1979 (9)

1.e4 c6 2.Nc3 d5 3.Nf3 dxe4 4.Nxe4 Nf6 5.Nxf6+ exf6 6.Be2 Bd6 7.0-0 0-0 8.d4 Re8 9.Re1 Bf5 10.Be3 Nd7 11.h3
Be4 12.Nd2 f5 13.Nxe4 fxe4 14.c4 Qc7 15.Bf1 Bh2+ 16.Kh1 Bf4 17.Qd2 Bxe3 18.Qxe3 Nf6 19.Rad1 Rad8 20.Rd2
h6 21.g3 Qa5 22.b3 Rd7 23.Bg2 Red8 24.Red1 Qg5 25.Kg1 a6 26.Re2 Kf8 27.Qc3 Qf5 28.a4 Kg8 29.Qe3 Re8
30.Kf1 Qa5 31.Qd2 Qxd2 32.Rexd2 e3 33.Re2 exf2 34.Rxe8+ Nxe8 35.Kxf2 a5 36.Ke3 Kf8 37.Rb1 Ke7 38.g4 Kd8
39.b4 Re7+ 40.Kd3 axb4 41.Rxb4 Kc7 42.Rb1 Nf6 43.a5 Nd7 44.Ra1 Nb8 45.h4 Na6 46.Rb1 Nb8 47.Bf3 Nd7
48.Ra1 Re8 49.Rf1 Re7 50.Bg2 Nb8 51.Rf4 Nd7 52.Rf1 Nb8 53.Be4 Na6 54.Rb1 Nb8
55.Bf5!?

The same opponents had reached a position with these basic contours some time earlier, but then Karpov had not been
able to make any progress. Now he comes up with a new try.

55...Nd7

After 55...Na6 56.h5 (56.Rf1?! allows a draw after 56...g6 57.Be4 Nb4+ 58.Kc3 Na2+) 56...Nb8 57.Rg1 White
maintains some pressure.

56.Ra1 g6?

Black finally loses patience.

He should have waited passively with 56...Nb8 to see how White would try to break open the position. His chances of
holding remain good. It would have been interesting to see how Karpov was going to put pressure on Black’s position
by playing 57.g5 or 57.h5 with Rg1.
Also possible was 56...Nf6 57.Rf1 and it seems that Black has better drawing chances than White has winning ones.
57.Bxd7!

White exchanges in order to clear the way for the rook to invade. Surprisingly, this move is powerful enough to bring
Black down.

57...Rxd7

Black hopes to get counterplay on the queenside.

In case of 57...Kxd7, helping with the king on the kingside would have been very tough as well: 58.Rf1 and now:
A) 58...Ke8 59.d5
A1) 59...cxd5 60.cxd5 Rc7 61.Kd4 Rc2 62.Rb1! Ra2 63.Rb5! Kd8 64.Kc5
A11) 64...Kc7 65.d6+ Kc8 66.Kb6 Rd2
analysis diagram

67.Ka7 – White would have obtained extremely good winning chances with this subtle king manoeuvre;
A12) 64...Kc8 65.Kd6 f5 66.gxf5 gxf5 67.h5 and White’s advantage is so convincing, it is almost certainly winning.
A2) 59...Rd7 60.Kd4 cxd5 61.cxd5 Rd6 62.Rb1 Ra6 63.Rb5 Kd8 64.Ke5 – White’s advantage is clear, yet it requires
some effort to convert it into a win.
B) 58...Ke6 59.Re1+ and White wins the pawn ending: 59...Kd7 60.Rxe7+ Kxe7 61.Ke4 Ke6 (61...f6 loses to 62.d5
Kd6 63.h5) 62.d5+ cxd5+ (62...Kd6 63.g5) 63.cxd5+ Ke7 64.Ke5 f6+ 65.Ke4 Kd6

analysis diagram
And now 66.h5! wins just like in the famous game Botvinnik-Flohr, Moscow 1944.

58.Rf1! Kb8?

Hort looks for activity for his king on the queenside, but his play there is simply too slow compared to Karpov’s on the
kingside.
Better was 58...Kd8. Even though staying in the centre was very unpleasant, this should have been tried: 59.g5 hxg5
60.hxg5 Ke8

analysis diagram

Now if 61.Rf6 Ke7 62.a6 bxa6 63.Rxc6, Black struggles.

Planning to defend the d4-pawn with the rook and to march to b6 with 61.Kc3 can lead to an advantageous queen
ending: 61...Ke7 62.Rf4 Rd6 63.Kb4 f6 64.gxf6+ Rxf6 65.Rxf6 Kxf6 66.Kc5 g5 67.Kb6 g4 68.Kxb7 g3 69.a6 g2
70.a7 g1=Q 71.a8=Q and White wins.

59.Rf6!

Making Black aware of the weakness of the h6-pawn.

59...Ka7 60.h5! Ka6

In the event of 60...gxh5 61.gxh5 Ka6 62.Rxh6 Kxa5 63.Rh8 b5 (or 63...Kb6 64.h6 f5 65.h7 f4 66.Ke4 Rf7 67.Kf3)
64.h6 Kb4 65.c5 White has a decisive advantage.
61.g5!

Pushing the pedals hard. This breakthrough puts an end to Blacks resistance.

61...hxg5 62.h6 Kxa5 63.h7 Rd8 64.Rxf7 b5 65.cxb5 Kxb5 66.Rb7+! Ka6 67.Rg7 Rh8 68.Ke4

White just marches towards the g-pawns and wins.

68...Kb5 69.Kf3 Kc4 70.Rd7 Kd3 71.Kg4 Rxh7 72.Rxh7 Kxd4 73.Rd7+ 1-0

GAME 62
Karpov, Anatoly
Unzicker, Wolfgang
Bad Kissingen 1980

1.c4 e6 2.Nc3 d5 3.d4 Nf6 4.Nf3 Be7 5.Bg5 0-0 6.e3 h6 7.Bh4 b6 8.Qb3 dxc4 9.Bxc4 Bb7 10.0-0 Nbd7 11.Rfd1
Ne4 12.Bg3 Nxc3 13.bxc3 Bd6 14.a4 Bxg3 15.hxg3 Qf6 16.Nd2 c5 17.a5 Qe7 18.Bb5 Rfc8 19.axb6 Nxb6 20.Ba6
Rc7 21.Bxb7 Rxb7 22.Ne4 cxd4 23.cxd4 Nd5 24.Qd3 a5 25.Nc5 Rba7 26.Qb3 Nf6 27.Qb5 Qd8 28.Rdb1 Nd7
29.Qa4

29...Nb6?

When a player is worse and has the choice between playing a rook or a knight ending, most of the time he or she should
aim for the former. This game is no exception.
Better was 29...Nxc5! reducing the pressure on the queenside: 30.dxc5 Rc8! (planning to eradicate White’s queenside
pawn) and now:
A) 31.Rc1 Rac7 32.Qb5 (after 32.Qxa5 Qf8 White cannot hang on to the c-pawn):
A1) With 32...Qd5 Black can probably hold the position by directing all his firepower to c5: 33.Rxa5 Qd2 34.Raa1
Qd5 35.Qa6 Kh7 36.Rc2 Rc6 and White is probably unable to do more than hang on to the pawn;
A2) 32...a4! – this delightful pawn move makes sure Black is not losing, e.g. 33.c6 Qd6 34.Qa6 Kh7 and Black is safe,
or 33.Rxa4 Rxc5! and the position is a dead draw.
B) 31.Rb5 Qd2! 32.Rxa5 Rxa5 33.Qxa5 Qc2 and the c5-pawn drops.

30.Qc6

Karpov keeps up the pressure on the queenside.

30...Nd7?

Black is impatient. He wants to end the game too quickly. He will not be successful with this aim...
30...Nd5! was clearly better.

31.Nb7!

Karpov doesn’t miss such tactical resources in the ending.

31...Qe7

The alternative 31...Qe8 defends the a8-rook, yet it doesn’t solve Black’s problems: 32.Nd6! (32.Nxa5? would not
work now: 32...Rxa5 33.Rxa5 Rxa5 34.Rb7 Nf6 and Black wins) 32...Qd8 33.Rb7! (this subtle move gains material)
A) 33...Nf6 34.Nxf7 Qd5 35.Qxd5 exd5 36.Rxa7 Rxa7 37.Ne5 should do;
B) If 33...a4 34.Rxa7 Rxa7 35.Nb7 Qa8 36.Qxd7 Rxb7 37.Rxa4 Qb8 38.Qc6 – practice shows that the stronger side
has better chances with 5 pawns against 4 rather than 4 pawns against 3. This position is winning for White;
C) 33...Nb8 34.Qc5 Rxb7 35.Nxb7 Qd7 36.Nxa5 Nc6 37.Nb3. It is almost impossible to prove, but White should be
winning here as well.

32.Nxa5
Karpov chops off a pawn, which spells Black’s doom as after 32...Rxa5, 33.Qxa8+ Rxa8 34.Rxa8+ wins back the
knight.

32...e5 33.Nb3 Rxa1 34.Nxa1 Ra2

It was bad luck for Black that he had no time to isolate the white d-pawn, as his rook was hanging. Of course, even if he
had had the time, his defensive task would have been daunting.

35.Nc2

Karpov’s pieces are slightly disorganized, so he improves the knight and then creates a passed pawn.

35...exd4

36.Qc8+!

This is very typical of Karpov. First he prevents any activity from his opponent, then he initiates the execution.

36...Kh7 37.Qc4!

Pushing back Black’s only active piece.

37...Ra8

Or 37...Qe6 38.Qxe6 fxe6 39.Nxd4. Unzicker probably knew the Karpov-Hort game from the previous year. One can
appreciate he did not want to play with the same pawn structure.

38.Qd3+!

He even takes the opportunity to push back the king.

38...Kg8
Avoiding 38...g6, which would weaken Black’s structure.

39.Nxd4

The position is known to be winning for the stronger side when the heavy pieces have disappeared. It should be the
same with them on the board.

39...Nf6 40.Rc1 Qe4 41.Qe2 Rd8 42.Qf3 Qe5 43.Qf4 Qh5 44.f3 Nd5 45.Qe4 Nf6 46.Qf5 Qxf5 47.Nxf5

After exchanging the queens White can push his pawns and his rook is stronger than Black’s.

47...Rd3

Maybe 47...h5!? would have put up a better fight.

48.g4!

White starts gaining space with his pawns.

48...Nd7 49.Kf2

In 1980 Unzicker was already an ageing veteran. It was taxing to defend such a long game. Even if he had been
younger, though, his position would still be fundamentally lost.

49...Ne5 50.Rc5!

Pushing back the slightly active knight.

50...Ng6 51.Rc8+ Kh7 52.f4 Rd5 53.g3 Ra5 54.Rc7 Kg8 55.Nd6 Nh8

White can be more relaxed with the black knight forced into passivity.
56.e4 Ra2+ 57.Kf3 Ra3+ 58.Kg2 f6

Black’s position would deteriorate further if White were allowed to play e4-e5.

59.Re7 Rd3 60.Re8+ Kh7 61.Nf5 Nf7 62.Re6 Kg8 63.Kf2 Kf8

64.Ra6

Karpov pushes back the rook and he can start improving his king as well. It takes a few moves, but the conclusion
cannot be altered.

64...Rd8 65.Ke3 Rb8 66.Ra7 Rd8 67.Nd4 Re8 68.Ra6 Rb8 69.Kd3 Rd8 70.Kc3 Rc8+ 71.Kb4 Re8 72.Ne6+ Kg8
73.Kc4!

This is a clear sign that Karpov is not interested in swapping knights. If 73.Kc5 Nd8.

73...Nd8 74.Nc5! Nf7

75.Kd5!

Karpov has improved his king beautifully.

75...Rc8

Or 75...Rd8+ 76.Ke6.

76.Nd7 Nd8 77.Nc5 Nf7 78.Rc6 Ra8

Alternatively, 78...Rd8+ 79.Ke6 Ra8 80.Rc7 loses as well.

79.Rc7 Kf8 80.Ne6+ Kg8

81.e5!

White has optimized the potential of his pieces. Now it is time to create the passed pawn.

81...fxe5 82.fxe5 Re8 83.Rd7 Ra8 84.Nc5 Re8 85.e6 Ng5 86.Kd6 Kf8

After 86...Nh7 87.Ne4 Nf8 88.e7 Ng6 89.Nc3 wins.

87.Ra7 Rd8+ 88.Nd7+ Ke8 89.Ra4!

Preventing a check and seeking to invade on the f-file.

89...Nh7

After 89...Nxe6 90.Kxe6 Rxd7 91.Ra8+ the pawn ending is hopeless for Black.
90.Ke5!

Karpov uses his king in majestic style. He changes the angle of his attack and invades on the weakened kingside.

90...Nf8 91.Kf5 Nxd7

At this point, Unzicker may well have reflected on the missed opportunity to exchange knights on move 29. The
circumstances have changed considerably.

92.exd7+ Rxd7

After 92...Kf7 93.Rd4 g6+ 94.Ke5 Ke7 95.Rd1 Black loses because of zugzwang.

93.Kg6!

The exclamation mark is not for subtlety, but for his correct assessment of the ending.

93...Kf8

After 93...Rd3 94.Kxg7 Rxg3 95.Re4+ Kd7 96.Kxh6 Black is hopelessly cut off.

94.Ra8+ Ke7 95.Kxg7 Rd4 96.Ra7+ Ke8

If 96...Ke6 97.Ra6+ (not 97.Kxh6?? Rxg4 98.Rg7 Rxg7 99.Kxg7 Kf5 and White loses his pawn) 97...Ke7 98.Kxh6
Rxg4, 99.Rg6 wins after 99...Ra4 100.g4. 99.Ra3 Ke6 100.Kh5 Rg8 would be a well-known draw.

97.Kxh6

White’s g-pawn reaches the fifth rank and the frontal attack no longer works with the black king cut off. There is no
escape.

97...Rxg4 98.Rg7 Ra4 99.g4 Kf8 100.g5 Rg4

If his king were on the rook file here, Black would still have had drawing chances.

101.Ra7 Rg1 102.Kg6 Rg2 103.Ra8+ Ke7 104.Rg8 1-0

 GAME 63
Karpov, Anatoly
Ivkov, Borislav
Bugojno 1980 (5)

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.h3 Nd7 10.d4 Nb6 11.Nbd2 exd4
12.cxd4 Nb4 13.Nf1 c5 14.a3 Nc6 15.Be3 Na5 16.Bc2 Nbc4 17.Bc1 cxd4 18.Nxd4 Bf6 19.Rb1 d5 20.exd5 Bb7
21.Nf5 Qxd5 22.Qxd5 Bxd5 23.b3 Nb6 24.Be3 Rab8 25.N1g3 Be6 26.Ne4 Nd5 27.Nxf6+ Nxf6 28.Bc5 Rfe8
29.Nd6 Red8 30.f4 Nb7 31.Nxb7 Rxb7 32.f5 Bd5 33.Rbd1 Rbd7 34.Rd2 h6 35.Kh2 Rc8 36.b4 a5

37.g4!

Gaining space.

37...axb4 38.axb4 Bc4 39.Red1 Rcd8?

Maybe the Yugoslav grandmaster was short of time. 39...Rxd2+ would have been better. With the rooks on the board
White would not be able to move his minor pieces as freely as in the game. White’s advantage is smaller in this line.
With 40.Rxd2 Ra8 Black saves one rook.

40.Rxd7! Rxd7 41.Rxd7 Nxd7

42.Bd6

Now nothing disturbs the domination of the two bishops.

42...Nb6 43.Kg3 Bd5

According to Ivkov 43...Bf1! was better, although it doesn’t change matters much.

44.Kf4 Nc4 45.Bc5 Bg2 46.Bd3?!

The Yugoslav grandmaster suggests 46.h4! as winning: 46...f6 47.Be4 Bxe4 48.Kxe4 h5! (waiting passively with
48...Ne5 does not help Black: 49.Kd5 Kf7 50.h5! Ke8 51.Kd6! Kd8 (51...Kf7 52.Kc7; 51...Nxg4 52.Kc6 Ne5+
53.Kxb5 Kd7 54.Bf8) 52.Bb6+ Kc8 53.Ke6 and White wins) and now:
A) If 49.g5 fxg5 50.hxg5
analysis diagram

50...g6!! is a surprising blow:

A1) 51.fxg6 h4 52.Bd4 h3 53.Kf3 Nd2+ 54.Kg3 h2 55.Kg2 Nf3 56.Bf6 Kf8 57.g7+ Kf7 58.g6+ Kg8 59.Bc3 (59.Kh1
Ne5) 59...Ne5 60.Kxh2 Nxg6 61.Kg3 Ne7 62.Kg4 Nd5 draws;
A2) 51.f6 h4 52.Kf4 Kf7 53.Kg4 Ke6 54.Kxh4 Ne5 – the knight goes to f7. Black holds the draw by creating a
fortress, as Ivkov has pointed out.
B) 49.gxh5!

analysis diagram
49...Kh7 50.Be3 (50.Kd5 Kh6 51.Kc6 Kxh5 52.Kxb5 Ne5 53.Bd6 or 53.Ka6 Nd3=) and now:
B1) With some nice variations Ivkov demonstrates what is wrong with 50...Nd6+. After 51.Kd5 Nxf5 52.Bf2 Kh6
53.Kc5 (with the surprising 53.Ke6!! White can hunt down the knight; Ivkov misses this unusual switch, for example:
53...g6 (53...Nxh4 loses to 54.Bxh4 Kxh5 55.Bxf6) 54.hxg6 Kxg6 55.Kd5 Kh5 56.Kc6 Nxh4 57.Kxb5 Ng6 58.Kc5
Ne5 59.b5 Kg4 60.Bd4 and White wins) 53...Kxh5 54.Kxb5 Kg6 55.Kc6 Kf7 56.b5 Ke6 (56...Ne7+ 57.Kd7 Nd5
58.Kd6 Ne7 59.h5 Ke8 60.Bc5 and White wins) and now:
B11) 57.Kc7 Ne7 58.Bc5 Nd5+ 59.Kc6 f5 and Black survives;
B12) 57.Bc5 Nxh4 58.b6 Nf3!! (Ivkov missed this great saving move; after 58...Ng6 59.Bd6 White promotes the b-
pawn) 59.Kc7 Ne5 60.b7 Nd7=;
B13) 57.b6!. This is the right way to play: 57...Ne7+ 58.Kc5 (only move) 58...Kd7 59.h5! (59.Kb5 Nc8! and Black
can destroy the b-pawn with the knight) 59...Nc6 60.Bg3 Nd8 61.Kd5 and remarkably, White now invades on the
kingside: 61...Nb7 62.Be1 Nd8 63.Bb4 and White wins.
B2) 50...Nxe3!? was the best practical chance: 51.Kxe3 Kh6 52.Kd4 Kxh5 53.Kc5 Kg4 54.Kxb5 Kxf5 and now:
B21) 55.Kc6 g5;
B22) 55.h5? Ivkov gives an exclamation mark to this move, but actually this is a mistake that turns out to squander the
win.

analysis diagram

He missed the saving move 55...Ke4!! (Ivkov finishes after 55...Ke6 56.Kc5, correctly evaluating the position as
losing) 56.Kc5 f5 57.b5 f4 58.b6 f3 59.b7 f2 60.b8=Q f1=Q 61.Qe8+ Kf3 (with the king on e6 Black would lose his
queen here) 62.Qf7+ – both the pawn and the queen endings are drawn;
B23) 55.Kc5! g5 (55...Ke6 56.h5) 56.h5 g4
analysis diagram

57.Kd4!! Kf4 58.Kd3! (this way White drives the black king to a losing square) 58...Kf3 59.h6 g3 60.h7 g2 61.h8=Q
g1=Q 62.Qxf6+ and White wins.
Back to the game.

46...f6

If 46...Bxh3, 47.Kg3 traps the bishop.

47.h4 Bc6
Ivkov provided the following evaluation of 47...Kh7!?: Black might survive in analysis, but in an over-the-board game
he has little chance. But there is a surprisingly active idea hidden behind this move:
A) 48.Be4 Bxe4 (this time the exchange works for Black) 49.Kxe4 h5 50.g5 fxg5 51.hxg5 g6 52.f6 Kg8 53.Kd5 Kf7
(if 53...h4?? 54.Ke6 wins) 54.Bf2 (54.Kc6 h4) 54...Nd2 55.Kc6 Nf3 56.Kxb5 h4! and now:
A1) 57.Bxh4 Nxh4 58.Kc5 (58.Kc6 Nf3 59.b5 Nxg5 60.b6 Ne6=) 58...Nf3 59.b5 Nxg5 60.b6 Ne6+ 61.Kd6 Nd8
62.Kc7=;
A2) 57.Kc6 h3 58.Bg3 Nxg5 59.b5 Ne6 60.b6 Kxf6 61.Bd6 g5 62.Kd7 Nc5+ and Black holds;
B) 48.h5 Kg8
B1) 49.Be4 Bf1 (49...Bxe4 50.Kxe4 Ne5 loses to 51.Kd5 Nxg4 52.Kc6) 50.Bd5+ Kh7 51.Bd4 Be2 52.Bc6 Ne5?
According to Ivkov, Black holds with the help of a witty pawn sacrifice. Still, worse positions usually lose in same-
coloured bishop endings. Black can fight to survive with the move 52...Nd6. After 52...Ne5? there follows 53.Bxe5
fxe5+ 54.Kxe5 Kg8 55.Bd5+ Kf8 56.Kf4 Ke7 57.Bf3 Bc4 58.Ke3!! (this move was probably missed in the post
mortem) 58...Kd6 (58...Bf7 59.Be2 Be8 60.Kd4 Kd6 61.Bd3 Bd7 62.g5! Ke7 63.f6+ Kf8 64.fxg7+ Kxg7 65.gxh6+
Kxh6 66.Kc5 Kxh5 67.Bxb5 and White wins) 59.Be2! Bxe2 (59...Kd5 60.Bxc4+ Kxc4 61.g5) 60.Kxe2 Ke5 61.Ke3
Kd5 62.g5 and White wins.
B2) 49.Bd4
B21) 49...Kf7 50.Be4 (50.Bxc4+? bxc4 51.b5 Ke8 52.Bc3 Kd7 53.Bb4 Bf1! (53...Bh3 54.Kf3 wins) 54.Bf8 (54.Ke3
c3) 54...c3 and Black escapes) 50...Bf1 51.Bd5+ Ke7 52.Bc5+ Nd6 and Black is living very dangerously;
B22) 49...Kh7 – this is Ivkov’s recommendation, but moving away from the centre seems awkward. Black manipulates
his king as though he were playing a queen ending – he just puts it aside in a convenient shelter. It was not possible to
crack Ivkov’s move, yet it remains suspect:
B221) 50.Bxc4?! bxc4

analysis diagram

51.b5. The opposite-coloured bishop ending looks lost. However, Black can save himself despite the presence of the
dangerous b6-pawn and despite the fact that all his kingside pawns are fixed on the colour of White’s bishop. One
needs to look deeply when it comes to assessing these endings: 51...Kg8 52.b6 Kf7 53.Ke3 (in case of 53.Bc3 Ke8
54.Bb4 Bd5 55.Ke3 Bg2 56.Kd4 Bf3 (or 56...Kd7 57.Bf8 (57.Kxc4 Kc6) 57...Bf3 58.Bxg7 Bxg4 59.Bxh6 Bxh5 and
Black also takes enough material) 57.Kxc4 Kd7 58.Kb5 (58.Bf8 Kc6) 58...Kc8 59.Ka6 Kb8! 60.Bd6+ Ka8 61.Bf8
Be2+ 62.Ka5 Kb7 Black is safe) 53...Ke8 54.Bc5 Kd7 55.Bf8 c3 (55...Bh3? 56.Kf3 c3 57.Bxg7 and it’s all over)
56.Bxg7 (56.Kd3 Bf3=) 56...c2 57.Kd2 Bf3 58.Bxf6 (58.g5 fxg5 59.f6 Bxh5=) 58...Bxg4 59.Bg7 Bxf5 60.Bxh6 Bg4=;
B222) 50.Be4 Bf1 51.Bc6 Ne5 52.Bxe5 fxe5+ 53.Kxe5 Be2 and according to Ivkov, Black can hold.

48.h5!!

Karpov wants to grind down his opponent. He does not simply wait for a mistake but carries out his own plan. What
then is the point of this move? The disadvantages are clear. It puts one more pawn on the colour of the black bishop and
loses flexibility by giving up the possibility of g4-g5. However, as the game continues the positive elements of this very
deep move are revealed.
In case of 48.Be4 Bxe4 (48...Bd7 49.Bd5+ is an even better version of the game) 49.Kxe4 White wins after 49...Kf7
50.Kd5 g6 51.Kc6 Na3 52.h5 or 49...Ne5 50.Kd5 Kf7 51.h5 Ke8 52.Kd6, but Black can try 49...h5!?. Karpov likes to
deny his opponent any chances, but this probably wouldn’t solve Black’s problems as the line transposes to the analysis
of 46.h4, which wins for White after he has overcome some obstacles.
But why play this move with the h-pawn? For one, Karpov’s text move stops all black pawn moves for once and for all.

48...Kf7 49.Be4 Bd7

After 49...Bxe4 50.Kxe4 Ne5 51.Kd5 Nxg4 52.Kc6 Black has no passed pawn, and White’s b-pawn will decide.

50.Bd4!

Karpov now prevents ...Ne5 as well.

50...Ke7
51.Ba8!!

This is a very pleasing move aesthetically. The bishop goes to the only square on the diagonal that leaves a path open
along which the white king can penetrate.

51...Be8 52.Ke4 Bf7

53.Ba7!?

The other bishop follows suit. This turns out to be good enough to create winning chances, yet it is not the best
continuation. It provides Black with the chance to survive by finding some difficult moves.
Best was 53.Bg1!!. The bishop should move to the other corner. This guarantees the win: 53...Kd7 54.Kd4 Ne5 55.Bd5
Be8 (after 55...Bxd5 56.Kxd5 Nxg4 57.Kc5 Ne5 58.Kxb5 White wins) 56.Kc5 Ke7 (56...Nxg4 57.Bc6+ Ke7
58.Bxe8 Kxe8 59.Kxb5 and White wins) 57.Bd4 Nxg4 58.Bc6 Bxh5 59.Bf3 Kd8 (59...Kf7 60.Kxb5 g6 61.fxg6+
Kxg6 62.Kc5 and the b-pawn looks unstoppable).

analysis diagram

Now 60.Kxb5 Be8+ 61.Ka6 Ne5 is just a draw, but after 60.Kb6!, avoiding possible checks and preparing to play Bc5-
f8 Black is in trouble, for example 60...Kc8 61.Ka6! Kc7 62.Bc5 Kd8 63.Bf8.

53...Kd7 54.Kd4 Ne5 55.Bd5!

Another preventive move, this time guarding against ...Bc4.

55...Nc6+??

In olden days the second time control came at move 56. The Yugoslav grandmaster, a subtle player, would have
benefited from that option had it existed. He makes a bad mistake.
Ivkov shows with some impressive analysis what is wrong with 55...Bxd5. There would follow 56.Kxd5 Nxg4 57.Bg1
Ne5 58.Kc5 Nf3 (58...Nc4 59.Kxb5 Nd6+ 60.Ka6 Nxf5 61.Bc5) 59.Bf2 Nd2 60.Kxb5 Ne4 61.Be1 Nd6+ 62.Ka6!
(62.Kc5? Nxf5 63.b5 Kc7=) 62...Kc6 (on 62...Nxf5 63.b5 Nd6 64.b6 Kc6, 65.Bb4 with the idea Bf8 wins) 63.Bf2
Nb5 64.Bb6 Nd6 65.Bc5 Nb5 (65...Nxf5 66.Bf8) 66.Bf8 Nc7+ 67.Ka5 Nd5 68.Bxg7 Nf4 69.Bxh6 (or 69.Bxf6 Nxh5
70.b5+ Kb7 71.Be5!+–) 69...Nxh5 70.Ka6 Kc7 71.b5 and White wins.
It would be interesting to know how Karpov would have reacted to 55...Be8!. This was a resilient option, leaving White
no clear way to win.
analysis diagram

A) 56.Bb8 Nxg4 57.Bf3 and now:

A1) 57...Bxh5? 58.Kc5 Kc8 59.Bd6 (59.Bg3 Nf2!!) 59...Kd8! (in case of 59...Kd7 60.Bf8 (60.Be2 Nf2) 60...Ke8
61.Bxg7 Kf7 62.Bxh6 Nxh6 63.Bxh5+ Ke7 64.Bg6 Ng4 65.Kxb5 White wins) 60.Kb6! (White first removes his king
from the knight checks before the bishop gets to f8. 60.Kxb5 Be8+ 61.Ka6 Ne3; or 60.Bf8 Ne5!. Here is the point of
54...Kd8: the d7-square is vacant. 61.Bxh5 Nd7+ 62.Kxb5 Nxf8 63.Kc6 Kc8!= rather than 63...Nd7 64.Kb7!?)
60...Kc8 (this keeps the e8-square available for the bishop. After 60...Kd7 61.Bf8 Ke8 62.Bxg7 Kf7 63.Bxh6 Nxh6
64.Bxh5+ Ke7 65.Bg6 or 60...Ke8 61.Kxb5 Kf7 62.Kc5 g6 63.b5, White wins) 61.Ka6! Kd7 62.Bf8 Kc7 63.Bxg7
Kd6 64.Kxb5 Ke5 65.Kc5 Kxf5 66.b5 Kg6 67.b6 Ne5 68.Bxh5+ Kxh5 69.b7 and White wins;
A2) 57...Nf2 58.Bg3 Nh3 59.Kc5 Kd8 60.Bd6 Ng5 61.Bd5 Bxh5 62.Bf8 g6 and Black gets away with it;
A3) 57...Ne5!? is probably the simplest: 58.Bxe5 fxe5+ 59.Kxe5 Ke7=;
B) 56.Kc5 Kc7 57.Bb6+ Kb8 58.Kd6 Nxg4 59.Ke7 (59.Bc6 Bxh5 60.Bxb5 Ne5=) 59...Bxh5 60.Bf3 Ne3 61.Bxh5
(61.Bxe3 Bxf3 62.Kf7 Be4 63.Kg6 h5=) 61...Nd5+ 62.Kf7 Nxb6 63.Kxg7 Nd5 64.Bf3 Nxb4 65.Kxf6 Kc7 66.Be4
(66.Ke7 Nd3 67.Be2 Nf4=) 66...Kd6 67.Kg7 (67.Kg6 Na6 68.f6 Nc7 69.Kxh6= or 69.Bf5 Ne8 70.f7 Ke7=) 67...Na6
68.f6 Nc7 69.Kxh6 (69.f7 Ne6+ 70.Kxh6 Ke7=) 69...Ne8 70.f7 Ke7=;
C) 56.Bc5! (this shows another point of 53.h5: Black pawns are fixed on the kingside. Likewise there is little scope for
the g-pawn) 56...Nxg4 57.Bf8:
C1) 57...Bxh5
analysis diagram

58.Bf3! (this is such an attractive pin) 58...Ke8 59.Bd6 (59.Bxg7 Kf7 60.Bxh6 Nxh6 61.Bxh5+ Ke7 62.Ke4 Kd6
63.Be8 Ng4 64.Bxb5 Ne5 gives Black decent drawing chances) 59...Kf7 60.Kc5 g6 (60...Kg8 loses to 61.Kxb5 Be8+
62.Kc5 Ne3 63.Be4 Bd7 64.b5 Nxf5 65.b6) 61.Kxb5 (61.Bd5+ Kg7 62.Kxb5 Ne3 63.Bc6 Nxf5 64.Bc5 Be2+=)
61...Ne3 62.fxg6+ Kxg6 (in case of 62...Bxg6 63.Kc5 Bd3 64.b5 Bxb5 65.Kxb5 White can chop off one of the pawns,
with a likely win) 63.Bxh5+ Kxh5 64.Kc5 Nd1 65.Kd4 Nb2 66.b5 Na4 67.Bc5 f5 68.Kc4 Nxc5 (68...f4 69.Bd4)
69.Kxc5 f4 70.Kd4 Kg4 71.b6 f3 72.Ke3 Kg3 73.b7 and White wins;
C2) 57...Ne5!? (maybe this is the best Black can hope for) 58.Bxg7 Ke7 59.Bxh6 Bxh5 60.Kc5 Be2

analysis diagram
61.Be6 – Black is clearly worse, but he has chances to draw.

56.Bxc6+! Kxc6

57.Bc5

Fixing the pawns now pays off. Black’s pawns are going to fall.

57...Bc4

How about White’s pawns? They will remain unharmed:

58.Ke4 1-0

A reborn Tal was the sensation of 1979 and 1980. The next game was their only decided one in their entire career.

GAME 64
Karpov, Anatoly
Tal, Mikhail
Bugojno 1980 (8)

1.c4 e6 2.Nc3 d5 3.d4 c6 4.e3 Nf6 5.Nf3 Nbd7 6.Bd3 dxc4 7.Bxc4 b5 8.Bd3 a6 9.e4 c5 10.d5 c4 11.dxe6 cxd3
12.exd7+ Qxd7 13.0-0 Bb7 14.Re1 Bb4 15.Ne5 Qe6 16.Nxd3 Bxc3 17.Nf4 Qd7 18.bxc3
18...Nxe4

18...Qxd1, the other way to restore the material balance, was less attractive: 19.Rxd1 Nxe4 20.Ba3 Rd8 21.Bb4!? (thus,
White can keep up the tension around the enemy king. With 21.Re1 White could have aimed for a rook + opposite-
coloured bishops ending with an extra pawn, as in the note to 20...Rhe8 further on) 21...Rd7 22.Nd3! (22.Re1 Kd8!) and
Black is under heavy pressure as f2-f3 and Nc5 will come.

19.Qxd7+!

Karpov centralizes Tal’s king. But seriously; he wants to drive the king out into the danger zone. Alternatively 19.Ba3
0-0-0 20.Qh5 (20.Qxd7+ Rxd7) 20...g6 would be unclear.
19...Kxd7 20.Ba3 Rhe8

It would have been more natural to clear a passage for the king to safety via c8 with 20...Rad8. Unfortunately, this loses
a pawn: 21.f3! Nd6 (21...g5? 22.Nd3! would give White a great advantage, whereas after 22.Nh3?! g4 23.fxg4 Rhg8
Black would get dangerous counterplay) 22.Rad1 Kc6

analysis diagram

23.Ne6! (this neat tactical shot nets a pawn) 23...fxe6 24.Rxe6 Rhe8 25.Rdxd6+ Rxd6 26.Rxd6+ Kc7 27.Kf2. Karpov
has won many endgames with less advantage than this.

21.Red1+!

Why not bring out the other rook, which has not been developed yet, you may ask. White wants to avoid the exchange
of rooks in order to keep the heavy artillery to attack Black’s king. On the e-file Black could possibly force exchanges.

21...Kc7 22.f3 Nf6

Interestingly, no commentator mentions 22...Ng5!?. The point would be that Black wants to exchange an attacking
piece, i.e. the white knight. Opposite-coloured bishops can increase the domination of the attacker, but this time there
are no queens and it looks as though Black’s king can survive:
A) 23.Bd6+ Kb6 24.a4
analysis diagram

24...Rad8! – the rook comes into play before Black plays 24...Ne6, as then after 25.a5+ Kc6 26.c4! the black king
would be in danger: 25.c4 bxc4 26.Rab1+ Ka7 and Black is no longer in real danger;
B) 23.c4 and now:
B1) 23...Rad8 24.cxb5 Rxd1+ 25.Rxd1 axb5 26.Rc1+ Bc6 27.Nd5+ Kb7 28.Ne7 Bd7 (or 28...Ra8 29.Bb4 Ra6 and
White’s advantage is about to evaporate) 29.Rd1 Bc6 30.Nxc6 Kxc6 31.Rd6+ Kc7 32.Rd5 Ra8 33.Bb4 Ne6 and Black
escapes;
B2) 23...Ne6 24.cxb5 axb5 (with 24...Nxf4 25.Bd6+ Kb6 26.Bxf4 axb5 27.Rd7 Re2 Black could create counterplay)
25.Rac1+
B21) 25...Kb8 26.Bd6+ Ka7 27.Nxe6 Rxe6 28.Rc7 Kb6 29.Rxf7 Rxa2 30.Rxg7 Rg6 31.Rxg6 hxg6 gives enough
counterplay for the pawn;
B22) 25...Bc6 26.Rd6 Rxa3 (26...Nd8 keeps the position together as well) 27.Rdxc6+ Kb7 and Black is just a fraction
worse.

23.Bd6+ Kb6

If 23...Kc6!? 24.c4 (after 24.a4 Rad8 25.axb5+ axb5 26.Bb4 Nd5 Black’s problems are over) 24...bxc4 (24...Rad8??
loses to 25.cxb5+ Kxb5 26.Rab1+) 25.Rac1 Rac8 26.Rxc4+ (in case of 26.Rd4 Kb6 27.Rdxc4 Rxc4 28.Rxc4 Nd5 Black
survives) 26...Kb6 27.Rb1+ Ka7 28.Rcb4 White has unpleasant pressure, yet Black has chances to hold. You would not
normally welcome this position with Black unless you had given it some careful analysis beforehand;
A) 28...Ba8 29.Nd3 Nd7 30.R1b2 Bc6 (30...Re6 31.Bg3) 31.a4 Re6 32.Bg3 Ka8 33.Kf2 and White still has some
pressure;
B) 28...Bc6 29.Bc5+ Ka8 30.Bd4 Nd5
analysis diagram

31.Ra4! (after 31.Nxd5 Bxd5 32.Ra4 Re6 White’s edge would evaporate) 31...Nc7 (31...Bxa4 32.Nxd5) 32.Ra5 f6
33.Kf2. Black is still a bit worse. And the position would suit Karpov much better than the Magician from Riga.

24.c4 Rac8

Karpov recommends 24...Rad8, but I think he would not have minded much: 25.c5+ Kc6 (after 25...Ka7 26.a4 b4 27.a5
b3 28.Rab1 Bc6 29.Rd4 wins the pawn) 26.Nd3 Kd7 27.Ne5+ and Black’s king has yet to find an opportunity to relax.
24...bxc4 is a much better try: 25.Rab1+ Kc6 (maybe this line is objectively better than the text, maybe this can even be
proved with the help of a computer. However, very few players would like to find their king in the centre like this)
26.Rb4 (Kasparov prefers 26.Rd4, e.g. 26...Red8 27.Rb4 Rxd6 28.Rdxc4+ Kd7 29.Rxb7+ Ke8 30.Rcc7 Rd7 and Black
exchanges this rook and frees his other rook. To a human this looks less promising than his next improvement on
Karpov’s line) 26...Rac8.
analysis diagram

A) If now 27.Rxc4+ Kb6 (after 27...Kb5?! 28.Rb4+ Kc6 29.a4! Black’s king is in serious danger. Maybe not too long,
though, as it will be mated) 28.Rb1+ Ka7 29.Rcb4 Ba8 – this probably looks worse for Black than it actually is;
B) Kasparov examines 27.a4!. This move aims to keep up the pressure against the black king. Kasparov would know,
he has proved this in so many of his games; 27...g5 28.Rxc4+ Kb6 29.Bc7+ Ka7 30.Rd6 Ne4 31.fxe4 gxf4 32.Rd7
Rxe4 33.Rxe4 Bxe4 34.Rxf7 Kb7 35.Bxf4+ – according to Kasparov Black has good chances of a draw here. Let me
add that a professional player may play on forever in such a position and still not win.

25.cxb5 axb5 26.a4!

He keeps going after the king, leaving Black no time to get organized.
26...Rcd8?

Tal wants to ease his troubles by swapping a rook. This loses in a number of ways. Most commentators, at this point,
have no longer looked for improvements. However, the game should not have been over yet. Black should have tried to
double rooks on the second rank:.
A) 26...bxa4 is a risky way to accomplish the doubling:
A1) 27.Rdb1+ helps the black king to reach the kingside: 27...Kc6 28.Rxa4 Kxd6 29.Rxb7 Rf8 30.Raa7 Ke5 and Black
holds;
A2) 27.Rxa4 Rc2 (of course Black is living dangerously, but he can still resist) 28.Nd3 and now:
A21) 28...Bc6 29.Rb4+ Bb5 30.Ne5 Re2 31.Nxf7 Re1+ 32.Rxe1 Rxe1+ 33.Kf2 Re2+ 34.Kg3 – Black is a pawn down
and his king is far away; that is too much trouble;
A22) 28...Nd5 29.Rb1+ Kc6 30.Bg3
analysis diagram

30...Ree2
A221) 31.Ra7 Bc8 32.Ne1 Rc4 33.Rxf7 Re7 34.Nd3 Rxf7 35.Ne5+ Kd6 36.Nxc4+ Ke6 37.Nd6 Rf8 38.Rb8 Ne7 and
Black stabilizes his position in an endgame with 2 against 3 pawns. Quite torturous for Black!;
A222) 31.Ne5+ Kc5 32.Ra5+ Kd4 33.Nxf7 Rxg2+ 34.Kh1

analysis diagram

Now 34...Rgd2? 35.Ra4+ Kc5 36.Nd6 loses. The entire line looks very dangerous for Black, but with the counter-thrust
34...Ne3! he may escape as it does not give White much time to build up his attack. His king can also come under fire.
B) Keeping the b-file closed with 26...Rc2!? is the safer way to get to the second rank:
B1) 27.a5+ Ka7 and Black has reached a higher level of safety than ever in this game;
B2) after 27.Rdb1 Bc6 28.Nd3 (28.axb5 Bxb5) 28...Ree2 Black is also still kicking;
B3) 27.Rab1 Bc6 (27...Ba6 28.Ba3) 28.h4 and White keeps up the pressure, but this is much better for Black than the
game; in fact he has chances to hold.

27.axb5!

At this point Karpov most probably saw the winning plan and he now carries it out with his usual admirable
purposefulness.
Also good was 27.Rac1!, which is Kasparov’s way to introduce the rook. One can see the differences in approach. Both
methods are extremely effective. In Karpov’s line, his opponent feels there is some breathing space, but not for long. In
Kasparov’s case, the opponent has no time to feel much as the blows are coming too quickly, viz. 27...Rd7 (27...Rc8
28.Bc7+) 28.Bc7+ Ka7 29.Rxd7 Nxd7 30.axb5.

27...Rd7

If 27...g5 28.Rac1 Rd7 29.Bc7+! wins as Karpov points out.

28.Rd4 Red8 29.Rad1 Rc8

29...g5 would lose to 30.Be7! Rxd4 31.Bxd8+.

30.Be5!

This is the final move of Karpov’s plan. It leaves Black without any chance. Now White exchanges and he either wins
further material or ensnares the king.

30...Re7

In case of 30...Rxd4, 31.Bxd4+ Kxb5 32.Bxf6 gxf6 33.Rd7 wins two pawns.

31.Rd6+ Kxb5
32.Rb1+!

Karpov could have won by collecting the pawn, but he starts an easy mating attack instead.

32...Kc4

If 32...Kc5 33.Nd3+ Kc4 34.Rb4 mate, and if 32...Ka5 33.Rd2 Rc4 34.Ra2+ Ra4 35.Bc3+ Ka6 36.Rxa4 mate.

33.Rd4+

33.Rd2 leads to a quicker checkmate: 33...Rxe5 34.Rc2+ Kd4 35.Rb4+.

33...Kc5 34.Nd3+ 1-0

Black cannot avoid the mate: 34...Kc6 35.Rc1+ Kb6 36.Rb4+ Ka7 37.Ra1+ Ba6 38.Bd4+. Once again, Karpov has
shown how well he can organize an attack on the opponent’s king in the endgame.
At the Amsterdam IBM tournament, Karpov beat Van der Wiel by keeping an equal position alive and when the chance
presented itself, he didn’t let it slip. Despite his loss against Ribli, Karpov won the event.
 GAME 65
Van der Wiel, John
Karpov, Anatoly
Amsterdam IBM 1980 (1)

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.c3 Nf6 5.d4 exd4 6.cxd4 Bb4+ 7.Bd2 Bxd2+ 8.Nbxd2 d5 9.exd5 Nxd5 10.Qb3
Nce7 11.0-0 0-0 12.Rfe1 c6 13.Ne4 h6 14.Ne5 Qb6 15.Nd6 Qxb3 16.Bxb3 Rb8 17.Re2 Nf5 18.Nxf5 Bxf5 19.Bxd5
cxd5 20.g4 Bh7 21.Rae1 Rbd8 22.h4 f6 23.Nf3 g5 24.Re7 Rf7 25.hxg5 hxg5 26.Re8+ Rxe8 27.Rxe8+

27...Kg7

Karpov plays for a win by simply maintaining the status quo. With this pawn formation, the knight should be better.
However, the g-pawn is fixed on the bishop’s colour.

28.Ne1 Rc7 29.f3 Kf7 30.Rb8

John ties Karpov’s rook to the defence of his pawns.

30...Ke6 31.Kf2 b6

The position is equal, but the opponents just keep on playing.

32.Ra8 Bb1 33.a3 Ba2 34.Ke3 Rh7 35.Re8+ Kd7 36.Rf8 Re7+ 37.Kd2 Ke6 38.Rh8 Rc7 39.Re8+ Kf7 40.Ra8 Bb3
41.Nd3

Black has not been able to make any real progress.

41...Ba4 42.Rh8 Ke6

43.f4

Just like Hort in Game 61, Van der Wiel runs out of patience doing nothing and changes the inherent structure of the
position. After 43.Ra8 it would have been interesting to know how long Karpov would have kept trying.

43...Rc2+ 44.Ke3 Rg2!

Karpov spots the new target and attempts to exploit the weakness.

45.f5+

White has several safe ways to continue:

A) 45.Kf3 – this is one of them. After 45...Rg1 46.f5+ Kd6 47.Nf2 Bd1+ 48.Nxd1 Rxd1 49.Ke3 Rb1 50.Rh6 White
cannot lose;
B) 45.fxg5 Rg3+ 46.Kd2 Rxg4 looks frightening for White, but...
analysis diagram

B1) 47.g6 Bb5 48.Nf4+ Kf5 49.g7 Rxg7 50.Nxd5 is equal;

B2) 47.Rh6! (from a famous Karpov-Kasparov game in the 1984 World Championship we know that with the d4-pawn
and a knight against d5-pawn and a light-squared bishop, White can consider sacrificing pawns) 47...Rxg5?! 48.Nf4+
Ke7 49.Rh7+ Kd6 50.Rxa7. Now maybe White can start squeezing.

45...Kd6 46.Rh6 Rg3+ 47.Kd2 Ke7 48.Nb4?

Opening up the position, which favours Black. With 48.Nf2! White could still tie Black’s hands and gain a draw:
48...Rb3 49.Rh7+ Ke8 50.Rxa7 Rxb2+ 51.Ke3 Rb3+ 52.Kd2 Rxa3 53.Ra8+ Ke7 54.Ra7+ with a perpetual.

48...Rxg4 49.Nxd5+?

It was possible to stop his free-fall with 49.Kc3 Rh4 (49...Rg3+ 50.Kd2) 50.Rg6 Rh3+ 51.Kd2 Bb3 52.Nc6+ Ke8
(52...Kf7 53.Nd8+) 53.Nxa7 and White is still very much in the game.

49...Kd6 50.Nxf6

50.Ne3 was problematic: 50...Rxd4+ 51.Kc3 Re4 (or 51...Ke5 52.Rxf6 Kxf6 53.Kxd4 Bc6 and White may soon run out
of moves) 52.Rxf6+ Ke5 53.Re6+ Kf4 54.b3 Bb5 55.Rxe4+ Kxe4 56.f6 Be8 and Black is likely to win.

50...Rxd4+ 51.Kc3

If 51.Ke3 Ke5 52.Nh7 Re4+ 53.Kf3 Bd1+ 54.Kf2 Kxf5 Black wins.
51...Rh4!

John had probably overestimated the power of his battery. This move pulls out the sting.

52.Ne4+

After 52.Rxh4 gxh4 53.Kd4 h3 54.Ng4 Bc2 Black wins. This also seems the case after 52.Rg6 Ke5, but White keeps
drawing chances with the surprising 53.Ng8!!. It’s really hard to trust a move like this and even if John saw it, one can
understand that he did not like to play it: 53...Kxf5 54.Ne7+ Ke5 (Black can try 54...Kg4 55.Rg7 Rh3+ 56.Kd4 as
well, but it’s hard to tell whether he can win with the extra pawn, as there are very few pawns left) 55.Nc8 g4 56.Nxa7
Rh7 and now:
A) 57.Nb5 Bxb5 58.Rg5+ Kf4 59.Rxb5, leaving Black three choices:
A1) 59...Rg7 60.Rxb6 g3 61.Rf6+ Kg4 62.Rf1 g2 63.Rg1 Kf3 64.a4 Kf2 65.Ra1 g1=Q (65...Re7 66.b4) 66.Rxg1 Rxg1
is only a draw;
A2) 59...g3 60.Rb4+! (forcing the king away or blocking the third rank; 60.Rxb6 loses to 60...g2 61.Rg6 Rh3+ 62.Kc4
Rg3) 60...Kf3 (60...Kf5 61.Rb5+ Kg4 62.Rxb6 Rg7 63.Re6 g2 64.Re1 and White can again sac the rook) 61.Rxb6 g2
(after 61...Rg7 62.Rf6+ Ke2 63.Re6+ Kd1 64.Rd6+ Kc1 65.Re6 White holds) 62.Rg6 Rh3 63.Rxg2 Kxg2+ 64.Kc4 is a
draw. Interestingly, 64.Kb4 would lose, but shouldering with 64.Kd4 would also hold;
A3) 59...Rh6! (not that hard to find, but there are several tempting possibilities) 60.Kd2 Re6! (decisively cutting off the
king) 61.a4 g3 62.a5 bxa5 63.Rxa5 g2 and Black wins.
B) 57.Rxg4!. White has only one pawn left and his king is standing well. Maybe this is an objective draw, it certainly
offers reasonale drawing chances over the board, for example: 57...Rxa7
B1) 58.Rg6 Ra6 (58...Rb7 59.Kb4 Be8 60.Rh6 Kd5 61.a4 and White probably holds) 59.Kb4 Be8 60.Rh6 Ra4+ 61.Kc3
b5 – probably Black can win this;
B2) 58.Kb4! Be8 59.b3 Rh7 60.a4 – this is very close, maybe it is a draw.

52...Kd5 53.Rxh4 gxh4

The distant passed pawn decides the outcome.

54.Ng5 Bd7! 55.f6 Kd6 56.Kd4

56.Nh7 Ke6 wins after 57.Nf8+ Kxf6 58.Nxd7+ Kf5. The h-pawn needs three moves to promote, but White can do
nothing to stop it.

56...h3 57.f7 Ke7 58.Nxh3

A desperate attempt. But despite the wrong colour of the bishop, Black wins effortlessly.

58...Bxh3 59.Ke5 Bd7 60.f8=Q+ Kxf8 61.Kd6 Ba4 62.Kc7 Ke7 63.Kb7 Kd6 64.Kxa7 Kc5 0-1

The following endgame is an exception. When René Olthof offered his opinion on my proposed Judit Polgar book, he
mentioned that a loss by Judit might be included in the collection. I also selected a loss for my Kasparov books and
now this idea is repeated with Karpov. I wanted to select an endgame in which Karpov lost as a World Champion, to
show that he is, after all, human. He who lives by the sword dies by the sword.
My other candidate games were three endgames won by Kortchnoi at Baguio or Timman’s win in the Buenos Aires
tournament, 1980. Finally I went for Ribli’s victory, as this reminded me of Karpov’s own play when he squeezed a
win out of virtually nothing.

GAME 66
Ribli, Zoltan
Karpov, Anatoly
Amsterdam IBM 1980 (6)

1.d4 Nf6 2.c4 e6 3.g3 d5 4.Bg2 Be7 5.Nf3 0-0 6.0-0 dxc4 7.Qc2 a6 8.Qxc4 b5 9.Qc2 Bb7 10.Bf4 Nd5 11.Nc3 Nxf4
12.gxf4 Nd7 13.Rfd1 Qc8 14.Ne4 c5 15.dxc5 Nxc5 16.Nxc5 Qxc5 17.Qxc5 Bxc5 18.Rac1 Rfc8

19.Ne5!
Ribli goes for simplification in order to invade with a rook along the seventh rank. His king will be slightly better and
the e5-knight will be strong. It is easy to underestimate the danger Black will soon face.

19...Bxg2 20.Kxg2 f6

A small move, yet it is hard to assess all the consequences it conceals. It opens a route to the centre for the king, but it
also opens the seventh rank for a white rook. Furthermore, it chases the knight away but weakens the e6-pawn.
20...Rc7?! is Dragan Ugrinovic’s suggestion in Chess Informant. It looks right to fight for the open file; however, the
plan has tactical drawbacks: 21.b4! Bd6 22.Rxc7 Bxc7 23.Nxf7! Bxf4 24.Rd8! (after 24.Nd8! White is also clearly
better) 24...Rxd8 25.Nxd8 Bd6 26.Nxe6 Kf7 27.Nc5 a5 28.Nb3. White keeps his extra pawn and Black’s activity
probably doesn’t give enough compensation.

21.Nf3

After 21.Nd3 the knight would block any invasion along the d-file. This can be an important element.

21...Bf8

Ugrinovic gives 21...Kf7 a question mark. However, things are not that simple: 22.Rd7+ Be7 23.Rcc7 Rxc7 24.Rxc7
Rd8! (rooks belong on open files) 25.e3 Rd6 26.Nd4 Ke8 and Black may not have too big a problem. However, one
normally doesn’t like to calculate such positions, even though a player may possess Karpov’s special calculating
abilities. Here White can try to play f4-f5 at once or mount the pressure with 27.Kf3.
21...Kf8! was found to be the simplest way to neutralize White’s initiative: 22.Rd7 Be7 23.Rcc7 Rxc7 24.Rxc7

analysis diagram

24...Bd6! (this is the point of approaching along the 8th rank) 25.Rd7 Bxf4 26.e3 (26.Nd4 Re8) 26...Be5 27.Nxe5 fxe5
28.b3. Now, in Novak-Prandstetter, Prague 1981, White played 28.Rb7 and a draw was agreed. Black is a pawn up and
wants to bring the rook to the second rank via the c-file. By moving the b-pawn from the second rank White could keep
the tension, but I think that Black’s extra pawn should be sufficient for the draw.
22.e3!

This is a simple and strong improving move in Karpov’s style. It removes a pawn from the second rank and defends f4.
It will also protect the knight on d4.

22...g6?

This is a mysterious move; maybe it is a waiting move. Quoting Ribli: ‘Karpov offered a draw, I did not accept it.’ That
is all the Hungarian grandmaster shared with the readers. This move shows that Black doesn’t clearly see what he
should do. Karpov commits a common error. Players tend to make a bad move when they offer a draw, as they are
thinking about the latter and that diverts their concentration from the moves they should play. Be careful with your
move when you offer a draw; also check the opponent’s move when he offers!
After 22...Kf7!? 23.Rd7+ Ke8 24.Rxc8+ Rxc8 25.Ra7 Rc6 26.Nd4 White would also be better. Obviously, it would not
be a simple task to exploit the advantage.

23.b3!

Ribli doesn’t hide his intentions but starts his subtle plan on the queenside. Ugrinovic mentions 23.h3, which would be a
waiting move.

23...Bb4

Blocking the a-pawn with 23...Ba3 does not promise much. After 24.Rxc8+ Rxc8 25.Rd7 it is unpalatable for Black to
have his king restricted to the eighth rank. After 25...Bc5 26.Nd2 White is pressing.
Alternatively, 25...Rc2 26.Nd4 would also give White an edge (only not 26.Ra7, when 26...Bc1!? (26...Rxa2 27.Rxa6, or
26...Rc6 27.Nd4 Rb6 28.Kf3 and White is a bit better) 27.Nd4 Rxa2 28.Kf3 Bd2 gives Black enough counterplay):
26...Rd2 27.Rd8+ (27.Ra7 Bc5) 27...Kf7 28.Nxe6 Rxa2 (in case of 28...Rxd8 29.Nxd8+ Ke7 30.Nc6+ Kd6 31.Nd4,
White hangs on to the extra pawn) 29.Nc7 f5 30.Rd7+
analysis diagram

30...Be7. No-one likes to invite a pin like this unless it is necessary. However, Black has good drawing chances here:
31.Nd5 Ke6 32.Rxe7+ Kxd5 33.Rxh7 a5 and Black’s queenside pawns are quick.

24.h3!?

In case of 24.Rxc8+ Rxc8 25.Rd7 Rc2 26.Nd4 Rxa2 27.Nxe6 Ra3 28.Rg7+ Kh8 29.Ra7 Kg8 it is not easy to extract
more from the position.

24...Kf8

Every move looks a bit worse, so it is hard to choose which disadvantage is the least damaging.
24...Be7 gives reasonable chances as well: 25.Rxc8+ Rxc8 26.Rd7 (once again things look frightening from Black’s
point of view) 26...Kf7 27.Ra7 Rc6 28.Nd4 and Black has to be very careful.
24...Ba3 would lead to lines similar to the ones with 23...Ba3.

25.Nd4 Kf7 26.a4!

Ribli softens up Black’s queenside admirably.

26...bxa4

Not 26...Bc5? 27.axb5 Bxd4 28.Rxc8 Rxc8 29.bxa6! when White can go for more than just a superior rook endgame (as
after 29.Rxd4 axb5 30.Rd7+ Kg8 31.Rb7 Rc5 32.b4): 29...Bb6 30.Rd7+ Ke8 31.Rb7 Bc7 32.b4 and Black is in trouble.

27.bxa4 Bc5?!

Karpov wants to get rid of the knight.

A) 27...Bc3 would not allow the game continuation and it serves the same purpose, yet it is still not appetizing: 28.Ne2
Bb4 29.Rd7+ Be7 30.Rcc7 Rxc7 31.Rxc7 and Black is kept very passive;
B) 27...a5!? 28.Kf3 Ke7 29.Nc6+ Ke8 30.Rc4 (after 30.Ke4 Ra6 31.Nd4 Rxc1 32.Rxc1 Kd7 Black is safe) 30...Rc7
31.Rdc1 Rac8 32.e4 Kd7 and Black is likely to survive;
C) With 27...Rc5!? Black could have tried to double rooks on the c-file.

28.Rc4!

White improves his rook. Ribli writes in his analysis that he likes this move.

28...Ba3

In case of 28...Bxd4 29.Rcxd4 Re8 30.Rb4 Re7 31.a5 Black’s position remains rather passive.

29.Rxc8

With 29.Nc6! White could have limited Black’s freedom of movement. Possibly this was even stronger than the text,
for example: 29...Ke8 30.Rb1! and White keeps his advantage here as well.

29...Rxc8 30.Rb1! Rc4

Karpov is looking to activate his rook. He probably anticipated that he would lose a pawn, but hoped that his bishop
would function better than the knight.
After 30...Rc7 31.Rb6 (with 31.Rb8!? the White rook could cause a lot of headaches on the eighth rank, yet Black might
not be lost) 31...Bc5 32.Rxa6 Bxd4 33.exd4 Rd7 Black has decent drawing chances.
31.Rb7+

Finally Ribli’s rook gets to the seventh rank, even with a check.

31...Be7 32.Ra7 e5

Black had no choice but to weaken his pawn structure.

33.fxe5 fxe5 34.Nf3 Rxa4 35.Nxe5+ Kf6

Karpov overestimates his chances. He probably missed that his king can still be vulnerable here.
With 35...Ke6!? he would have moved further away from g6 but, more importantly, he would have kept the f6-square
vacant for the bishop: 36.Nc6 (36.f4 Bf6) 36...Bf6! (the bishop will support the a-pawn from the diagonal) 37.Rxh7
Rc4 38.Na5 Rc5 39.Nb3 Rb5 40.Nd2 Be7 41.Rg7 Kf6 42.Rg8 a5 and White is better, but it is hard to tell whether he
can win.

36.Nc6 Bc5?!

Karpov probably missed Ribli’s finesse and got over-optimistic. He might have started to think about winning the game.
He had recorded so many successes in endgames that this would not be a miracle. After 36...Bd6!? 37.Rxh7 Rc4
38.Nd4 Be5 39.Nf3 Bc3 40.Ra7 Rc6 41.e4 (41.Kg3 Kf5) 41...a5 42.Kg3 White has decent winning chances.

37.Rxh7 Ra2 38.Kf3

Not only stepping away from ...Bxe3.

38...a5 39.h4! a4

If 39...Ke6 40.Nd8+ Kf5 41.h5 Bb4 42.h6 and White wins.

40.Ke4!

Karpov is faced with his own recipe for success. Ribli not only does a good job exerting pressure, but he also seizes his
tactical chances superbly, just like Karpov often does so well. I am certain that Karpov did not fully appreciate the
power of this move.

40...Bf8?!

Preventing the mate but moving away his active bishop. Uncharacteristically Karpov misses his best chance. Possibly he
had missed the mate threat and now panicked. 40...Rxf2 (41.Nd8!) or the pawn push 40...a3 would have created
obstacles.
analysis diagram after 40...a3

41.Nd8!!. Karpov must have missed this lovely study-like tactic somewhere in his calculations. It exploits the fact that
Black’s rook is out of the game. However the a3-pawn is close to promotion, and Black can try pushing it further:
41...g5 42.h5! (in case of 42.Rh6+ Ke7 43.Ne6 Bb6 44.Nxg5 Rb2 the a-pawn suddenly becomes very strong) 42...g4!
(Black tears open the mating net and he will have two pawns on the second rank) 43.Rf7+ Kg5 and now:
A) 44.Rf5+ Kh6 45.Nf7+! (in the event of 45.Rxc5? Rxf2, the g-pawn saves Black) 45...Kh7 (45...Kg7 46.h6+)
46.Rxc5

analysis diagram

46...Rxf2. The rook clears the way for the g-pawn and cuts off the white king. It is a frightening prospect to allow two
black pawns to get so close to promotion, yet White has enough resources to win the game. In an over-the-board game
the two passed pawns could confuse the opponent, especially if he were taken by surprise:
A1) 47.Ng5+? Kh6 48.Ne6 a2 49.Ra5 (49.Rc1 g3 50.Nf4 g2) 49...g3 50.Ra8 (50.Nf4 Rxf4+!) 50...Kxh5 51.Nf4+ Kg4
is equal;
A2) 47.Rc7 a2 48.Ra7 g3! (48...a1=Q 49.Rxa1 Rxf7 50.Rg1 Rg7 (in the event of 50...Kh6 51.Rxg4 Kxh5 52.Rg8!
White also wins) 51.Kf5 Rf7+ 52.Kg5 Re7 53.Rg3 Rg7+ 54.Kf4 Kh6 55.Rxg4 Rf7+ 56.Ke4 Kxh5
analysis diagram

57.Rg8!!. This is the only move to win as the eighth rank must be withheld from the black rook. Otherwise White
cannot start a frontal attack successfully) 49.Ng5+ Kh6 50.Nh3 g2 51.Ra8 transposes to line A3) below;
A3) 47.Ne5 a2 and now:
A31) 48.Rc1 g3 49.Nd3 Rd2 50.Nf4 Rf2 (50...Rb2 51.Ra1 g2 52.Nh3 Kh6 53.Kf3 and White wins) 51.Nh3 Rh2
52.Ng1 Rf2 53.Ra1 and White probably wins, but not automatically)
A32) 48.Ra5!? g3
A321) Now, in case of 49.Ra7+ Kh6 50.Ng4+ Kg5! (after 50...Kxh5 51.Nxf2 gxf2 52.Rh7+ Kg4 53.Rh1 Kg3 54.Ra1
White wins) 51.h6 Kxg4 (51...Rf7 52.Rxf7 a1=Q 53.Rg7+ Kh4 is equal as well) 52.h7 Rf8 53.Rxa2 Re8+ 54.Kd5 Rh8
Black will not lose;
A322) 49.Nf3! g2
analysis diagram

50.Ng5+ (this position would have been hard for White to evaluate objectively. Black’s two pawns on the second rank
are very threatening) 50...Kh6 (the triangulation with 50...Kg7 does not help: 51.Nh3 Kh6 52.Ra8! Kh7 (52...Kxh5
53.Nxf2; 52...Rd2 53.Kf3) 53.Ra6 and White wins) 51.Nh3 Kg7 52.Ra6 Kg8 (after 52...Rf1 53.Rxa2 g1=Q 54.Nxg1
Rxg1 55.Ra5 White wins) 53.Ra7 – White pushes his pawn to h6 and then starts pushing the e-pawn, which wins.
B) 44.Ne6+!. The simpler way to win, for example: 44...Kxh5 45.Nxc5 Kg6 (45...Rd2 46.Nd3 a2 47.Ra7 Kh4 48.Kd4
and White can start pushing his e-pawn) 46.Rf4 Rc2 47.Kd5 a2 (47...Kh5 48.Nd3 Rd2 49.Kd4) 48.Rxg4+ Kh5 49.Ra4
and White wins.
However, the black king can run away from the danger zone with 40...Ke6!!, when his passed pawn saves the game.
Ribli missed this escape in his analysis: 41.Nd8+ (in case of 41.Rh6 Rxf2 42.Rxg6+ Rf6 Black escapes) 41...Kd6
42.Nf7+ (or 42.Nb7+ Kc6 43.Nxc5 Kxc5 44.f4 a3 and as the a-pawn is so far advanced, the position is equal)
42...Kc6 43.Ne5+ Kb5 44.Rb7+ Ka6 and White cannot win.
Back to the game.
41.Ra7 Bd6

It looks like Black cannot take the key f-pawn, but tactics are on his side: 41...Rxf2!! (my fellow Hungarian also missed
this move in his analysis) 42.Ne5 (after 42.Rxa4 Rh2 there are too few pawns on the board for White to win) 42...Rb2!!
(this brilliant move saves Black) 43.Nd7+ (43.Ra6+ Kg7 44.Rxa4 Rb7 and Black holds. He could also exchange rooks
on b4) 43...Ke7 44.Nxf8+ Kxf8 45.Rxa4 and White can play on with his extra pawn, but objectively this is a draw.

42.f4 Rh2 43.Ra6 Kf7

After 43...Rxh4?, 44.Ne5 Ke6 45.Nc4 wins.

44.Ne5+ Bxe5 45.Kxe5 Kg7 46.Ra7+

The position was adjourned here. Somewhat surprisingly, Karpov continued.

46...Kh6 47.Rxa4 Rxh4

Black is not far from the other side of the river, but he will never get there. With his rook on the queenside he would
draw, but both his king and his rook are now out of the game. If 47...Rb2, 48.Ra8 wins.

48.Kf6! Rh5

Black’s next two moves serve to keep White busy and give him no time for Ra8. After 48...Rh1 49.Ra8! Kh7 50.Ra7+
Kh6 51.Rg7 Rg1 52.f5 g5 53.e4 wins easily.

49.e4 Rh4 50.e5 Rh5 51.e6

With 51.Kf7 Ribli could win in a more fancy way, but the text is safer. Why choose a winning line which depends on a
promotion with check? There follows 51...Rf5+ 52.Kg8 g5 53.Ra6+ Kh5
analysis diagram

54.Rf6! Rxf4 (or 54...Kg4 55.Rxf5 Kxf5 56.fxg5) 55.Rxf4 gxf4 56.e6 f3 57.e7 f2 58.e8=Q+ and White wins.

51...Rf5+ 52.Ke7 Kg7

After 52...Rb5 53.Kf7 Rf5+ 54.Kg8 wins.

53.Kd6 Rf8

The rook has no time to get to the queenside. After 53...Rb5, 54.Ra7+! Kh6 55.e7 wins.

54.Ra7+ Kf6 55.Rd7

In a lethal zugzwang, Karpov resigned.

GAME 67
Karpov, Anatoly
Hübner, Robert
Tilburg 1980 (5)

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Nf3 c5 8.Be3 Qa5 9.Qd2 0-0 10.Rc1 cxd4 11.cxd4
Qxd2+ 12.Nxd2 e6 13.Nb3 Rd8 14.Bg5 f6 15.Be3 f5 16.exf5 gxf5 17.Bb5 Nd7 18.0-0 Nf6 19.Bg5 Bd7 20.Bc4 b6
21.Nd2 Re8 22.Nf3 Bc6 23.Ne5 Bd5 24.Bb5 Rec8 25.a4 Ne4 26.Bf4 Bf8 27.f3 Nf6 28.Bg5 Kg7 29.Ba6 Rxc1
30.Rxc1 Bd6 31.Nc4 Bb4 32.Ne3 Re8 33.Bb5 Rf8 34.Bf4 Kg6 35.Kf2 Bb7 36.Rc7 Rf7 37.Rc2 Rf8 38.Bc4 Re8
39.Bb3!

Now the rook becomes a little stronger.

39...Re7

After 39...Nd5 40.Be5 White keeps up the pressure.

40.h4 h6 41.g3

Karpov also improves his pawn structure. He is exceptionally strong at tying up opponents, but this is a remarkable
example even by his standards!

41...Ba3?

In Chess Informant, Zdenko Krnic recommends 41...Kh7 as the least damaging move to Black’s position. The king
moves away from the centre. Black has many options, but no useful one.

42.Nc4 Bd5

After 42...Bb4 43.Ne5+ Kh7 44.Nc6 wins.

43.Ba2 Bxc4 44.Bxc4 h5 45.Bb3

Taking control of the c-file. It is just another significant ‘little’ move.

45...Kf7 46.Rc6 Bb2 47.Ke3

Further improving his king’s position.

47...Nd5+
This loses material in a lost position. If 47...Re8 48.Be5.

48.Bxd5!

Now the time has arrived to take direct action.

48...exd5+ 49.Be5 Re6 50.Rc7+ Re7 51.Rc2 Ba3 52.Kf4

This final improving move wins material. Hübner makes a few moves out of inertia, but his fate is sealed.

52...a6 53.Kxf5 b5 54.Rc6 b4 55.Rb6 Ke8 56.g4

Karpov doesn’t bother to win further material.

56...Kd7 57.gxh5 Rf7+ 58.Kg4 1-0

GAME 68
Karpov, Anatoly
Quinteros, Miguel
La Valetta Olympiad 1980 (12)

1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 0-0 6.Be2 e5 7.Be3 Qe7 8.dxe5 dxe5 9.0-0 c6 10.Qc2 Bg4 11.Rfd1 Na6
12.Rd2 Bxf3 13.Bxf3 Nc5 14.Rad1 Ne6 15.g3 Rfd8 16.Rxd8+ Nxd8 17.Rd2 Ne6 18.Qd1 Bf8 19.Bg2 Qe8 20.Bh3
Bb4 21.f3 Qe7 22.Kg2 Bc5 23.Bxc5 Nxc5 24.b4 Ne6 25.Bxe6 Qxe6
26.c5

White takes firm control of the d-file and gains an edge. Black has no weaknesses, though.

26...Qc4

Black quickly forces the queen exchange. His position requires careful play.
A) 26...a5?, trying to open the a-file, would have failed to 27.Rd8+ Rxd8 28.Qxd8+ Kg7 29.Qxa5 Qd7 (29...Qc4
30.Qa3; Black hopes to take advantage of White’s absent queen and slightly airy king) 30.Ne2! (but White can prevent
the invasion and then return with the lady) 30...Qd3 (30...Qd1 31.Kf2 Qh1 32.h4) 31.Kf2 g5 32.g4 and Black gets
nowhere;
B) With 26...Kf8 Black can involve the king, even if it is somewhat risky: 27.Rd6 (or 27.Qe2 Qe7 28.Qd3 Ke8 and
...Rd8 to follow) 27...Qc4 28.Qb3 (28.Rxf6 Kg7! 29.Rd6 (or 29.Qd6 Qxc3 and the White king is exposed to a
perpetual) 29...Qxc3 30.Qd2 Qxd2+ 31.Rxd2 a5 32.a3 axb4 33.axb4 Ra4 and all White’s advantage has evaporated)
28...Qxb3 29.axb3 Ke7 (the endgame is harmless for Black) 30.b5 cxb5 31.Nxb5 Rc8 32.b4 a6 33.Nc3 a5 – Black
exchanges all the queenside pawns and gains the draw;
C) Finally, 26...Kg7, preparing ...a5, looks like a reasonable option: 27.Rd6 Qc4 (the inserted moves favour Black. He
can try 27...Qe7 with the idea of ...Rc8, ...Rc7 and ...Rd7 as well) 28.Qb3 Qxb3 29.axb3
analysis diagram

29...a5! (again the rook does not stand really well on d6) 30.b5 cxb5 31.Nxb5 Rc8 and Black has absolutely nothing to
worry about.

27.Qb3 Qxb3 28.axb3 Kf8

A) After 28...a6:
A1) Grandmaster Geller considers 29.Rd6. This leads to an advantage, e.g. 29...Kg7 30.f4 (30.Na4?! a5 may provide a
kick) 30...Re8 31.Kf3 and this is plus/minus according to Geller;
A2) 29.Na4 (to transfer the knight to c4) 29...a5 30.Nb6 Ra6 31.bxa5 Rxa5 32.b4 Ra1 (32...Rb5 33.Rb2 traps Black’s
rook) 33.Nc4 Rb1 34.Nxe5 Rxb4 35.Rd8+ Kg7 36.Rc8 and Black is in trouble as he will lose the f7-pawn;
B) With 28...a5! Black could get rid of the pressure on the queenside, e.g. 29.Ra2 (29.b5 cxb5 30.Nxb5 Rc8 or 29.bxa5
Rxa5 30.b4 Ra3 and Black obtains counterplay) 29...Nd7 30.f4 (in case of 30.Rxa5 Rxa5 31.bxa5 Nxc5 32.b4 Na6
33.Na4 Kf8 Black holds the knight ending without any special difficulty) 30...f6!? and Black will bring the king to the
centre and attain the draw. 30...Ra6 is also reasonable.
29.b5!

Karpov usually achieves what he wants on the other side of the board or in the centre first, and only then brings his king
across.

29...Ke7!

In some previous games, Karpov’s opponents missed chances to equalize by temporarily sacrificing a pawn or by
becoming aggressive and exploiting the fact that their king was already in the centre. Here I have tried to find play for
Black based on 29...Rc8. Then, after 30.b4 cxb5 (30...Ke7 31.Ra2) 31.Nxb5 a6 32.Nd6 Rc7 33.f4 White is clearly
better.
Alternatively, 29...cxb5 30.Nxb5 a6 31.Nd6 Rb8 32.f4! Ke7 33.fxe5 Nd7 34.Rd5 and White dominates, e.g. 34...Ke6?
35.Nxf7!.

30.Ra2

Karpov’s play is sometimes likened to the embrace of a boa constrictor. Here he starts encircling his opponent with fatal
constriction in mind.

30...Ne8?!

A) If 30...Nd7 31.b4 a6 32.bxc6 (32.b6 Rd8 33.Rd2 (not 33.Na4 Nf8, when Black’s control of the d-file is important)
33...a5! 34.bxa5 Nxc5 35.Rxd8 Kxd8 is nice for Black) 32...bxc6 33.Kf2 and White is better, but it may not be enough
for a win;
B) With 30...cxb5! it was still possible to loosen the suffocating grip, but Black has only a small window of
opportunity: 31.Nxb5 (31.b4 a6 32.Nxb5 Ne8) and now:
B1) 31...a5 32.b4 a4 33.Ra3! (33.Nc3 a3) and Black cannot achieve any decent counterplay;
B2) 31...Rc8 32.Rxa7 Rxc5 33.Rxb7+ Ke6 34.g4 and Black does not get enough for the pawn;
B3) 31...a6! and now:
B31) 32.b4 Ne8 33.Nc3 (after 33.Na3 Rd8 34.Nc4 Rd4 Black becomes very active) 33...Rd8 34.Nd5+ Ke6 35.b5 axb5
36.Ra5 Nf6 and Black gets rid of the pressure;
B32) 32.Nd6 b6 33.Nc4! (33.Nb7 Nd7 34.c6 Nc5 35.Rd2 Rc8 36.b4 Nxb7 37.cxb7 Rb8 38.Ra2 Kd7 39.Rxa6 Kc6 and
Black has no problems) 33...bxc5 34.Nxe5 Nd7 and Black is slightly worse, but he should hold:

analysis diagram

35.Nc4!. White is still better and this is the type of position one doesn’t really like to defend when Karpov sits on the
other side of the board. Even so, it is likely to be tenable: 35...Rb8 (35...Ke6 36.Ra5) 36.Ra3 and now:
B321) 36...Nf8 37.f4 Ne6 38.Rxa6 Nd4 39.Ra7+;
B322) 36...Nb6 37.Ne3 Ra8 38.Ra5;
B323) 36...Rb4, intending to hold the rook ending with 3 pawns against 4. However, the rook becomes embroiled in the
tactical manoeuvres of the knight: 37.Kf2 Nb6 (of course Black can try to save the passive position with 37...Nb8, or
also 37...f6 38.Ne3 Rb7, but it remains unpleasant for him) 38.Ne5! (avoiding simplifications. After 38.Nd2 c4
39.Nxc4 Nxc4 40.bxc4 Rxc4 41.Rxa6 h5 Black draws) 38...Rb5 39.Rxa6 Rxb3 40.Ra7+ is unpleasant for Black;
B324) 36...Rb5 37.Kf2 Nb6 38.Nd2 (38.Ne3 a5) 38...a5 and Black holds, though White is still a fraction better.

31.b6

Fixing b7 and the pawn gets closer to promotion itself. Black now quickly runs out of air.

31...a6
32.Nb1!!

Karpov’s best ever Nb1 remains the one against Spassky, but this one is also very effective indeed.

32...Ng7!

Black has better chances with the rooks on the board, as he controls the only open file. After 32...Rd8 33.Rd2 (also, after
33.Na3 Rd3 34.Nc4 Ke6 35.Na5 Rd7 Black is very passive) 33...Nf6 (after 33...Rd4 34.Rxd4 exd4 35.Nd2 Ng7
36.Nc4 Ne6 37.b4 Black will go down, too) 34.Rxd8 Kxd8 35.Nd2 Nd7 36.b4 a5 37.bxa5 Nxc5 38.Nc4 f6 (38...Nb3
also loses: 39.Nxe5 Nxa5 40.Nxf7+ Kd7 41.Ne5+ Kd6 42.f4 Kc5 43.f5) 39.Nd6 Kd7 40.Nxb7! and wins.

33.Nd2 Ne6 34.b4 Rd8 35.Nc4 Rd4 36.Nd6 Rxb4?

Knuckling under. 36...Nd8 was necessary in order to defend the b7-pawn.

analysis diagram

A) 37.Ra4 Rd2+ 38.Kf1 (38.Kh3 Rd3) 38...Rxh2 39.Nxb7 Nxb7 40.Rxa6 Rd2 41.Ra7 Rd7 and Black can handle the
queenside problems; 42.b5 Nd8 43.Rc7 cxb5 and Black wins;
B) White can try the very interesting sacrifice 37.Nxb7. Black must play very carefully now. However, it is hard to
believe that it works for White: 37...Nxb7 38.Rxa6 Ke6 (38...Kd7? 39.Ra7 Kc8 40.b5 is slightly unpleasant) 39.Ra7
Nd8 (after 39...Rd7 40.b5 Nd8 41.Rc7 cxb5 42.c6 White is better) 40.b5 cxb5 41.b7 Nxb7 42.Rxb7 b4=;
C) Geller gives 37.Rb2! an exclamation mark and evaluates the position as plus/minus. It is a pity he did not exert
himself more. The rook will be tied to the defence of the b4-square for some time and should later be activated by
giving up the pawn. Black is worse, but it is not easy to crack his position. It would be nice to know how Karpov would
have forced a breakthrough.
There could follow: 37...Ke6 38.f4 (38.Kf2 f5 39.Ke3 fxe4 40.fxe4 (40.Nxe4 Nf7=) 40...h6) 38...f6 39.Kf3 Ke7
40.Ke3. White is pressing, there is no doubt about it. However, Black has chances to hold as it is not easy for White to
bring the b2-rook into play. If Black can hold, chess reveals one of its paradoxical faces. Karpov has played really
subtly, Black has committed several inaccuracies and still the game is drawn. Of course, often after one small mistake
the game is gone, especially against the big Ks.
37.Nxb7

Taking the pawn breaks the backbone of Black’s position.

37...Rb5 38.h4 h5

After 38...Nxc5 39.Nxc5 Rxc5 40.Rb2 the b-pawn promotes.

39.Kf2 Kd7??

This loses at once, but Black would lose anyway after 39...f6 40.g4.

40.Nd6 Rxc5 41.Rb2 1-0

Third Reign (1981-1983)

After a less impressive year in 1980 Karpov tried to regain his control by winning Linares and the Moscow super
tournament. In Amsterdam he shared 2nd-3rd place with Portisch behind Timman.
The year 1982 saw the rise of Kasparov to top level. Karpov’s superiority became less obvious. In Mar del Plata he
came second after Timman, to whom he lost. In the Phillips & Drew he shared first place with Andersson, but again
lost one game, against Seirawan. In Tilburg Karpov was looking for a clean victory again, and he achieved this with a
7½-11 score. At the Lucerne Olympiad Karpov produced an endgame masterpiece against Quinteros (Game 75). This
game was played when stakes were very high. The Soviet Union had to catch up with Hungary to win the gold medal.
Still, Karpov entered an equal endgame, relying on his skill. He didn’t mind going for an endgame where his opponent
objectively had good chances to hold. In the last round, Karpov went on to win a theoretically drawn rook endgame,
against IM Jakobsen.
Karpov came second behind Spassky at the Linares tournament in 1983, but his win against Larsen (Game 77) was
magnificent. From his victory in the 50th USSR Championship we take the endgame versus Romanishin (Game 78). At
Tilburg he beat Polugaevsky for the first time since their 1974 quarterfinal match (Game 79).

GAME 69
Karpov, Anatoly
Quinteros, Miguel
Linares 1981 (1)

1.d4 Nf6 2.c4 c5 3.Nf3 cxd4 4.Nxd4 e6 5.g3 d5 6.Bg2 e5 7.Nf3 d4 8.0-0 Nc6 9.e3 Bg4 10.h3 Bxf3 11.Bxf3 Be7
12.exd4 exd4 13.Bf4 0-0 14.Nd2 h6 15.Re1 Bd6 16.Bxd6 Qxd6 17.Nb3 Rad8 18.c5 Qc7 19.Rc1 Rd7 20.Bg2 Rfd8
21.Qd3 Re7 22.Rxe7 Qxe7 23.Rd1 Qe6 24.Kf1 Rd7 25.Qe2 Qf5 26.Kg1 Re7 27.Qf1 Ne4 28.Nxd4 Nxd4 29.Rxd4
Nxc5 30.Rd8+ Kh7 31.Qd1 b6 32.Rd4 Rd7

33.b4! Rxd4 34.Qxd4 Ne6

In case of 34...Qb1+ 35.Kh2 Ne6 36.Qe4+ Qxe4 37.Bxe4+ g6 38.f4 White is slightly better as his bishop is superior to
the knight.

35.Qd7!

The queen on d7 causes unpleasant problems and although these are not unsolvable, they run deep.

35...a5

After 35...Kg8?! 36.Qc8+! (not 36.Qxa7? Qb1+ 37.Kh2 Qxb4 and White has squandered most of his advantage, or
36.a3 Qb1+ 37.Kh2 Qb2 38.Qe8+ Nf8 39.Qe3 Ne6 and Black is worse, but he is still in the game) 36...Kh7 37.Qb7!
(it would have been hard to find this queen improvement. Now the bishop is threatening something as well) 37...Kg8
(37...Qb1+ 38.Kh2 Qxb4 39.Qxf7+–) 38.a3 Qb1+ 39.Bf1 Nd4 40.Qd5! (40.Qxa7? one must be very cautious when
placing the queen off-side like this: with 40...Nf3+ 41.Kg2 Ne1+ Black has a perpetual) 40...Qa1 41.Kg2 Qc3 42.Bc4
Ne6 43.Qa8+ Kh7 44.Bxe6 fxe6 45.Qxa7 the queen ending should be winning.
It would be unconventional to defend the pawn from the front with 35...Kg6!? (Black must be quite alert from now on)
36.Bf1 (in case of 36.Qb7 Qb1+ 37.Bf1 Qxa2 Black is not worse in this variety) 36...Qb1 37.a3 (in case of 37.Qd3+
Qxd3 38.Bxd3+ Kf6 Black’s king is centralized now: 39.f4 g5 40.fxg5+ Nxg5 41.h4 Ne6 42.Kf2 Ke5 43.Ke3 f5
44.a3 f4+ and the position is equal)
analysis diagram

37...Ng5 38.Qd6+ (38.h4 Nf3+ 39.Kg2 Qe4 and the battery is good for a perpetual) 38...f6 is a bit daunting, but White
is too tied up to exploit the weakened king: 39.Qd5 f5 with equality.

36.a3 axb4 37.axb4 Kg6?

The king wants to help the overburdened queen, but this fails to save the game. Among several unattractive moves,
there was one good enough to hold the position.
A) The saving move is not 37...Qb1+, as reducing the number of pawns is no remedy in this case: 38.Kh2 Qxb4
analysis diagram

39.Qxf7 and Black’s king loses such an important brick in the wall that it cannot be saved:
A1) 39...Nf8 40.h4 Ng6 41.Qf5! Qf8 42.Be4 Qxf5 43.Bxf5 h5 44.g4 Kh6 45.g5+ Kh7 and Black loses the knight;
A2) 39...Ng5 40.Qf5+ Kh8 41.Qc8+ Kh7 42.Bd5 Qd4 43.Qf5+ g6 44.Qd7+ Qg7 45.Qe8 Qh8 46.Qc6 and Black is
unlikely to survive with his exposed king, despite the limited material.
B) Nor is it 37...Kg8 38.Qd6 (38.Qc8+ Kh7 39.Qb7 Qb1+ 40.Kh2 Qxb4 41.Qxf7 may well be winning too)
38...Qb1+ 39.Kh2 Qb2 40.Qxb6 Nd4 41.Qb8+ Kh7 42.Qf4 f5 43.b5 and White consolidates his extra pawn;
C) The only move is 37...g6!?. This barely keeps Black in the game: 38.Qb7 b5 (in case of 38...Qb1+ 39.Kh2 Qxb4
40.Qxf7+ Ng7 41.Bd5 Black doesn’t survive the attack) and now:
C1) 39.Bf1 Nd4;
C2) 39.Bd5 Nc7 40.Ba2 Qd7 41.Qf3 Kg7 and Black manages to get organized;
C3) 39.g4 (winning a pawn, but at the cost of a weakening of his kingside) 39...Qf4 40.Qxb5 (40.Bd5 Nc7 41.Bc6
Qe5) 40...Kg7!. This is a subtle move. It is not easy to see that White can make no progress once the king defends f7:
C31) 41.Bf1 Nd4 42.Qc4 Nf3+ 43.Kg2 Nh4+ and Black has a perpetual because of the weakened h4-square;
C32) White can try to clear the way for the pawn: 41.Qb7 Qc1+ (with 41...Ng5 42.Qb5 (42.b5 Qc1+) 42...Ne6 Black
could repeat moves) 42.Bf1 Ng5 43.Qg2 and now Black can defend in an unusual way: 43...Qf4 44.b5 Nf3+ 45.Kh1

analysis diagram

In case of 45...Ne1 46.Qg3 Qe4+ 47.f3! Nxf3 48.Qg2 Black is still facing problems, but he can force a draw with
45...Qc1! (his motif seems to be unique) 46.b6 (46.Qxf3 Qxf1+) 46...Nh4.

38.Qd6 Qb1+

After 38...b5 39.Bf1 the pawn drops anyway.

39.Kh2 Qf5
40.Qxb6

The game is over now.

40...h5 41.h4 Kf6 42.Bh3 Qe4 43.Bxe6 fxe6 44.Qc5 e5 45.b5 Qb7 46.Qd6+ Kf5 47.b6 1-0

GAME 70
Ljubojevic, Ljubomir
Karpov, Anatoly
Linares 1981 (5)

1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.Nf3 Nd7 7.h4 h6 8.h5 Bh7 9.Bd3 Bxd3 10.Qxd3 e6 11.Bf4
Qa5+ 12.Bd2 Qc7 13.0-0-0 Ngf6 14.Ne4 0-0-0 15.g3 Nc5 16.Nxc5 Bxc5 17.Qc4 Bd6 18.Qa4 Kb8 19.Ne5 Nd5
20.f4 Nb6 21.Qb3 Bxe5 22.dxe5 Nd5 23.c4 Ne7 24.Be3 c5 25.Rxd8+ Rxd8 26.Rd1 Rxd1+ 27.Qxd1 b6
28.Qd3?!

Ljubojevic keeps the d-file. Maybe he was flirting with the idea of Qh7. In the game the initiative will swing to Karpov.
28.Qf3 Qd7 29.Bf2 was a better plan and should have been tried immediately. Karpov mentions 28.g4! as testing for
Black. White’s plan is to push Black back on the kingside. Very unusual complications could occur. They are
extremely hard to judge even for these specially gifted players.
Karpov now mentions only the two queen moves that occupy the long diagonal. These indeed lead to extremely
interesting positions, but they are not the only moves for Black:
A) 28...Nc6! (28...Kc8 is also all right). Black can prepare for 29.Qd6 this way, as after 29...Qxd6 30.exd6 f5 Black
prevents all white activity on the kingside. In case of 29.g5 hxg5 (with the retreat 29...Ne7! the knight blocks the
kingside and Black has a nice endgame) 30.fxg5 Qxe5 31.Qf3 Nd4 32.Qf2 Kc8 33.h6 gxh6 34.gxh6 Qe4 Black is not
worse;
B) It appears that Black can hold the game with 28...Qb7, too. Both lines are tight, but this is the shorter method, with
fewer chances to lose along the way: 29.Qd6+ Kc8 30.b3 (30.g5? Qe4; or 30.a3 Qf3 31.Kd2 Qg2+ 32.Kd3 Qf1+ with
a perpetual) 30...Qf3 (30...g6 loses to 31.g5! hxg5 32.h6 Qh1+ 33.Kb2 Qh2+ 34.Ka3 gxf4 35.Bc1!) 31.Bd2 (31.Qxe7
Qxe3+=) 31...Qh1+ 32.Kb2 Qg2 33.Ka3 (33.Kc3 Qf3+) 33...Nc6 34.g5 Nd4 35.Bc3 Nc2+ 36.Ka4
analysis diagram

36...Nd4 37.Ka3 (37.Bxd4?? Qxa2+ 38.Kb5 Qxb3+ 39.Kc6 Qa4 mate) 37...Nc2+ and after a few decent defending
moves Black finally gets a perpetual;
C) 28...Qc6 29.Qd6 Qxd6 30.exd6

analysis diagram

C1) Karpov doesn’t mention 30...Nc6, but it is good enough to draw: 31.g5 Nd4 32.Bxd4 cxd4 33.Kd2 f6 34.Kd3 e5
35.fxe5 fxe5 36.gxh6 gxh6 37.b4 Kc8 38.Ke4 Kd7 39.c5 bxc5 40.bxc5 Ke6 – despite the d6-pawn being closer to
promotion than the d4-pawn, White cannot leave Black’s pawns unguarded;
C2) 30...Ng8? – the outcome of this move is somewhat irrelevant to the position, nevertheless Karpov analyses it and
some breathtaking endgame motifs arise:
C21) 31.b4!? cxb4 (if 31...Nf6 32.bxc5 bxc5 (32...Nxg4 33.cxb6!) 33.Bxc5 a6 (33...Nxg4 34.Bxa7+!) 34.Bd4 wins)
and now:
C211) 32.g5 Kb7 33.c5 Kc6 34.Bd4 g6 35.hxg6 fxg6 36.Kc2 bxc5 37.Bxc5 a5 or 36.cxb6 axb6 37.Be5 h5 38.Kd2 and
in both cases the white king must stay in the square of the black h-pawn, so it’s a draw;
C212) 32.c5 Kb7! (32...bxc5 33.g5!?) 33.Bd4 (33.cxb6 axb6 34.Bd4 g6) 33...bxc5 34.Bxg7 f6 and Black stops the
onslaught.
C22) 31.g5! – fixing the knight on g8.

analysis diagram

C221) 31...f6?!. The champion explains what is wrong with this: 32.g6 Kb7 33.b4 cxb4 34.f5! e5 (34...Kc6 35.c5 bxc5
36.Bxc5 exf5 37.Bxb4 loses as well) 35.c5. Karpov stops here, evaluating the position as a small plus/minus. A closer
look reveals that it actually wins as Black runs out of moves: 35...Kc6 (or 35...b5 36.Kc2 Kc6 (36...a5 37.d7 Kc7
38.c6) 37.Kd3 a5 38.Bd2 Kd7 39.Ke4 Kc6 40.Be1 – poor Black! He has to make a move) 36.Kc2 bxc5 37.Bxc5 a5
38.Kd3 Kd7 39.Kc4 Kc6 40.a3 and White wins because of zugzwang;
C222) 31...Kb7 32.b4! Kc6 (if 32...cxb4 33.Bd4 f6 34.g6 Kc6 35.c5 bxc5 36.Bxc5 a5 37.Kc2 f5 (37...e5 38.f5 Kd7
39.Kd3 Kc6 40.Kc4 Black runs out of moves) 38.Kb3 Nf6 39.Bd4 e5 40.Bxe5 Nxh5 41.Ka4 and White wins because
he still has the a-pawn) 33.bxc5 bxc5 34.Bxc5 g6.
analysis diagram

Karpov gives this move an exclamation mark and calls the position unclear in his minimal Chess Informant analysis. In
my opinion, this is a complete misjudgement.
After 35.hxg6 fxg6 36.Kd2 White simply wins by marching to e5, for example 36...h5 37.Ke3 Kd7 38.Ke4 Kc6
39.Ba3 Kd7 40.c5 Kc6 41.Bc1 Kd7 42.Be3 Kc6 43.Ke5 Kd7 44.Bf2 and Black is hopelessly lost;
C223) 31...a5! (Karpov doesn’t mention this move, but it is worth considering as it slows down b2-b4. It seems the best
practical chance as White has to find some spectacular moves to win) 32.a3 (if 32.Kc2 Kb7 33.Kb3 Kc6 34.Ka4 Kxd6
35.Kb5 Kc7 36.Bd2 f6 (36...Ne7 37.Bc3 Nf5? loses to 38.Bxg7) 37.Bc3 (after 37.b3, 37...Ne7! solves Black’s
problems. Other moves are risky. There follows 38.gxf6 gxf6 39.Bc3 Nf5 40.Bxf6 Ng3 41.Bg7 Nxh5 42.Bxh6 Ng3
and Black is not under pressure anymore) 37...Ne7 38.gxf6 gxf6 39.Bxf6 Nf5 White has no hopes of winning) we have
arrived at another major crossroads:
analysis diagram

I) Not 32...Kb7? 33.b4 axb4 34.axb4 cxb4 (34...Kc6 35.bxc5 bxc5 36.Bxc5 again leaves Black without moves)
35.Kc2! Kc6 36.c5 bxc5 37.Bxc5 f6 (after 37...g6 38.hxg6 fxg6 39.Bxb4 White’s king holds the h-pawn, which would
not have been possible after 35.Kb2) 38.g6 f5

analysis diagram

IA) 39.Bd4? Kxd6 40.Bxg7 Kd5 41.Kb3 Ke4 42.Be5 Kf3 43.Kxb4 Kg4 44.Kc5 Kxh5 45.g7 Kg6 and Black will not
lose;
IB) It looks like White can go after the b-pawn with 39.Kb3? Nf6 40.Bd4!
analysis diagram

40...e5!! (this is an unbelievable move, yet it is the only way to draw. After 40...Nxh5 41.Be5 Kd7 42.Kxb4 White wins
by either invading from a5 or carefully walking to h4 to hunt down the knight) 41.Bxe5 Nxh5 42.Kxb4 Ng3 43.Bxg7
Kxd6. Now the point of the 40...e5 pawn sacrifice becomes apparent – the Bf8 check doesn’t win as the e6-square has
been vacated;
IC) 39.Bxb4! Nf6 40.Bc3! Nxh5 (in the event of 40...e5 41.fxe5 Nxh5 42.Bd2 White probably wins) 41.Be5 wins.
II) 32...a4 (trying to slow down b2-b4) 33.b4 axb3 and now:

analysis diagram
IIA) 34.Kb2 Kb7 35.Kxb3 Kc6 36.Ka4
IIA1) 36...Kxd6 37.Kb5 Kc7 38.a4 (38.Bd2 Ne7 39.Bc3 f6=, but not 39...Nf5 40.Bxg7!) 38...Ne7? 39.a5 bxa5
40.Bxc5 Nf5 41.Bf8! f6 (Black vacates the f7-square for the knight, and as a consequence the bishop cannot take on g7
now)

analysis diagram

IIA11) 42.c5. White covers the d6-square from the knight before he launches his threat. But Black has a very subtle
escape: 42...e5!! (not 42...hxg5 43.fxg5 fxg5 44.Bxg7!) 43.Bxg7 (43.fxe5 fxg5 44.Kxa5 g4 45.e6 g3 46.e7 Kd7 47.c6+
Ke8 48.c7 Nd6 and the tide has turned) 43...exf4 44.gxf6 f3 45.f7 f2 46.Be5+ Kd7 47.f8=Q f1=Q+ 48.Kb6 Qb1+ and
maybe White can play on, but Black should be able to hold;
IIA12) 42.Kxa5 (this way of stepping aside requires more skilful play to win) 42...Kc6 43.Bxg7 fxg5 44.fxg5 hxg5
45.Kb4 (45.h6? Nxh6) 45...g4
analysis diagram

46.Be5! (or 46.h6 g3 47.h7 g2 48.h8=Q g1=Q 49.Qc8+ Kd6 50.c5+ Ke7 51.Qc7+ Ke8 52.Bf6 – yes, White keeps the
initiative, but there is so little material left that Black should be able to hold) 46...Kd7 (46...Nh6 47.Bf4 Nf7 48.h6 e5
49.Be3 g3 50.h7 g2 51.c5 and White wins) 47.c5 (47.Kb5 g3) 47...Kc6 48.Bg3! (48.Bf4 Ne7) 48...Nh6 (48...Kd5
49.Kb5; 48...Kd7 49.Bf4) 49.Kc4 Nf7 50.Bf4 e5 51.Be3 and Black is in trouble;
IIA13) 42.Ka4!! (this elegant return of the king is the most effective way to prevent a ...Nd6 check) 42...Kc6 (after
42...hxg5 43.fxg5 fxg5 44.Bxg7 g4 45.h6 Nxh6 46.Bxh6 Kc6 47.Be3 White wins) 43.Bxg7 fxg5 44.fxg5 hxg5 45.h6
Nd6 46.h7 Nf7 47.h8=Q Nxh8 48.Bxh8 Kc5 49.Kb3 g4 (49...a4+ loses to 50.Kc3 a3 51.Bd4+ Kc6 52.Kb3) 50.Be5
a4+ 51.Kc3 a3 52.Bd4+ Kd6 53.Kb3 and White wins.
analysis diagram after 36.Ka4

IIA2) 36...f6! (keeping the white king away from b5 as long as possible) 37.g6 e5 38.d7 Kxd7 39.Kb5 Kc7 40.a4
(40.fxe5 fxe5 41.Bf2 Nf6 42.Bg3 Nd7=) 40...Kb7 (or also 40...Ne7! 41.a5 bxa5 42.Bxc5 Nf5 43.fxe5 fxe5 and Black
is safe) 41.a5 bxa5 42.Kxc5 a4 43.Kb4 e4 44.Kxa4 Kc6 45.Kb4 Ne7 and Black might even try to win from within his
fortress.
Back to the analysis diagram after 32...a4 33.b4 axb3.
IIB) As the black knight is only a spectator for the moment, White can lash out with a beautiful shot: 34.Bxc5!!.

analysis diagram

White’s pawn on a3 is still far removed from its promotion square. That is what makes this sacrifice so wonderful and
surprising.
34...bxc5 (34...Kb7 loses to 35.d7 Kc7 36.Bf8 f6 37.Bxg7) 35.a4 hxg5 (35...Kb7 36.a5 Kc6 37.a6 and one of the pawns
promotes) 36.fxg5 f5 37.a5 f4 38.a6 f3 39.d7 Kc7 40.a7 f2 41.d8=Q+ and White wins.
Returning to the game.
28...g6!

Karpov is happy to resolve the situation on the kingside.

29.hxg6 fxg6 30.a3

White could prevent his kingside pawns from being fixed with 30.g4!?. There follows 30...h5 31.Bd2 Qc6 (with
31...hxg4 32.Qg3 Nf5 33.Qxg4 Black will not take control) 32.gxh5 gxh5 33.Qh3 Nf5 34.Qxh5 Qa4 35.Bc3!? Qxc4
36.Qe8+ Kb7 37.Qd7+ and White keeps a perpetual up his sleeve.

30...a5 31.b3

31.g4!? still looks feasible.

31...h5! 32.Qe4
32...Nf5

Black has fixed the pawns and that is good news for him, as the white bishop has limited activity.

33.Bf2 Qd7 34.a4?!

Ljubojevic plans to create a fortress. He stops Black’s play on the queenside, but at the same time he relinquishes any
opportunity to keep Black busy on that side.

34...Kc7 35.Kc2

Not 35.Qa8?. As Karpov explains, White achieves nothing with the sole queen. Black plays 35...Qd3!, then interposes
the knight on e7 and White’s play is gone, whereas Black’s is about to begin.

35...Qd8!

Ljubomir may either have missed this, or he did not fully appreciate all the implications.

36.Kc1?

This is a very unfortunate waiting move – although Karpov has no comment to make.
After 36.Qg2! Karpov would have had to look for another plan: 36...g5 37.fxg5 Qxg5 (unlike in the game Black now
captures without check) 38.Qa8 and White holds.
36...g5!

He immediately exploits White’s mistaken last move.

37.fxg5

White cannot just wait with 37.Kc2? because of 37...g4 38.Kc1 h4 39.gxh4 g3 40.Be1 Qxh4 and White is brushed
aside.

37...Qxg5+ 38.Kc2

38.Kb2!? might have offered greater resistance.

38...Ne7!

This square clearance was the manoeuvre that had escaped Ljubo’s attention. White’s position is already lost.

39.Qh7 Kd7

Karpov doesn’t want to give the smallest of counterchances with 39...Qg6+!?. There would follow 40.Qxg6 Nxg6
41.Be3 Kd7 (41...Nxe5?? 42.Bf4) 42.Bg5 Nxe5 43.Bf6 Nc6 44.Kd3 Nd4 (44...e5 45.Ke4 Nd4 46.Bxe5 Nxb3 47.Ke3
and things are not so simple) 45.Ke4 (45.Bxd4 cxd4 46.Kxd4 Kd6 47.Ke4 Kc5 48.Kf4 (48.Ke5 Kb4) loses to
48...Kd4) 45...Nxb3 46.Bc3 Nd4 47.Be1 (47.Bd2 e5) 47...Nf5 48.Kf4 Nd6 and Black wins;
With 39...Qf5+ he could exchange queens at once, just as soon hereafter in the game: 40.Qxf5 exf5 41.Be3 (41.Kd3
Ng6) 41...Kd7 is like the game continuation.

40.Qe4 Qf5 41.Qd3+ Kc6 42.Qxf5 exf5 43.Be3 Ng6 44.e6 Kd6 45.Bg5 Kxe6 46.Kd2

If 46.Bd8 f4.

46...f4!
This forces the win.

47.gxf4 h4 48.Ke3 h3 49.Kf3 Kf5 50.Kg3 Nxf4!

Nice little tactics can be useful in endgames.

51.Bd8 Ne2+ 52.Kxh3 Nd4 53.Bxb6 Nxb3 54.Bd8 Ke4 55.Kg4 Kd3 56.Kf4 Kxc4 57.Ke4

57...Kc3!

Karpov plays superbly; the king shoulders his adversary away first. If 57...Kb4?! 58.Kd3 c4+ 59.Kc2 Kxa4??
(59...Nd4+ 60.Kd1 c3 (not 60...Kxa4?? 61.Bf6!=) 61.Bf6 Kc4 still wins) 60.Kb2 White can draw.

58.Bf6+ Kc2

Also 58...Kb4 59.Kd5 c4 (59...Kxa4?? 60.Kc4) 60.Ke4 Kxa4 wins.

59.Be5

If 59.Kd5 Kd3.

59...c4 60.Ke3 c3 61.Bf6

61.Bg7!? would set a trap. Black cannot use the same winning method as in the game: 61...Nc5?? 62.Kd4 Nxa4
63.Kc4! Nb6+ 64.Kb5 a4 65.Kb4 and now the bishop is standing on the right square.
But 61...Nd2! 62.Kd4 Kb3 63.Kd3 Nb1 64.Bf6 c2 65.Bg5 Nc3 66.Bh6 Na2 does win.

61...Nc5

61...Nd2 62.Kd4 Kb3 63.Kd3 Nb1 wins as well.

62.Ke2
This loses right away. Karpov shows that other moves were also losing, for instance: 62.Bd8 Kd1 63.Bxa5 c2 64.Bd2
Nb3, or 62.Kd4 Nxa4 63.Kc4

analysis diagram

63...Nb6+! (Karpov shows that Black can spoil even this position if he plays 63...Nb2+??. Then White plays 64.Kb5 a4
65.Kb4 and eats up the pawns) 64.Kb5 a4 65.Kb4 Nd5+ wins.

62...Kb3 0-1

Here is a game from the match against Kortchnoi in Merano, where Karpov experienced less problems to defend his
title than he had three years earlier.
 GAME 71
Kortchnoi, Viktor
Karpov, Anatoly
Merano World Championship 1981 (9)

1.c4 e6 2.Nc3 d5 3.d4 Be7 4.Nf3 Nf6 5.Bg5 h6 6.Bh4 0-0 7.Rc1 dxc4 8.e3 c5 9.Bxc4 cxd4 10.exd4 Nc6 11.0-0 Nh5
12.Bxe7 Nxe7 13.Bb3 Nf6 14.Ne5 Bd7 15.Qe2 Rc8 16.Ne4 Nxe4 17.Qxe4 Bc6 18.Nxc6 Rxc6 19.Rc3 Qd6 20.g3
Rd8 21.Rd1 Rb6 22.Qe1 Qd7 23.Rcd3 Rd6 24.Qe4 Qc6 25.Qf4 Nd5 26.Qd2 Qb6

27.Bxd5?

This is a strange mistake in a World Championship final. Tension often causes blunders, but a mistake that contradicts
principles is less understandable, particularly in such a well-known type of position.
In an isolated-pawn position, White should keep as many minor pieces on the board as he can, as they prevent the rook
from exerting pressure on the d-pawn. In this game Kortchnoi makes this mistake for the second time: he has
voluntarily exchanged a knight on move 16 as well. To make things even weirder, Polugaevsky, who was a truly great
player, doesn’t give a question mark to either move in his Chess Informant analysis.
This second unfortunate exchange helps Black as it gives him more space to organize an attack against the d4-pawn.
27.a3 should have been played.

27...Rxd5 28.Rb3?!

White position is critical as ...e6-e5 will come and then the d4-pawn falls, but putting the rook on the b-file makes the
white king more vulnerable. After 28.f4 Black would play 28...Qd6 with ...f6 and ...e5 to follow.

28...Qc6 29.Qc3 Qd7 30.f4

White is in so much trouble that he has to weaken his king in order to stop ...e5 – a move that would win a pawn.
30...b6 31.Rb4

This rook will now be out of the game.

31...b5!

And Karpov doesn’t let it return.

32.a4 bxa4!

32...a6 also looks good and was probably more in Karpov’s style. Probably, he already smelled blood.

33.Qa3 a5 34.Rxa4 Qb5

Karpov defends the a5-pawn before launching the knockout blow.

35.Rd2
35...e5!

For a change Karpov doesn’t make any of his usual ‘improving’ moves. He simply checkmates his opponent. When
Fischer was asked about what was required to become a great player, he mentioned several mental aspects. Chesswise
he said you need to know when to press and when to dig in. Karpov knows this very well. It became apparent to me
while writing this book, that Karpov often attacks the king with killer precision.

36.fxe5 Rxe5 37.Qa1

In case of 37.Rf2, one rook is not enough to defend the king from the threat of the heavy pieces: 37...Re1+ 38.Kg2
Qc6+ 39.Kh3 Rg1 40.Rxa5 Qe6+ 41.Rff5 Rxd4 42.Qc3 Rgd1 and Black wins.

37...Qe8!

Karpov creates threats on the e-file and keeps an eye on the a4-rook at the same time. This strong move blows apart
White’s defence.

38.dxe5

If 38.Kf1 Qc6 39.Kg1 Rc5 wins.

38...Rxd2

It is all over. Kortchnoi’s king cannot escape the crossfire. He pays for putting his heavy pieces out of play on the a-file.

39.Rxa5

After 39.Qe1 Qd8 wins.

39...Qc6 40.Ra8+ Kh7 41.Qb1+

Maybe Kortchnoi did not know that they had already passed the time limit. He could have resigned.

41...g6 42.Qf1 Qc5+ 43.Kh1 Qd5+ 0-1

GAME 72
Karpov, Anatoly
Miles, Anthony
London Phillips & Drew 1982 (9)

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 g6 6.Be3 Bg7 7.f3 0-0 8.Qd2 Nc6 9.g4 Be6 10.0-0-0 Nxd4 11.Bxd4
Qa5 12.a3 Rab8 13.h4 Rfc8 14.Nd5 Qxd2+ 15.Rxd2 Bxd5 16.exd5 a6 17.Be2 Nd7 18.f4 Nc5
19.Rh3!

This is the most effective way to bring the rook into the game.

19...Rc7

Karpov mentions 19...Bxd4!?. An active break-up of White’s centre follows after 20.Rxd4 e5 21.fxe5!? (21.dxe6 Nxe6
22.Ra4 and White is still preferable) 21...dxe5 22.Rd1. Karpov evaluates this as slightly advantageous to White.

20.Re3!

Karpov first improves his pieces. The rook stands well on e3.

20...b5 21.Bxg7 Kxg7 22.Rd4!

Improving his second rook as well.

22...a5?!

Miles wants to play ...b5-b4 and restrain White. But he weakens the b5-pawn.
Karpov mentions 22...Na4!? but unfortunately stops there. What would White do with his small initiative? After 23.c3
(23.Rde4 Kf8) 23...Rbc8 White has a couple of options. One of them is 24.Rde4 (others are 24.Bd1 Nb6 25.Bb3 or
24.Kc2 h6 25.Bf3 Nb6) 24...Nb6 (24...Kf8! looks better) 25.Rxe7 Rxe7 26.Rxe7 Nxd5 27.Re4 and White is a little
better.
23.b4!

Stopping ...b4 and fixing the b5-pawn.

23...Na4?!

An unforced pawn sacrifice. Thus, Black puts himself in a situation where any mistake becomes expensive.
A) In case of 23...e5? 24.fxe5 dxe5 25.d6 exd4 (25...Rc6 26.Rd5 wins according to Karpov) 26.dxc7 Rc8 (or 26...Rf8
27.Re5 Ne6 28.bxa5 Nxc7 29.Re7 Rc8 30.Bf3 Kf8 31.Rd7 Ne6 32.a6 Nc5 33.Bb7 Nxd7 34.Bxc8 and White wins the
knight) 27.Re5 Ne6 28.bxa5 Rxc7 29.a6 White would have a clear advantage;
B) 23...axb4! 24.Rxb4 (if 24.axb4 Na4 25.Kd2 (in case of 25.Bd3 Nc3 it looks like Black will grind White’s play to a
halt) 25...Rbc8 26.Bxb5 Rxc2+ 27.Kd3 R8c3+ 28.Ke4 Rh2 Black gets counterplay) and now all that Karpov says is that
this is plus/ minus, without any further analysis. Kasparov contradicts him and evaluates the position as balanced. But
Kasparov stops after 24...Rcb7. It is possible to build up pressure with White, but it is inconclusive:
B1) 25.Kd2 – White is slightly better after 25...e6 26.Bf3 Rb6, but Black’s position is stable. There is no clear way to
increase the advantage;
B2) In the event of 25.c4 bxc4 26.Rxb7 Rxb7 27.Bxc4 Ra7 Black is not worse;
B3) 25.g5 h6 26.Kd2 hxg5 27.hxg5 e6 28.Bg4 exd5 29.Bf3 Ne4+ 30.Bxe4 dxe4 31.Rexe4 Ra8 and Black holds.

24.bxa5!

White not only wins an ordinary pawn, but in certain cases it can grow into a very strong passed pawn, backed by a
light-squared bishop. The b-pawn wasn’t even an ugly duckling, yet it will turn into a handsome prince who later
decides the outcome of the game.

24...Nc3 25.Bf1 Kf8

Black has the option to play for a fortress with 25...Ra8 as he has no time to recapture the pawn. After 26.Kb2 the White
king intervenes (Kasparov recommends 26.Rdd3 as an attempt to improve: 26...Na4 27.Rb3 Rac8 28.Bd3 Nc5 29.Rxb5
Nxd3+ 30.Rxd3 Rxc2+ 31.Kb1 Rh2 unclear, according to Kasparov): 26...Na4+ 27.Kb3 Nc5+! – Kasparov points out
that this check disturbs White’s king march (after 27...Rxa5 28.Kb4! Raa7 29.Bxb5 the king decides the issue) 28.Kb4
Na6+ 29.Kxb5 Rb7+ 30.Kc4 Nc5 and Black’s pieces firmly control the queenside.

26.Kb2

26...Rbc8?

It is easy to miss Karpov’s next move and it is even easier to fail to appreciate its power. This move allows the white
king to become more active.
Kasparov points out that Black was still able to stay in the game by annoying the king with knight checks. So:
26...Na4+! 27.Kb3 Nc5+ 28.Ka2 Nb7 29.a6 (29.c3 Nxa5 30.Rb4 Rc5 31.a4 Ra8 32.Bxb5 Rxd5 and Black has enough
pieces on the queenside to keep White modest there) 29...Rxc2+ 30.Kb1 Rf2 31.Bxb5 Nc5 32.Rb4 Ra8. White is tied
up and cannot make any progress on the queenside, therefore Black is not worse.
27.Kb3!

This is a very powerful king move, Black has a lot of forces along the c-file and that makes this advance even more
beautiful.

27...Rc5 28.a6!!

Tony did not fully appreciate the strength of this pawn push during the game. Very soon he would realize how
devastating it was.

28...Nxd5 29.Rxd5! Rxd5

30.Rc3!!

Exchanging the c8-rook in order to clear the way for the a-pawn is the point of White’s idea.
After 30.Bg2 Rd4 31.Bb7 (31.a7 d5) 31...Rd8 32.Rd3 Rxf4 33.Rd5 White still has the better chances, but the game
continuation is more convincing.

30...Rd8

After 30...Rxc3+ 31.Kxc3 Rc5+ 32.Kb4 Rc7 White has a couple of wins. 33.Bg2 is Karpov’s choice.

31.Rc7!

The safest way to collect the full point. Pushing 31.a7 would win as well, viz. 31...b4 32.axb4 Rd1 33.Ba6 Ra1 34.b5
Ra8 35.Rc7.

31...Rd1 32.Bxb5

The rest is trivial, Black cannot cope with the passed a-pawn.

32...e5 33.a7 exf4 34.Rb7 Rb1+ 35.Ka4 Rxb5 36.Rxb5 f3 37.Rb8 f2 38.Rxd8+ 1-0

GAME 73
Bouaziz, Slim
Karpov, Anatoly
Hamburg TV-A 1982 (5)

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.d3 Nf6 5.0-0 d6 6.c3 a6 7.Bb3 Ba7 8.Nbd2 0-0 9.Nc4 h6 10.Ne3 Be6 11.Re1 Re8
12.Bxe6 fxe6 13.h3 Qd7 14.Ng4 Rf8 15.Be3 Nxg4 16.hxg4 Bxe3 17.Rxe3 Qf7 18.Qe2 Qg6 19.Nh2 Rf6 20.Rf3 Rxf3
21.Qxf3 Qg5 22.Qe3 Qxe3 23.fxe3 g5 24.Rf1 Kg7

As usual Karpov starts by improving his pieces. The text prevents 25.Rf6.

25.Kf2
25...Nb8!!

It was the knight’s turn to find a better square. It is heading for the outpost c5.
The following example remains Karpov’s best and probably most important knight manoeuvre.

Karpov - Spassky
Leningrad Cand. Semi-final 1974 (9)

24.Nb1!! Qb7 25.Kh2 Kg7 26.c3 Na6 27.Re2 Rf8 28.Nd2 Bd8 29.Nf3 f6 30.Rd2 Be7 31.Qe6 Rad8 32.Rxd8 Bxd8
33.Rd1 Nb8 34.Bc5 Rh8 35.Rxd8 1-0
Returning to the game.

25...Nb8!! 26.Nf3 Nd7 27.Kg3?

Like a goalkeeper who dives too early in one direction at a penalty, the long-time best African player defends the g4-
square prematurely – before it is attacked. 27.b3!? was an alternative.

27...Nc5!

The king has left the centre, so Karpov turns his attention there.

28.Rd1

The rook is known to be a poor defender.

28...a5!

This not only gains space but has another function too. Remember, little Tolya already played ...a5 when he was only
ten.

29.Kf2

The king returns to the weakness that is under pressure.

29...Ra6!

Karpov employs the rook in an original way. There is already a threat.

30.Ke2?

Hastening to defend d3 and relieve the rook of this burden. However, he misses something. Instead 30.b3 would restrict
Black’s advantage to a minimum.
30...Na4!

Suddenly Karpov wins a pawn. Bouaziz had not noticed that his weakness was under imminent threat.

31.d4

White looks for counterplay for the pawn, but this just accelerates the end.
After 31.Rd2, 31...Rb6 wins the pawn. Playing 32.c4 against Karpov is unlikely to work. At a lower level it might save
the half point. When something goes wrong, ‘hacking’ may work. However, in a position like this where there are no
other problems, White has to consider acting like a hedgehog! Nowadays there are no adjournments, so the opponent
must win the game with limited time. There is a chance that he goes astray during the process.

31...Rb6

Returning to football terminology: Black has scored the first goal. As he not only takes a pawn but improves his rook as
well, he keeps on attacking to gain another pawn. 31...exd4 looks winning as well.

32.dxe5 Rxb2+ 33.Kf1 Nxc3 34.exd6!?

This is no doubt witty but it changes little.

34...cxd6

We cannot know whether Karpov saw this in advance (I think this was a one-hour game) but Black wins easily in any
case.
Not 34...Nxd1? 35.d7 Nxe3+ 36.Ke1 Nxg2+ 37.Kf1 Rb1+ 38.Ke2 Nf4+ 39.Ke3 and Black has no more than a
perpetual. Your author once had a similar ending against Gelfand. The young Boris showed a remarkable calculating
ability, which has helped me understand how important this is, during my career as a trainer.

35.Rxd6
35...Rb1+!

Just as he did so many times in his early years, he pins the opponent’s piece to prevent it from becoming active.

36.Ne1 Kf6

Activating his king and making sure the rook has no active moves. To win this position now doesn’t require the ultimate
in technique.

37.Rd2 b5 38.Rc2 b4 39.Kf2 Ra1 40.e5+ Kxe5 41.Nf3+ Ke4 42.Nd4 Kd3 0-1

GAME 74
Torre, Eugenio
Karpov, Anatoly
Tilburg 1982 (3)

1.d4 Nf6 2.Nf3 e6 3.c4 b6 4.g3 Ba6 5.b3 Bb4+ 6.Bd2 Be7 7.Bg2 0-0 8.Ne5 c6 9.Bc3 d5 10.Nd2 Bb7 11.e4 Na6
12.0-0 c5 13.exd5 exd5 14.Ng4 Qd7 15.Ne5 Qc8 16.Re1 dxc4 17.Ndxc4 Bxg2 18.Kxg2 cxd4 19.Bxd4 Rd8 20.Qf3
Rxd4 21.Nc6 Bc5 22.Nxd4 Bxd4 23.Nd6 Qf8 24.Rad1 Bc5 25.Nxf7 Kxf7 26.Qb7+ Kg8 27.Qxa6 Kh8 28.Re2 h6
29.Rd3 Qf7 30.Qa4 Qb7+ 31.Kg1 a5 32.Qf4 Rf8 33.Qf5 Qc6 34.Qg6 Qc8 35.Rf3 Qb7 36.Qd3
36...Qa8!

Karpov will occupy the d-file with this subtle move. Black has a small material advantage; furthermore, his bishop is
really strong and there is no way of removing it. However, there are relatively few pawns and White has no
weaknesses. Karpov has to create one.

37.Rf4

37.Rd2 is no better, as it concedes another file after 37...Re8.

37...Rd8 38.Qf3 Rd1+ 39.Kg2 Qd8

Karpov feels his chances are better with the queens on.

40.Qc6 Rd6 41.Qf3 Qd7 42.Rf5

42...Rd3!

Chasing the queen away from the king.

43.Qa8+ Kh7 44.Rfe5

Other rook moves are less resilient.

A) After 44.Rf4? Nd5 45.Rf3 Rxf3 46.Kxf3 Nc3 47.Re8 Qf5+ wins;
B) If 44.Rf3?! Rd5! – Black wants to switch the rook to the h-file and have the queen join it there; and after 45.Qa6
Black is clearly better. Karpov prefers 45...Rd4 to 45...Qc6, which also looks very promising for Black.
44...Qg4!

So now we can all see the weakness that Karpov spotted. It is the white king itself.

45.Qb7

If 45.Qc6 Nd5 wins.

45...Bd6

According to Karpov he should have played 45...h5! using the pawn to soften up White’s king further: 46.h3 Qg6 47.h4
Qg4! (this is better than Karpov’s 47...Bxf2. White’s position collapses) 48.Qa8 (48.f3 Qd4) 48...Bxf2!.

46.h3

White temporarily slows down the attack, but at the cost of weakening the kingside. After 46.R5e3 Nd5 wins.

46...Qg6 47.R5e3 Rd1 48.Re1?

The long-time Filipino number one player wants to reduce the strength of the attack with exchanges, but he opts for a
less fortunate way of doing this. With 48.Rf3! – Karpov gives this move without exclamation marks – White finds the
only way to stay in the game. He can force the exchange of queens and put up resistance in the endgame: 48...Nd5
(48...Bc5 49.Qc8 Ne4 50.Qf5 and White reaches the endgame) 49.Qf7 Qxf7 (49...Qb1 50.Qf5+) 50.Rxf7 Bc5 and
Karpov indicates minus/plus. Black has good prospects of converting his advantage.

48...Nd5 49.Re6

After 49.Rxd1 Nxe3+ 50.fxe3, 50...Qc2+ wins.

49...Nf4+ 50.Kh2 Nxe6 51.Rxd1 Bc5 52.Rd2

After 52.Kg2 Nf4+ White sheds a pawn. If 52.Qg2 Qc2 53.Rf1 Nd4 54.f4 Qd3 White is in big trouble.

52...Ng5!

Black simply has too many pieces around White’s king.

53.Kg2

Perhaps the ugly 53.Qg2 was relatively the best. 53...Qe4 wins easily, though.

53...Qf5! 54.Qd7

Defending the h3-pawn, but leaving the king on his own. If 54.g4 Qf4 55.Re2 h5!.
54...Qe4+

It has taken Karpov 18 moves of streamlined play to reach a situation where he can finish off.

55.Kf1 Nf3 56.Re2

After 56.Rd1 Nd4 57.Kg1 Ne2+ 58.Kf1 Nxg3+ 59.fxg3 Qf3+ 60.Ke1 Qf2 mates.

56...Nh2+ 57.Ke1 Bb4+ 0-1

A typical Karpov game. Already at a very young age he knew how to go after the enemy king, even in the endgame.
This feature of his play has hardly been noticed. He kept his sting in his adult years.

GAME 75
Karpov, Anatoly
Quinteros, Miguel
Lucerne Olympiad 1982 (11)

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 e5 7.Nb3 Be7 8.0-0 0-0 9.Be3 Be6 10.Qd2 Nbd7 11.a4 Nb6
12.a5 Nc4 13.Bxc4 Bxc4 14.Rfd1 h6 15.Nc1 Rc8 16.Nd3 Be6 17.Bb6 Qe8 18.f3 Nd7 19.Be3 f5 20.exf5 Rxf5
21.Nb4 Nf6 22.Nbd5 Nxd5 23.Nxd5 Bf8 24.b3 Qf7 25.Nb6 Rc6 26.c4 Be7 27.Qe2 Bd8 28.Nd5 Bxd5 29.Rxd5 b5
30.axb6 Bxb6
31.Kh1!

As usual Karpov prevents as many tactical possibilities as he can. With the king in the corner Black cannot hope for a
check. White is somewhat better as he controls d5 and Black has more pawn islands. In addition, Black’s king is
slightly exposed to attack by the heavy pieces.

31...Bxe3?!

White was unlikely to take on b6. Most of the time one would advocate letting the opponent take and so build up one’s
own position rather than taking yourself and building up the opponent’s game for him. After the text move the queen is
freed earlier than it should have been. But this is probably not the decisive mistake.
Preferable was 31...Qb7!?, as White has only one weakness – the b3-pawn. Black should look for a way to attack it,
tying White’s hands. After 32.Rad1 Rf6 33.R1d3 a5 Black can think of easing the grip with ...a5-a4 at some stage (also
after 33...Qa7 34.Bd2 Qb7 35.Qe4 Bd4 Black’s position is solid): 34.Rb5 (34.Bd2 Qc7) 34...Qc7 35.Bxb6 Rxb6
36.Rdd5 Rxb5 37.Rxb5 Rf4 38.Qa2 Rd4 39.h3 Qf7 and Black obtains reasonable play.

32.Qxe3 Qc7

32...Qb7!? is Sasa Velickovic’s suggestion in his Chess Informant analysis. It has its merits, as the queen exerts
influence against the b3-pawn. It also saves a tempo compared to the game continuation:
A) 33.c5 dxc5 34.Qd3 (in case of 34.Rad1 Rf8 35.Rxe5 c4 Black exchanges the queenside pawns) 34...Rf8 35.Rxe5
Qf7 and if White is better at all, his edge is insignificant;
B) 33.Rad1 and now:
B1) 33...Rb6 34.R1d3!? (34.Rxd6 Rxb3 35.Qe2 Rc3 36.R6d5 (or 36.h3 Rf7 37.Ra1 Qc7 38.Qxe5 Qxc4, as Black has
safely taken White’s last queenside pawn, he can hold the position) 36...Qc6 37.Qe4 Rxc4 and Black is saved because
of the first rank) 34...Rf6 35.h3 – it is very hard for Black to meet c4-c5;
B2) 33...Qb4!. The queen helps to carry out ...a5 and ...a4, and is well positioned on b4: 34.R1d3 (34.Qd3 Rf6)
34...Rf6. Black’s position is solid. It is difficult for White to make any progress.
33.Rad1

Karpov starts building up pressure on d6.

33...Rf6?!

The Argentinian grandmaster protects d6 in advance, but this is somewhat passive and costs a tempo. The text gives
Karpov free rein to break open the position.
A) In case of 33...a5 34.Ra1 Rc5 35.Qd2 White dominates the queenside;
B) Alternatively, with 33...Rf8 Black could have tried bringing this rook to the queenside to expose White’s only
weakness – the b3-pawn. If its defence ties White’s hands, he cannot concentrate all his forces to break open the black
position. However, this move doesn’t solve all of Black’s problems either: 34.Qd3!? Rb8 35.h3 (White first creates a
luft before he acts) and now: 35...Qb6
B1) 35...Rcb6 36.Rxd6 transposes to line B4) ;
B2) 35...a5 36.Ra1! Ra8 (36...Rcb6 37.Ra3) 37.Qe4 a4 (37...Rca6 38.Rad1 is very unpleasant for Black)
analysis diagram

38.c5! dxc5 39.Qc4 Kh8 40.bxa4 and White has a clear edge;
B3) 35...Rbb6 36.f4 exf4 37.Rf5 Rc5 (37...Rb8 38.Rxf4 Rcb6 39.Qd5+ Kh8 40.Rdf1 will give Black little pleasure,
even though he might survive) 38.Rxf4 Re5 39.Rdf1 and White keeps his advantage;
B4) Black can simplify to a rook ending a pawn down with 35...Qb6, 36.Rxd6 (36.c5 dxc5 37.Rxe5 Rf6 is also
unpleasant for Black) 36...Rxd6 37.Qxd6 Qxd6 (after 37...Qxb3 38.c5, the c-pawn is more dangerous than the a-pawn)
38.Rxd6 Rxb3 39.Rxa6 Rc3 40.Rc6 Kf7 and Black has real drawing chances despite White’s extra pawn.

34.h3 Qb6 35.Qd3 Qb4

36.Kh2!

A signature move. Karpov puts the ball in his opponent’s court.

36...Qb8?

It is hard to tell, but it looks as though after this mistake Black can no longer save the game. It is amazing, and to exploit
the error requires the touch of a maestro.
A) 36...Rb6? would have been too ambitious now: with 37.c5! White can already successfully expose Black’s king:
37...dxc5 38.Rd8+ Kf7 39.Qd7+ leads to a computer-assisted win: 39...Kg6 40.Qe8+ Kf5 41.Qh5+ g5 42.Qe8! Qxb3
43.Qd7+ Kg6 (43...Qe6 44.Qh7+) 44.Rd5 and Black’s kingside is taken apart.
Incidentally, 39.Qh7 also wins: 39...Qf4+ 40.Kh1 Rb7 41.R1d7+ Rxd7 42.Rxd7+ Ke6 43.Qxg7 and Black has no
perpetual: 43...Qc1+ 44.Kh2 Qf4+ 45.Kg1 Qe3+ 46.Kf1.
B) Velickovic now recommends 36...Qb6!. This move has a point, but once again it is very hard to anticipate what
Karpov was going to come up with. For instance, after 37.Re1 Rc5! Black eases the pressure: 38.Rxc5 Qxc5 39.Ra1 and
White is still better but his edge is smaller than in the game.

37.Re1!!

Karpov injects energy in a magnificent way. His intention is to create a direct attack on the king. His ferocity is second
only to Kasparov. It is no longer possible to see how Black can save the game, and if this is true, it is simply stunning.

37...Rf8

It is too late now to play Velickovic’s simplifying 37...Qb6? because of 38.Re4 Rc5 39.b4! Qxb4 (39...Rxd5 is
somewhat better, but still rather tough for Black: 40.Qxd5+ Kh7 41.c5 Qb5 42.Rc4 Rg6 43.Qe4 and Black is in deep
trouble) 40.Rxd6 Rxd6 41.Qxd6 a5 (41...Qa5 42.f4 wins) 42.Rg4. Velickovic evaluates this as plus/minus, but his nice
line is actually winning for White.

38.Re4! Qc7
If 38...Qb6 39.c5! Rxc5 40.Rxd6 Qb5 41.Qe3! – this time there is agreement with Velickovic’s evaluation.

39.Rg4!

So this is the square that Karpov wanted to transfer the rook to. The h6-pawn is hanging. Do you believe that this is the
final station of the rook’s journey?

39...Rf6 40.b4!

A superb switch to the opposite wing.

40...Qb6?!

40...Rf8 would also have been devoid of hope for Black, but still this would have been better than the game
continuation: 41.b5 axb5 42.cxb5 Rb6 43.Rg6 Rd8 44.Qe3 (this is one of several options for White to break open the
position) 44...Kf7 45.Qe4 Qb7 46.f4.

41.c5 Qc7

Not 41...Qb5? 42.Qb3.

42.Kh1!?

The king leaves the diagonal of the queen. Blocking the diagonal with 42.Re4!? was promising as well: in case of
42...Re6 (42...dxc5 43.Rd7) 43.f4 wins.

42...a5

Black doesn’t wait, as White intends to play 43.Re4 and go after the e-pawn.
43.Rc4!!

Brilliant usage of the rook. All Black can do now is choose on which part of the board he wants to lose.

43...Qb8

Black decides to lose the pawn on the queenside.

A) After 43...dxc5? Black would lose on the kingside: 44.Rd8+ Rf8 45.Rd7 Qc8 46.Rg4 and the king is caught;
B) 43...axb4 would lose in the centre: 44.cxd6 Qd7 45.Rxc6 Qxc6 46.d7 and White promotes;
C) 43...e4!. Black has suffered so much in this game that he probably did not even think of any active possibility. This
loses as well, but it gives White a chance to spoil his winning position: 44.fxe4! (in case of 44.Rxe4 axb4 45.cxd6
Rfxd6 46.Rxd6 Qxd6 Black saves the game. Also, after 44.Qxe4 axb4 45.Rxd6 Rcxd6 46.cxd6 Qxd6 Black escapes)
44...axb4 45.cxd6 Rfxd6. Now the e-pawn may promote with surprising speed.
analysis diagram

46.e5! Rxc4 47.exd6 (47.Rxd6 Rf4) 47...Rc1+ 48.Kh2 Qb8 (48...Qd7 49.Qb3) 49.Ra5 Kh8 50.Qd5 Rc8 51.Rb5 and
White wins.

44.cxd6 Rcxd6 45.bxa5 Qd8 46.Rcc5 1-0

 GAME 76
Gheorghiu, Florin
Karpov, Anatoly
Lucerne Olympiad 1982 (12)

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.b3 Bb4+ 6.Bd2 Be7 7.Bg2 c6 8.0-0 d5 9.Ne5 0-0 10.Bc3 Bb7 11.Nd2 Na6
12.e3 c5 13.Qe2 Rc8 14.Rfd1 Rc7 15.e4 cxd4 16.Bxd4 dxc4 17.Ndxc4 Qa8 18.f3 Rfc8 19.Nd3 Nb8 20.Nde5 Ba6
21.Qe1 Ne8 22.Ne3 Bc5 23.Bxc5 Rxc5 24.Nd7 Nxd7 25.Rxd7 R5c7 26.Rxc7 Rxc7 27.Rc1 Qc8 28.Rxc7 Qxc7
29.Qd2 Qc5 30.Bf1 Bxf1 31.Kxf1 Kf8

32.Ke2

The alternative 32.Nd1 was generated by Karpov’s effect on your author. The question arose: how can White improve
the knight? 32...b5 33.Qe3 Qa3 34.Qd2 and the knight goes to c3, which seems to be a better square than e3. White
does all right here.

32...Ke7 33.Qc2 Qh5!?

These players knew each other well. Karpov provokes his opponent, who, at this stage of his career, was mainly looking
to maintain peaceful coexistence on the board and sign the scoresheet after sharing the point.

34.Nf1?!

White merely tries to defend. It was healthier to play normally. After 34.h4 g5!? (or 34...Qe5 35.f4 Qa1 36.e5 and
White has improved his position) 35.hxg5 Qh2+ (in case of 35...Qxg5 36.Qb2 (36.Kf2) 36...Qxg3 37.Qa3+ White is
not worse) 36.Kd3 Qxg3 37.Qc6 Qxg5 38.Qb7+ White has nothing to worry about.

34...Kd7 35.Qc3

Trying to tie the knight to the defence of g7.

If 35.Qd3+!? – Karpov supplies these marks, as with this move White intends to do something active – 35...Nd6
36.Qa6 Qb5+ 37.Qxb5+ Nxb5 Karpov evaluates the position as ‘equal plus’. Let me translate. The position is equal
and the plus is because he is playing the position with black. This is only partly a joke.

35...f6!

This simple and strong move allows Karpov to bring in the knight soon.

36.h4!

White intends to free the knight from its burden.

36...Nd6 37.Kf2

If 37.Qd3 Qa5.

37...Nb5

So Karpov has achieved something tangible – his knight comes into play.

38.Qd3+ Ke7 39.Ne3 Qc5 40.Qd2?!

Going backwards is passive.

With 40.Qc4! White could almost force a draw: 40...Qe5 (otherwise White swaps queens with an easily drawn position)
41.a4! Nc3 (in case of 41...Nc7 42.f4 or 42.Qc6 White is doing all right) 42.Ng4 and White can force the queen
exchange here as well.

40...a5 41.Qd3
41...Nd4

Karpov relentlessly improves his knight.

42.Kg2?

Moving out of the pin, but leaving the centre.

A) In case of 42.a4?, 42...Qb4 43.Nc2 Qxb3 wins the pawn as the champion has pointed out;
B) With 42.Qc4! – a move which Karpov doesn’t mention – the long-time Romanian number one could have obtained
a playable position: 42...Kd6 (after 42...Qd6, 43.Nd5+! kills all the play. And in the event of 42...Qe5 43.Nc2 Nb5
44.a4 Black loses his advantage) 43.Qa6 Nc6 (43...Nc2 44.Qd3+) 44.Qc8 (White can play adventurously, but 44.Qd3+
should also be fine) 44...Qd4 45.Qf8+ Kd7 46.Qxg7+ Ne7 47.Qh6 (47.f4 Qxe4 48.Ke2 should be OK as well)
47...Qd2+ 48.Kf1 Qxa2 49.Ng4 with a sharp and complicated position, but White’s prospects are not worse.

42...Nc6 43.a4?

Not ideal, as it is easier to attack b3 than a2. Actually White may no longer be able to save the game!
With 43.Qc4 he could keep some play. In case of 43...Qxe3 44.Qxc6, White holds easily without the knights.

43...Ne5!

Karpov knows exactly when to swap a piece and when to keep it on the board. He analyses 43...Qd4, offering the queen
exchange.
A) Now after 44.Qxd4? Nxd4 45.Nc4 Nxb3 46.Nxb6 Kd6 Black wins according to Karpov. The disadvantage of
dispatching the king to g2 becomes clear;
B) 44.Qc2 and now:
B1) 44...Qxe3 45.Qxc6 Qe2+ 46.Kh3 Qxf3 47.Qc7+ Kf8 48.Qd8+ Kf7 49.Qd7+ Kg6 50.Qxe6 h5
analysis diagram

51.Kh2!. This is much easier than Karpov’s suggestion, for whenever the black queen moves, White keeps a perpetual.
The champion analyses 51.Qe8+ Kh7 52.Kh2 and stops here, claiming equality. White has chances to obtain a draw
but he must accept a difficult endgame: 52...Qe2+ 53.Kh3 f5 54.e5 Qg4+ 55.Kh2 f4 56.gxf4 Qxf4+ 57.Kg2 Qg4+.
There is no point in finding out whether White can hold this;
B2) 44...Kd7 45.Nc4 (Karpov assesses 45.Qc1? as ‘equal/plus’ with no further moves. After 45...Qd3! White just loses
a pawn) 45...e5 46.h5 and White is just a fraction worse.

44.Qd2 Qd6!

This time Black offers the exchange. This is the correct approach as it leads to a win.
After 44...h5 45.f4 Nc6 46.f5 White gets some play, as Karpov points out. For example: 46...Qd4 47.Qc2.

45.Qc3

A) Karpov hints that 45.Qc2!? may have been better, but this is very tough for White as well: 45...Qd3 and now:
A1) 46.Kf2 h5! (in his analysis Karpov prevents his opponent’s pawn from getting close to promotion. After 46...Kd6
47.h5! Qxc2+ 48.Nxc2 Nd3+ 49.Ke3 Nc5 50.Nd4 White gets counterchances) 47.Qc7+ Nd7 48.Nc4 (48.Qc2 Nc5)
48...Qc2+ 49.Kg1 Qxb3 50.Qd6+ Kd8 51.Qxe6 Qxf3 – the black queen’s invasion probably wins; at least it yields a
huge advantage, as Karpov points out;
A2) 46.f4!? Qxc2+ 47.Nxc2 Nd3 48.Kf3 (or 48.Ne3 Nc5 49.Nc4 Nxe4) 48...Nc5 and Black’s advantage gives him
very good winning chances.
B) Alternatively, 45.Nf1 Qd3! 46.Qxd3 (after 46.Qe3 Qxe3 47.Nxe3 Kd6 Black’s king is very strong) 46...Nxd3
47.Nd2 Kd6 48.Kf1 Kc5 49.Ke2 Kd4 and this kind of advantage is usually enough for a win.
45...Qd3!

Karpov bases his play on the distant white king. He removes the queen, the strong defending piece, and then invades.

46.Qxd3

White exchanges queens and that is the signature to his own death penalty.
Keeping the queens on with 46.Qc7+ was not much better, though. Black just has to find the right place for his king:
46...Kf8!? 47.Qb8+ (47.Qxb6 loses to 47...Qe2+ 48.Kh3 Qxf3) 47...Kf7 48.Qc7+ Kg6 49.h5+ Kh6 50.Ng4+ Nxg4
51.fxg4 Qxe4+ and Black wins.

46...Nxd3 47.Nc4
47...Nc1!

One can imagine that the Romanian grandmaster had missed this subtle approach in his calculations. After 47...Nc5?
48.Nd2 White is worse, but he remains in the game.

48.Nxb6

After 48.Nd2 Kd6! the point of 47.Nc1 can be seen – the route for the king is open.
Karpov mentions 48.b4 as a way of giving up the pawn. After some skirmishes White goes down anyway: 48...axb4
49.Kf2! Kd7 50.Nxb6+ Kc6 51.Nc4 Kc5 52.Nb2 b3 53.Ke3 Kb4 54.Kd2 Na2 and Black wins.

48...Nxb3

Black is clearly superior because of his better king. In a knight ending the weaker side seldom escapes with a draw.
When you are worse, try to avoid getting into a knight ending!

49.e5

Reducing the material doesn’t help.

White is not able to stop the Black king invasion with 49.Nc4 because after 49...Kd7 50.Kf2 (or 50.Na3 Kc6 51.Nb5
Kc5 52.Nc7 Nd4) 50...Kc6 51.Ke3 Kc5 52.Kd3 Kb4 Black wins.

49...fxe5

It is a pawn, but more importantly Black clears the way for his king.

50.Nc4 Kd7 51.Kf2 Kc6 52.Nxe5+

After 52.Ke3 Kd5 53.Nb6+ Kc5 54.Nd7+ Kb4 wins.

52...Kd5 53.Nd3 Nc5!

Karpov offers an exchange for the last time. He has controlled all exchanges in this game with style. The text wins the
a-pawn and White has no more hope of saving the game.

54.Ke3 Nxa4 55.Nf4+ Ke5 56.Nd3+

If 56.Nh5 g6.

56...Kd6 57.Kd4 Nb6 58.Ne5 h6!

Black could allow 59.Nf7-Ng5-Ne4-Nc3 as well, but then his pawn is one square further removed from promotion.
Both methods win, but Karpov’s is more precise.

59.Nf7+ Ke7 60.Ne5 a4 61.Kc3 Kd6 62.Nd3

After 62.Nf7+ Kc5 63.Nd8 Nd5+ 64.Kb2 Kd6 the knight soon invades to take White’s kingside pawns.

62...Nd5+ 63.Kb2 Ne3

Black resigned.

GAME 77
Karpov, Anatoly
Larsen, Bent
Linares 1983 (9)

1.e4 c6 2.d4 d5 3.Nd2 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.h4 h6 7.Nf3 Nd7 8.h5 Bh7 9.Bd3 Bxd3 10.Qxd3 e6 11.Bf4
Ngf6 12.0-0-0 Be7 13.Kb1 a5 14.Ne4 Nxe4 15.Qxe4 a4 16.Ne5 a3 17.b3 Nxe5 18.dxe5 Qa5 19.Bd2 Qc7 20.f4 0-0
21.f5 exf5 22.Qxf5 Qc8 23.Qxc8 Rfxc8 24.c4 b5 25.Kc2 Kf8 26.Rhf1 Ke8 27.Rf4 Ra7
28.Be1!

It is not easy to see the point behind this move.

28...Rb7 29.Rg4! Bf8 30.Rgd4 b4

Karpov writes that 30...Bc5 should have been played. After 31.Rg4, the rook has to stay on the 4th rank because of the
threat of ...bxc4 bxc4 Rb2 with check. He stops his analysis after 31...Bf8, claiming the position is equal. That may well
be the case, but I think Karpov would have played on. For example, 32.c5!? Rd7 33.Rxd7 (if the queens were on the
board, 33.Rd6 would cause a headache, but here Black should not be worse after this exchange sacrifice) 33...Kxd7
34.b4 Ke6 35.Rg3 Ra8 36.Bc3 – Black’s pieces are tied to the defence of their pawns, but it is hard for White to invade,
or to collect the a3-pawn. However, I think Karpov would have tried something here.
31.Bh4

So Karpov has found play against Black’s king.

31...Be7 32.c5 Bxh4?

Black relinquishes a tempo. Maybe Larsen was already short of time and just wanted to play an easy move to get closer
to the time control.
Better was 32...Rb5! 33.Bxe7 Kxe7 after which the king would be one step closer to the centre: 34.Rc4 (34.Rd7+?!,
aiming to double the rooks on the seventh rank, cannot embarrass Black: 34...Ke8 35.e6 Rxc5+ 36.Kb1 fxe6 37.Rxg7
Rd5 – Black prevents the doubling and stands well) 34...Ke6 35.g3 (in case of 35.Rdd4 Kxe5 36.Rg4 Kf6 Black is
safe) 35...Rc7 36.Rd8 Kxe5 37.Re8+ Kd5 (37...Kf5?? 38.Rf4+ Kg5 39.Re5+ produces an unusual checkmate) 38.Rd8+
– the position is balanced, for if Black avoids the perpetual by returning to d7, Rg8 wins the g-pawn.

33.Rxh4 Rb5?!

Here the rook is passive and out of play. The text ties up the white rook for a while, but it will be out of the game shorter
than its adversary.

34.Rc4 Rc7 35.Rdd4 Re7

After 35...Rd7 Black gets into trouble in an unexpected way: 36.Kd3! Rxd4+ 37.Kxd4 Ke7
analysis diagram

38.Rc2! (it is easy to miss this subtle regrouping) 38...Rb7 39.Kc4! Ke6 40.Rd2 and suddenly Black is in trouble.

36.Re4 Kd7

Black has to do something, and centralizing the king seems natural. However, White’s two rooks are immensely
powerful and the black b5-rook is excluded from the battle in the centre. So Black needs to be very careful with his
plans there.
Karpov mentions the possibility of 36...Rd7 which aims to place the rook on d5: 37.Rxb4 Rxc5+ 38.Rbc4 Rdd5 and here
he stops, evaluating the position as plus/minus. This is worth looking deeper into: 39.b4:
A) In case of 39.Kc3? Rxe5 40.Rxe5+ Rxe5 41.Rxc6 Re2 Black is active;
B) After 39.Rxc5 Rxc5+ 40.Kd3 Ke7 41.b4 Rd5+ 42.Kc4 Rd2, why should Black have any disadvantage?;
C) 39.b4 Rxc4+ 40.Rxc4 and now:
analysis diagram

C1) 40...Rxe5 turns the game into a wild race that is hard to assess: 41.Rxc6 Re2+ 42.Kb3 Rxg2 43.Kxa3 Rg5
(43...Kd7 44.Rc3 g5) 44.Ka4 Kd7 45.b5 Rxh5 46.Rc2 Rh1;
C2) With 40...Kd7 Black can handle the position quietly without running the risk of losing: 41.Rg4 (in the event of
41.Re4 Ke6 42.Kb3 Rd3+ Black does all right as well) 41...Rxe5 42.Rxg7 Ke6 43.Kb3 Rxh5 – this is very likely to end
in a draw.

37.Kd3 Ke6 38.Red4

38...Kxe5?!
This makes life difficult for Black – maybe it is even a losing mistake. He allows the e7-rook to be exchanged.
Best was 38...Kf5!! with which the king looks for counterplay on the kingside:
A) 39.g4+ Kg5 40.Rd6 Re6 41.Rd7 Rxe5 42.Rxf7 Rd5+ 43.Rd4 Rbxc5 and Black’s rook is no longer passive;
B) 39.Rf4+ Ke6! (leaving the e5-pawn be. In case of 39...Kg5? 40.Rg4+ Kxh5? 41.Rxg7 Re6 (after 41...Rxe5 42.g3
Black gets checkmated) 42.g3 Rg6 43.g4+ Kh4 44.Rxf7 White has a solid advantage) 40.Rfd4 (40.g4?! Rd7+ 41.Rfd4
Rd5 – now Black will start pressing) 40...Kf5! is just a repetition;
C) 39.Rd6 Re6 (Karpov evaluates 39...Rxe5 40.g4+ Kg5 41.Rxc6 as plus/minus: after 41...g6 Black is living
dangerously, but it is not clear how White should continue) 40.Rd7 Rxe5 41.Rxf7+ Ke6 42.Rxg7 Rxh5 43.Rg6+ Kd7
44.Rd4+ Kc7 45.Rg7+ Kc8 and Black seems to draw.

39.Re4+ Kf6

40.Rf4+

Karpov probably did not want to make such an important decision at his last move before the time control, but 40.Rxe7!
was stronger. It is very hard, even with his special feeling and calculating ability, to assess positions in which both
players can push their pawns.
Karpov evaluates 40...Kxe7 41.Re4+ Kf6 42.Kc4 as plus/minus. He probably wrote his evaluation based on what he
was thinking during the game. To most people this position would be considered as slightly advantageous for White
even though he is a pawn down without immediate compensation. This correct assessment shows Karpov’s very subtle
ability to evaluate an endgame position.
We can look somewhat deeper: 42...Rb7 (42...Kg5 43.Re7!) 43.Rd4 Kg5 44.Rd6 Rc7 (the race after 44...Kxh5 45.Rxc6
Re7 46.Kxb4 Re2 47.Kxa3 Rxg2 is close, White is slightly ahead: 48.b4 f5 49.b5 f4 50.Rc8 Rg1 51.b6 Rb1 52.Rf8 g5,
and now 53.Ka4 wins) 45.Kxb4 Kxh5:
A) White even has time to take the a-pawn: 46.Kxa3 Kg4 47.b4 (47.Kb4 h5 48.a4 Kg3 49.Ka5 Kxg2) 47...Kg3
analysis diagram

48.b5! (in case of 48.Rd2 f5 49.Kb3 g5 50.a4 f4 51.Kc4 Ra7 Black is not slower) 48...cxb5 49.Kb4 Kxg2 (49...Ra7
50.Rd2) 50.c6 h5 51.Kxb5 Rc8 52.Kb6 Rb8+ 53.Ka7 Rb2 54.a4 Rc2 55.Kb7 wins;
B) After 46.Ka5 White is faster. He can win by quickly pushing his c-pawn: 46...g5 (in case of 46...Kg4 47.Kb6 Re7
48.Kxc6 Re2 49.Kb7 Rxa2 50.c6 Rb2 51.c7, or 46...Re7 47.Rxc6 Re2 48.Ra6 Rxa2 49.Kb6 Rxg2 50.c6 wins) 47.Kb6
(47.b4 Rc8 48.b5 cxb5 49.c6 f5 50.Kb6 b4 51.c7 also wins) 47...Rc8 48.Kb7 Re8 49.Kxc6 Re2 50.Kb7 Rxa2 51.Ra6
and White wins.

40...Kg5!
The king can be an attacker and a prey to be hunted down at the same time. After 40...Ke6, exchanging should be in
White’s best interests as the Black king is not active: Karpov gives the position after 41.Rce4+ Kd7 42.Kc4 a
plus/minus.

41.Rg4+ Kxh5

The black king’s predatory role has ended. Black has won an important pawn, but at too great a cost. His king will be in
great danger trapped on the h-file.
After 41...Kf5 42.Rcf4+ Ke6 43.Re4+ Kd7 44.Kc4, it’s tough going for Black.

42.Rxg7 Rbb7??

Black walks into the proverbial minefield. He allows a forced checkmate. Tragically for Larsen, he was actually still in
the game here. Not after 42...Re5? 43.g3! Rd5+ 44.Ke2 Re5+ 45.Kf3 Rf5+ 46.Kg2, when White checkmates. But
according to Karpov 42...Re1? would be plus/minus and Black would have drawing chances in a practical, over-the-
board fight. However, against a computer-aided analyst it still loses:
43.g3! Rh1 and now:
analysis diagram

A) 44.Rxf7 Rd1+ (44...Rh2 45.Rg7!) 45.Ke3 Re1+ 46.Kf2 Kg6 47.Rd7 Re6 and Black can live with this position;
B) 44.Ke3? Rh2! and the rook stays on the h-file;
C) 44.Ke2!!. The king wants to chase the rook away from h1. Black will pay heavily for having put his rook out on b5:
44...Rb8 (44...f5 45.Kf2 f4 46.Kg2) 45.Rf4!!. The b5-rook no longer attacks the c-pawn, so this rook can attack from a
different angle (45.Kf2 Rd8 46.Kg2 Rdd1 would give Black counterplay. But the alternative 45.Re4! f5 46.Rf4 wins as
well): 45...Re8+ (bringing the rook back in time to defend the f5-pawn) 46.Kf2 (but the king drifts towards the other
rook) 46...Re5 47.Kg2 Rg5 (Black still resists wittily) 48.g4+! (this last precise move seals Blacks fate) 48...Kh4
49.Rh7 and White wins.
The best continuation for Larsen was 42...Re6!! (Black must release the isolated king at once) 43.g4+ (after 43.g3 Rg6
44.g4+ Kh4 45.Rxf7 Kg3 Black also has play) 43...Kh4 44.g5+ Kh5 45.gxh6 Kxh6 46.Rxf7 Kg6 and Black can draw.
43.Rc1!

Larsen resigned, not waiting for the checkmate. Karpov has again waged a superb attack on the king in an endgame.
There could have followed: 43...Rbd7+ 44.Kc2 (if 44.Kc4?? Re4 mate) 44...Re2+ 45.Kb1 Rb2+ 46.Ka1 Rd4 47.Rh1+
Rh4 48.g4 mate.
This endgame was complicated and brimming with highly unusual motifs. Karpov orientated well in the various lines –
certainly much better than his opponent.

GAME 78
Romanishin, Oleg
Karpov, Anatoly
Moscow USSR Championship 1983 (8)

1.Nf3 Nf6 2.g3 d5 3.Bg2 c6 4.0-0 Bg4 5.c4 e6 6.Ne5 Bh5 7.d4 Nbd7 8.Nc3 Be7 9.cxd5 Nxe5 10.dxe5 Nxd5 11.Qc2
0-0 12.h3 Qa5 13.Nxd5 cxd5 14.g4 Bg6 15.Qb3 Rac8 16.Be3 Rc4 17.Rfc1 b5 18.a4 Rxc1+ 19.Rxc1 bxa4 20.Qb7 a3
21.Ra1 Bb4 22.bxa3 Bc3 23.Rd1 Qxa3 24.Qxa7 Qb3 25.Rc1 Bxe5 26.Qd7 h6 27.Bf3 Bf6 28.Rc8 Rxc8 29.Qxc8+
Kh7 30.Qc5 d4 31.Bf4 Qc3 32.Qd6 Qa5 33.Qc6 e5 34.Bg3 Qe1+ 35.Kg2 e4! 36.Bxe4 Qxe2 37.Bf3 Qe7 38.Qd5 d3
39.Bf4 Bg5 40.Bxg5 hxg5 41.Qa5 Qf6 42.Qb4 Qe5 43.Qd2
43...Be4!

The bishop on f3 is controlling the d1-square very effectively. It will be easier to push the d-pawn now.

44.Qe3

Black was threatening with 44...Bxf3 and then ...Qf4.

44...Bxf3+ 45.Qxf3

After 45.Kxf3 Qd5+ Black promotes the d-pawn.

45...Qb5!!

The queen both defends the pawn and covers the e8-square, beautifully preventing any perpetual without making any
concessions in the vicinity of his king.

46.Qe4+

After 46.Qxf7 d2 47.Qf3 Qb1 48.Qf7 Qe4+ wins.

46...Kh6 47.Qd4

The pawn advance 47.f4 should also lose as White cannot find access to Black’s king: 47...gxf4 48.Qxf4+ Kh7 49.Kf2
(49.Qxf7 d2 50.Qf3 Qc4!! 51.Qf5+ Kg8 52.Qf3 Qd4 53.Qd1 Qe3 and Black wins) 49...Qb2+ 50.Kf3 Qe2+ – this still
requires skill, but Black is winning.

47...Qc6+ 48.Kg3 Qc7+

Karpov plays it safe, but actually this move prolongs the game. Better was 48...Qc2! 49.Qd7 (49.Qe4 Qc7+ 50.Kg2
Qd7 51.Qf3 d2 52.Qd1 Qc6+ 53.f3 Qc1) 49...d2 50.Qxf7 Qd3+ 51.Kg2 (51.Kh2 Qd6+ 52.Kg2 Kh7!! 53.Qh5+ Qh6
and the d-pawn promotes) 51...Qe4+ (Karpov did not analyse this check) 52.f3 (52.Kg3 Qf4+) 52...Qe2+ 53.Kg3
Qe5+ 54.Kg2

analysis diagram

54...Kh7! (in a way this is a primitive move, stopping a mate in one; on the other hand it is subtle as Black had to assess
the ensuing position carefully to decide whether it is a win. And Karpov really needed to win this game) 55.Qb3 Qe2+
56.Kg3 d1=Q and Black wins.

49.Kg2 Qc2 50.Qe4

In the event of 50.Qd5, 50...d2 51.Qxf7 Qc6+ 52.Kg3 Kh7! 53.Qf5+ Qg6 54.Qd5 Qc2 wins as Karpov points out.
50...Qc3 51.Qd5

Computer programs have since discovered that queen endings with an extra h- and g-pawn often do not win. Had
Romanishin known this, he might have opted for 51.h4!? gxh4 52.Qf5 and now:
A) 52...h3+!?;
B) 52...Qc6+ 53.f3 Qg6 54.Qd5 f5 55.Qxd3 fxg4 56.Qe3+ Qg5 (56...Kh7 57.fxg4 Qxg4+ 58.Kh1 and White can
hold) 57.Qe6+ g6 58.fxg4 Qf4 59.g5+ Kh5 (all endings after 59...Kxg5 or 59...Qxg5+ are drawn, see below) 60.Kh1
Qf3+ 61.Kh2 Qg3+ (61...Kxg5 draws) 62.Kh1 Qxg5 (62...h3 63.Qxg6+!).

analysis diagram

Even if Black plays the best moves he cannot win. Would you believe it? White can draw with nine different moves!
C) 52...g6 53.Qxf7 Qc6+ 54.f3 (54.Kh2 Kg5!) 54...Kg5 (in case of 54...Qd6 55.f4 d2 56.Kh3 g5 57.Qh5+ Kg7
58.Qxg5+ Kf7 59.Qf5+ White probably draws) 55.Kh3 Qd6 56.f4+ Qxf4 57.Qd5+ (in the event of 57.Qe7+ Kh6
58.Qxh4+ Kg7 59.Qe7+ Qf7 wins according to Karpov. Soon computers will be able to tell for sure whether this
evaluation is correct or not) 57...Kh6 58.Qxd3 and, knowing all these drawn positions with two extra pawns, this can
probably be held as well.

51...d2 52.Qxf7 Qc6+ 53.f3

If 53.Kh2 Kh7! – Karpov’s win is pretty nice: 54.Qf5+ Qg6 55.Qd7 Qc2 56.Qf7 (56.Qe8 Qc7+ 57.Kg2 Qd6) 56...Qc8
57.Qd5 Qc7+ 58.Kg2 Qc2 59.Qf7 Qc6+.
The 53...Qd6 check delivers the win too: 54.Kg2 Kh7 55.Qf5+ g6! (55...Qg6 56.Qd5 Qc2 57.Qf7 Qe4+ also wins)
56.Qf7+ Kh8 57.Qe8+ Kg7 and White runs out of checks.
53...Kh7! 54.Qb3 Qd6 55.Qc2+ Kh6

Black has nicely forced the enemy queen into a passive position.

56.Qd1

56...Qd3!

Making sure the queen is fixed on d1 and that his own king can cross the e-file.

57.Kf2 Kg6 58.Kg2 Kf7 59.Kf2 Kf8

White is in zugzwang.

60.Kg2 Ke7 61.Kf2 Kd6 62.Kg2 Kc5 0-1

GAME 79
Karpov, Anatoly
Polugaevsky, Lev
Tilburg 1983 (2)

1.Nf3 Nf6 2.c4 c5 3.Nc3 Nc6 4.e3 e6 5.d4 d5 6.cxd5 exd5 7.Be2 cxd4 8.Nxd4 Bd6 9.0-0 0-0 10.Bf3 Be5 11.Qd3
Nb4 12.Qd2 Bxd4 13.exd4 Bf5 14.Bd1 Re8 15.Na4 Nc6 16.f3 Qa5 17.Qxa5 Nxa5 18.Kf2 Nc6 19.Be3 Nb4

20.Bg5

One feels that this position must be equal. It is worth examining how Karpov gradually takes over the initiative in this
transitional phase of the game.

20...Nd7

After 20...Nd3+ 21.Kg1 (in the event of 21.Kg3 Nh5+ 22.Kh4 Nhf4 Black’s pieces are rather active) 21...Nd7 22.Bb3
h6 23.Bh4 g5 24.Bg3 Nf6 Black is all right.

21.g4 Bg6

With 21...Bd3 Black could ‘kill’ the game: 22.Re1 Rxe1 (after 22...Nc2 23.Bxc2 Bxc2 24.Nc3 Nb6 25.Rac1 White is
somewhat better) 23.Kxe1 Bc2 and Black should be safe.

22.Bb3
22...Bc2?!

After this Black becomes slightly underdeveloped. After 22...Bd3 23.Rfc1 b5 24.Nc3 Bc4 25.Bc2 Bd3 26.Bd1 Bc4 he
would have been fine.

23.Bxc2 Nxc2 24.Rad1 f6

In case of 24...Nf8 25.Rd2 Nb4 26.Nc3 Ne6 27.Be3 Rad8 28.f4 White has something to build on.

25.Bf4 Nf8

25...g5!? (Karpov has made some pawn moves similar to this. The pawns find themselves on the bishop’s colour but
they restrict its movement) 26.Bc1 Rac8 (or 26...Nb4 27.Nc3 Nb6 28.f4 h6 29.Kg3 Rac8 and Black has a healthy
endgame) 27.Nc3 (27.a3 Rc4) 27...Nb6 28.f4 h6 29.Nb5 Re6 and Black is fully in the game.
26.Rd2! Ne6

In case of 26...Nb4 27.Nc3 Rad8 28.Nb5 Rd7 29.Nxa7 Ng6 30.Bg3 Nxa2 White’s minor pieces cause problems on the
dark squares.

27.Bg3 Ncxd4!

It is hard to tell whether Polugaevsky had planned the ensuing endgame, hoping he would be able to hold it, or perhaps
he had misjudged his chances and simply had to settle for it. At this point, the text is the right decision.
The 27...Nb4 retreat would be no fun for Black. 28.Nc3 Rad8 (or 28...Red8 29.Re1 Kf7 30.Nb5, causing trouble)
29.Nb5 Nc6 (29...a6 30.Nd6 Re7 31.Re1) 30.Nd6 Re7 31.Re1 and Black is under pressure.

28.Rfd1 Nc6 29.Rxd5 Rad8

Polugaevsky defended well from this point on, but on move 52 he blundered and eventually lost.

30.Rxd8 Rxd8 31.Rxd8+ Ncxd8 32.Ke3 Kf7 33.f4 g6 34.f5 gxf5 35.gxf5 Ng7 36.Ke4 Nc6 37.Bd6 Ne7 38.Nc5 b6
39.Na6 Ngxf5 40.Bb8 Ke6 41.Bxa7 Nd6+ 42.Kd3 Nd5 43.a4 f5 44.b4 f4 45.a5 bxa5 46.bxa5 Kd7 47.Nc5+ Kc6
48.Nb3 Nb4+ 49.Ke2 Kb5 50.Kf3 Nc6 51.Bb6 Nc4 52.Bc7

52...N4xa5? 53.Nxa5 Nxa5?

53...Nd4+ is still a draw.

54.Bxa5 Kxa5 55.Kxf4 Kb5 56.Kg5 Kc5 57.Kh6 1-0

First Two Kasparov Matches (1984-1985)

Three years had passed since Karpov’s last title-defence match against Kortchnoi, so he had to put his crown at stake
again. His opponent was a worthy one. Kasparov had gone through the qualifiers and was already stronger than any of
Karpov’s previous opponents.
In the match, which started in 1984, Karpov took a convincing 5-1 lead and was comfortable in the endings. Before the
match, however, by way of preparation, he had already sharpened up his play in tournament games. He had probably
realized he would have to confront his rival in such situations. The amount of sharp games would increase in their later
matches. On the other hand, Kasparov also learned from Karpov and improved his endgame.
Statistics show that Kasparov has won many endgames against Karpov, but usually he had obtained a winning ending
from the middlegame. Kasparov never really outplayed Karpov in an endgame, except maybe for the last Sevilla game
in 1987.
In the first match, Karpov again ran out of energy when Kasparov stabilized and then won his second game. After a long
series of draws, the latter won another game, and a third. Finally FIDE President Florencio Campomanes stopped the
match when the score was +5 =40 -3.
In the second match, Karpov was to retain the title with a 12-12 score. After a loss in the first game, Karpov broke even
in the fourth and took the lead in the fifth game, but his opponent took over in the 16th game. Karpov had trouble in the
openings in this match. In Game 20 he kept playing on in an equal endgame, but without success. In the 22nd game
(here presented as Game 85), he reduced Kasparov’s lead to one point and in the final game Karpov attacked, but
Kasparov found an excellent defence, winning the game and the title.
Earlier in 1984, London hosted a match USSR-Rest of the World.

GAME 80
Karpov, Anatoly
Andersson, Ulf
London USSR vs ROW 1984 (1.1)

1.d4 Nf6 2.c4 e6 3.Nf3 Bb4+ 4.Bd2 Bxd2+ 5.Qxd2 0-0 6.Nc3 d5 7.e3 Nbd7 8.cxd5 exd5 9.Bd3 Re8 10.0-0 Ne4
11.Qc2 Ndf6 12.b4 c6 13.Ne5 Bf5 14.Na4 g6 15.Qb2 a6 16.Rfc1 Re7 17.Nc5 Nxc5 18.bxc5 Bxd3 19.Nxd3 Rc8
20.Rc3 Rcc7 21.Rb3 Ne8 22.Qe2 f6 23.Qf3 Rf7 24.Kf1 Ng7 25.Ke2 Rce7 26.Kd1 Qc8 27.Rab1 h5 28.h3 Ne6 29.h4
Kh7 30.Qh3 Qe8 31.Kc2 Rd7 32.Kb2 Ng7 33.Nf4 Rfe7 34.Ka1 Qf7 35.Rg1 Ne6 36.Nd3 Ng7 37.g4 hxg4 38.Rxg4
Nh5 39.Rb1 Qe6 40.Qf3 Rg7 41.Rbg1 Rde7 42.Kb2 Kh6 43.Kc3 Qf7 44.Nf4 Nxf4 45.Rxf4 Re6 46.Kd2 Qe7
47.Ke2 Kh7 48.Kf1 Kh6 49.Rg3 Kh7 50.Rfg4 Qf7 51.Rf4 Kh6 52.Kg1 Kh7 53.Kh2 Kh6 54.Qg2 Kh7 55.Kg1 Re8
56.Qf3 Rf8 57.Kf1 Qe7 58.Qd1 Qe8 59.Qb1 Kh6 60.Ke2 Qd8 61.Rfg4 Rfg8 62.Kf1 Qe8 63.Qd1 Qe6 64.Qf3 Rf7
65.Kg1 Rfg7 66.a3 Re7 67.Kh2 Rf7 68.Rf4 Kh7 69.Qd1 Kh6 70.Qd3
70...Qe8?!

The Swedish grandmaster moves the queen away from the centre. Black has nothing to do except wait. But Karpov
suggests a better way of doing nothing: his recommendation is 70...Rff8, establishing a connection between the rooks.
After 71.e4 dxe4 72.Rxe4 Qd5! the black queen is centralized and he has decent chances to defend.

71.e4!

After manoeuvring for so long Karpov opens up the position.

71...dxe4

After 71...Qd7 72.exd5 Qxd5 73.h5! g5 74.Rf5 Qe6 75.Re3 Qd7 76.d5 cxd5 77.Rxd5, White has created some pressure.

72.Rxe4 Qd7 73.Qe3+ Kh7

74.Re6!

Karpov has made sure he controls the only open file. Now he starts exerting pressure on the sixth rank.

74...Rgg7

On 74...Qd5 75.Rd6 Qh5 76.Rxf6 White wins a pawn.

75.Rf3 f5?!

Black had more chances to keep his position together with 75...Qd8, keeping the 6th rank closed.

76.h5 gxh5 77.Qh6+

White also keeps promising prospects with 77.Rh3. Then, after 77...Kg8 78.Rxh5 Re7! (78...Rf8, Karpov’s suggestion,
simply loses to 79.Reh6) 79.Rxe7 Qxe7 80.Qf4 Black’s position is tough but not hopeless.

77...Kg8 78.Rfe3 Qc7+

Playing 78...Re7! at once was reasonable: 79.Rxe7 Rxe7 80.Qg6+ Kh8! 81.Rh3!? and although Black is living
dangerously, he may survive.
If 81.Re5 Qxd4 (in case of 81...Rf7 82.f4 Black can hardly move) 82.Qxh5+ Rh7 83.Re8+ Kg7 84.Re7+ Kf6, according
to Karpov, Black has a perpetual.

79.Kh3 Re7??

This is a bad mistake in a difficult position. Better was 79...Qd7! 80.Re8+ Rf8 81.Rxf8+ Kxf8 82.Qh8+ Kf7 83.Qxh5+
Kf6 84.Qh4+ Kf7 when Black’s position is unattractive, but unlikely to be lost.

80.Rxe7!
Karpov doesn’t miss a forced win like this.

80...Rxe7 81.Qg6+ Kf8 82.Qf6+ Ke8 83.Qh8+ Kd7 84.Rxe7+ Kxe7 85.Qg7+ Kd8

86.Qf8+ 1-0

Karpov will take the f-pawn with check before he swaps queens.
Exchanging at once with 86.Qxc7+ would also have won: 86...Kxc7 87.Kh4 Kd7 88.Kxh5 Ke6 89.Kg5 Kd5 90.Kxf5
Kxd4 91.f4 Kxc5 92.Ke5 when White promotes with check.

GAME 81
Kasparov, Garry
Karpov, Anatoly
Moscow World Championship 1984 (6)

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.b3 Bb4+ 6.Bd2 Be7 7.Bg2 0-0 8.0-0 d5 9.Ne5 c6 10.Bc3 Nfd7 11.Nxd7 Nxd7
12.Nd2 Rc8 13.e4 b5 14.Re1 dxc4 15.bxc4 Nb6 16.cxb5 cxb5 17.Rc1 Ba3 18.Rc2 Na4 19.Ba1 Rxc2 20.Qxc2 Qa5
21.Qd1 Rc8 22.Nb3 Qb4 23.d5 exd5 24.exd5 Nc3 25.Qd4 Qxd4 26.Nxd4 Nxa2 27.Nc6 Bc5 28.Bh3 Ra8 29.Bd4
Bxd4 30.Nxd4
30...Kf8

Kasparov has a dangerous passed pawn, so Karpov’s king moves closer.

31.d6 Nc3!?

Karpov wants to slog it out.

31...Rd8 would have eliminated all play: after 32.d7 Bb7 33.Nxb5 Bc6 34.Nxa7 Bxd7 the disappearance of the
queenside pawns makes sure the game is drawn.

32.Nc6?

This move gives away a free tempo. Kasparov must have missed something along the way.
A) White could have forced a draw by winning the b5-pawn with 32.Bg2. There follows 32...Rd8 33.Bc6 Bc8 34.Nxb5
and there is very little left to play for;
B) White could have won a piece for two pawns and entered an interesting endgame with 32.d7 Bb7 33.Nf5 Rd8
34.Re8+ Rxe8 35.dxe8=Q+ Kxe8 36.Nd6+ Kd8 37.Nxb7+ Kc7. Black should not be worse here, as his king can help
his passed pawns;
C) Other commentators do not mention 32.Re3, which is an interesting zwischenzug, e.g. 32...b4 33.Nc6 Bb5 34.Bg2
(34.Nxb4 Na4) and it looks like Black is in big trouble, but he has a tactical rescue: 34...Ne2+ 35.Kh1
analysis diagram

35...b3! 36.Rxb3 Bxc6 37.Bxc6 Nd4 and Black manages to simplify to a draw.

32...Bb7!

Karpov is suddenly able to take over.

After 32...b4 33.d7 b3 34.d8=Q+ Rxd8 35.Nxd8 Bd3 White must soon give up his extra rook and settle for a draw.

33.Bg2 Re8!

Karpov diverts the strong white rook.

34.Ne5

According to Yusupov 34.Ra1 gives better chances of survival: 34...Bxc6 35.Bxc6 Re6 36.Rxa7 Rxd6 and White is
merely a pawn down.

34...f6

This is an ‘only’ move, but it wins.

35.d7

Kasparov offered a draw here, but his position is lost.

After 35.Nd7+, 35...Kf7 36.Ra1 Bxg2 37.Kxg2 Ke6 wins according to Yusupov. Alternatively, 35.Bxb7 Rxe5 36.Ra1
b4 37.Rxa7 b3 also wins for Black.

35...Rd8 36.Bxb7 fxe5 37.Bc6

37...Ke7?

Karpov blunders in time-trouble.

Best was 37...e4! 38.Ra1 Ke7! 39.Rxa7 Kd6 40.Ra6 Kc7 and Black wins by pushing his b-pawn.

38.Bxb5! Nxb5

In time-trouble Karpov goes for the simplest continuation.

In the event of 38...Kd6 39.Bd3 Rxd7! 40.Bxh7 a5 Black is somewhat better; Yusupov prefers this idea to the game
continuation.

39.Rxe5+ Kxd7 40.Rxb5

40...Kc6

Not 40...Ra8?. This being the last move before the time control, it would have been very risky, even mistaken, to give
both kingside pawns: 41.Rb7+ Kc6 42.Rxg7 a5 43.Rxh7 a4 44.Re7 a3 45.Re1 a2 46.Ra1 Kd5 47.h4 Kc4 48.Kg2 Kb3
49.g4 Kb2 50.Re1 a1=Q 51.Rxa1 and White holds.

41.Rh5?!

In rook endings, when the time comes to harvest the pawns it is better if the opponent has his pawns on the same rank,
then the rook can reap them easily. Kasparov might have hoped to take the h-pawn on the sixth rank with a check,
otherwise it is hard to find an explanation for this move.
Timman gives an only move mark to 41.Re5 in New In Chess, and Yusupov recommends the same idea. The latter
claims that 41...Ra8 42.Re6+ (42.Kf1 a5 43.Ke2 a4 44.Kd3 a3 45.Re1 a2 46.Ra1 Kd5 wins for Black) 42...Kc5 43.Re7
a5 44.Rxg7 is slightly better for White, but it looks more like a losing position: 44...a4 45.Rxh7 (45.Rc7+ Kb4 46.Kg2
(46.Rb7+ Ka5 47.Rxh7 a3 48.Re7 a2 49.Re1 Rb8 and Black wins) 46...a3 47.Rc1 a2 48.Ra1 Kc3 49.f4 Kb2 and Black
wins) 45...a3 46.Rc7+ Kb4 47.Rb7+ Ka4 48.Rb1 a2 49.Ra1 Kb3 50.h4 Kb2 51.Re1 a1=Q 52.Rxa1 Rxa1+. With the
help of modern tools we can prove that the position is winning for Black (53.Kg2 Kc3) but just by one tempo. If it
were White’s turn now he would be able to draw.

41...h6

Black must keep his h-pawn.

42.Re5 Ra8!

Yusupov gives 42...Rd5!? as a winning continuation. It looks like it is, very narrowly. If White misses one tempo, it is a
draw. At this moment Karpov sealed his move. For him it would have been impractical to go for such a narrow path
with a close race. Even with his exceptional calculating ability it is possible to miss an intermediate check or a finesse.
Remembering his adjournment against Petrosian, one can easily imagine the Karpov team had analysed the position all
the way to the end:
43.Re7 a5 44.Rxg7 a4 45.Ra7 (45.Rg6+ Kb5 46.Rxh6 a3 47.Rh8 Ka4 and Black wins) 45...Kb5 46.Kg2 Kb4 (46...Rd6
47.h4) 47.h4 a3 48.g4 Ra5 49.Rb7+ Kc3 50.Rb1 a2 51.Ra1 Kb2 52.Re1 a1=Q 53.Rxa1 and now:
A) 53...Kxa1 54.f4 (54.Kf3 Ra4) 54...Ra3! 55.g5 (now if 55.h5 Kb2 56.g5 Kc3 57.Kf3 Kd4+ 58.Kg4 hxg5 59.fxg5
(59.Kxg5 Kd5) 59...Ke5, Black wins) 55...h5 56.f5 (56.g6 Ra6 57.f5 Ra4 wins) 56...Ra5 57.g6 Ra4 wins;
B) 53...Rxa1!? 54.Kf3 (54.g5 hxg5 55.hxg5 Kc3 loses).

analysis diagram

Now not 54...Kc3 55.Ke4 Ra4+ 56.Kf5 Kd3 57.f4 (57.h5 loses to 57...Ra5+ 58.Kg6 Rg5+ 59.Kxh6 Rxg4) 57...Ke3
58.g5 Rxf4+ 59.Kg6 hxg5 60.Kxg5 and White draws, but 54...Ra4! wins.

43.Ra5

The rook becomes very passive, but Garry had no choice.

After 43.Re6+ Kc5 (Timman gives 43...Kb5. His line transposes to Yusupov’s, see 44.Rg6 a5 45.Rxg7 a4 46.Rc7+ Kb4
47.Rb7+ Ka5 48.g4 a3 49.Rb1 a2 50.Ra1 Kb4 51.f4 Kb3:
A) 52.g5 hxg5 53.fxg5 Ra5! 54.Kf2 (Yusupov doesn’t mention this move. It gives some practical chances, whereas
54.h4 Ra4 55.Kg2 Rxh4 loses) 54...Rxg5 55.Kf3 Rh5! (the only move to win) 56.Kg3 Kb2;
B) 52.Kf2 Kb2 53.Rxa2+ (after 53.Re1 a1=Q 54.Rxa1 Rxa1 55.Ke3 Kc3 56.h4 Rh1 57.g5 h5 Black wins according to
Timman) 53...Rxa2 54.Ke3 Kc3 55.h4 Rh2 56.h5 Rh4 57.g5 Rxh5 58.g6
analysis diagram

58...Rh4! 59.f5 (59.Kf3 h5) 59...Rg4 wins, as Yusupov analyses.

43...Kb6 44.Ra2

Garry still has hope, if he can manoeuvre his king under the a-pawn and his rook can defend the kingside. This rook
endgame is reminiscent of the very last game of the Capablanca-Alekhine 1927 World Championship match in Buenos
Aires.
After 44.Ra1 a5 45.Kf1 a4 46.Ke2 a3 47.Kd3 Kc5 48.Kc3 (48.Kc2 Kc4!) 48...Kd5! (getting to the kingside pawns
quickly is vital; in case of 48...a2?! 49.Kb2 Kb4 50.f4! h5 51.Rxa2! Rxa2+ 52.Kxa2 Kc3 53.f5! Kd3 54.Kb3 Ke4
55.Kc4 Kxf5 56.Kd4 Kg4 57.Ke5 White’s king arrives in time) 49.Kb3 a2 50.Kb2 Ke4 51.Re1+ Kf3 Black wins.

44...a5 45.Kf1
45...a4

Yusupov recommends 45...Re8!? to cut off the king. This looks like a winning idea:
A) 46.Ra1 Kb5 47.Rb1+ Kc4 48.Ra1 Ra8 49.Ke2 a4 50.Kd2 a3 51.Kc2 a2 52.Kb2 Kd3 and Black wins;
B) 46.f4 Kb5 47.Rb2+ Kc4 48.Rb7 Ra8 49.Rxg7 a4 50.Rc7+ Kb5 51.Rb7+ Ka5 52.g4 a3 53.Rb1 Ka4 54.Kf2 a2
55.Ra1 Kb3 56.Ke3 Kb2 57.Re1 a1=Q 58.Rxa1 Rxa1 59.h4 and now:
B1) Not 59...Ra4? 60.h5 (or 60.g5! Ra3+ (60...h5 61.g6) 61.Ke4 Rh3 62.gxh6 and White is safe) 60...Kc3 (60...Ra5
61.Ke4 Kc3 62.f5) 61.g5 hxg5 62.fxg5 Rg4?? (62...Rh4! draws) 63.g6 – this would be a horror scenario for Black, as
he is losing now;
B2) Better is 59...Rh1! and now:
B21) 60.h5 Rh4 (in case of 60...Kc3 61.Ke4 Rh4 62.Kf5 Kd4 63.g5 Rxh5 64.Kg6 hxg5 65.fxg5 Rh1 66.Kf6 White
holds) 61.g5 Rxh5 (61...hxg5 62.fxg5 Rxh5 63.Kf4 Kc3 64.g6 is a draw) 62.gxh6 (62.g6 Rh4)
analysis diagram

Now, instead of the obvious 62...Rxh6?, 62...Kc3!! wins. It is remarkable that if the rook were on h4 it would be a draw;
B22) 60.g5 h5! (keeping the h-pawn, which can protect the rook on g4, is the right way to play. 60...Rxh4 61.gxh6
would be a draw) 61.f5 Rh3+! (this intermediate check wins the vital tempo and ensures the win) 62.Ke2 (62.Ke4
Rxh4+ 63.Ke5 loses to 63...Rg4 64.g6 h4 65.Kf6 (65.f6 Rg5+!) 65...h3 66.g7 h2 67.Kf7 h1=Q) 62...Rxh4 63.f6 Rf4
64.Ke3 Rf1! 65.Ke4 h4 66.Ke5 h3 67.g6 h2 68.g7 Rg1 and Black wins.

46.Ke2 Kc5 47.Kd2 a3 48.Kc1 Kd4 49.f4

After 49.Kb1 Rb8+ 50.Ka1 Rb2 51.Rxa3 Rxf2 52.Ra6 Rf6! 53.Ra7 g5 Black wins.

49...Ke4 50.Kb1 Rb8+ 51.Ka1 Rb2 52.Rxa3 Rxh2 53.Kb1

If 53.Ra7 g6 54.Ra6 g5 55.Rg6 (55.Kb1 Kf3 56.fxg5 hxg5) 55...Kf5 56.Rg8 Rg2 Black wins.

53...Rd2!

Black threatens 54.Rd3.

After 53...g5 54.Kc1 gxf4 55.gxf4 Kxf4 56.Kd1, the white king is close enough to draw.

54.Ra6

Moving one square further with 54.Ra7 g5 55.Rh7 loses to 55...g4 56.Rxh6 Rg2 57.Re6+ Kf3 58.f5 Kxg3.

54...Kf5 55.Ra7 g5

When the king is cut off, the defender has little chance against a g-pawn.

56.Ra6
56...g4!

Karpov plays with a subtle touch.

However, the prosaic 56...Rh2! would also have won: 57.Kc1 Kg4 58.Rg6! Rh3 59.Kd2 Rxg3 60.Rxh6 Kxf4! –
commentators at the time thought this would not work for Black, but it does.

57.Rxh6 Rg2 58.Rh5+

If 58.Kc1 Rxg3 59.Kd2 Rf3 60.Ke2 Kxf4! 61.Rf6+ Kg3 wins.

58...Ke4 59.f5 Rf2!

This is not the only way to win. 59...Rxg3 60.f6 Rf3 61.Rh4 Kf4 62.f7 (62.Kc2 Kg5) 62...Kg3 does the job as well.

60.Kc1 Kf3 61.Kd1

Kasparov could have set up more obstacles with 61.f6!?. However, Karpov would surely have handled the position
properly: 61...Kxg3 62.Rh6 Kg2! (the only move to win. Yusupov’s analysis of 62...Rf3? stands the test of modern
computers: 63.Kd2 Kf2 64.Rh2+ Kg1 65.Rh6=) 63.Kd1 g3! and now:
A) 64.Ke1 Kg1 – remarkably, if Black were to move in this position, only ...Kg2 would win. 65.Rg6 g2 (a mutual
zugzwang position) 66.Rh6 Rf5 67.Ke2 Re5+ 68.Kf3 Kf1 69.Rg6 Re6! wins;
B) 64.Rg6 Rf5 65.Ke2 Re5+! 66.Kd3 Kf3! 67.Kd4 (67.f7 Rf5 68.Rg7 Rf4!)
analysis diagram

67...Rh5!! – nicely losing a tempo to reach the mutual zugzwang position with the opponent to move. 68.f7 Rf5 69.Rg7
g2 (another mutual zugzwang) 70.Kd3 (70.Kc4 Kf2) 70...Rf4 – Yusupov mentions that this lovely piece of analysis
stems from grandmaster Sergey Dolmatov.

61...Kxg3 62.Ke1 Kg2

Not difficult to find, nevertheless this is the only winning move.

63.Rg5 g3 64.Rh5 Rf4 65.Ke2

65...Re4+!

Another ‘only’ move.

66.Kd3 Kf3 67.Rh1 g2 68.Rh3+ Kg4 69.Rh8 Rf4 70.Ke2 Rxf5 0-1
 GAME 82
Karpov, Anatoly
Kasparov, Garry
Moscow World Championship 1984 (9)

1.d4 d5 2.c4 e6 3.Nf3 c5 4.cxd5 exd5 5.g3 Nf6 6.Bg2 Be7 7.0-0 0-0 8.Nc3 Nc6 9.Bg5 cxd4 10.Nxd4 h6 11.Be3 Re8
12.Qb3 Na5 13.Qc2 Bg4 14.Nf5 Rc8 15.Bd4 Bc5 16.Bxc5 Rxc5 17.Ne3 Be6 18.Rad1 Qc8 19.Qa4 Rd8 20.Rd3 a6
21.Rfd1 Nc4 22.Nxc4 Rxc4 23.Qa5 Rc5 24.Qb6 Rd7 25.Rd4

25...Qc7

Garry seems to be impatient. At this stage, he was yet to win his first game. Something like that had probably never
happened to him before. He would like to play for a win; instead he has to defend a boring position.
25...Rc6 26.Qb3 Rc4 was also possible.

26.Qxc7 Rdxc7
27.h3

Karpov makes a little improving move.

White could not win a pawn with 27.Nxd5. After 27...Nxd5 28.Bxd5 Bxd5 29.Rxd5 Rxd5 30.Rxd5 Rc2 Black has
nothing to worry about.

27...h5

Stopping g3-g4.

28.a3 g6

If there were no other pieces than the white knight and the black bishop, then a good recommendation would be to put
the black pawns on h5, g7, f6, b6 and a5, just as Capablanca managed against Flohr in their famous 1935 game in
Moscow. Garry doesn’t follow Capablanca’s set-up.

29.e3 Kg7?!

This is a slight inaccuracy, as the king later goes to f8, losing a tempo. However, one tempo doesn’t matter much in this
position.

30.Kh2

Karpov doesn’t follow the standard patterns by centralizing his king. He starts a triangulation, not to lose a tempo but to
improve his bishop and king.

30...Rc4 31.Bf3 b5

Garry looks for an initiative on the queenside, running the risk that the position may get blocked. With so many black
pawns on the colour of Karpov’s bishop there is some cause for concern. This move has received harsh criticism from
some commentators, yet others did not believe it was a mistake. I tend to go along with the latter assessment.
32.Kg2

This is his last improving move for the time being. However, it is not the last time in the game that Karpov uses the g2-
square.

32...R7c5

33.Rxc4 Rxc4

Probably Kasparov was already shaken by the three losses he had suffered. He normally provokes dynamism like
nobody else, but this time he goes for the static option.
Black could have neutralized White’s play on the queenside by tactical means with 33...dxc4!? (even 33...bxc4 would be
possible). After 34.Rd6 a5 35.Rb6 (or 35.Na2 Nd5 36.Nc3 (in case of 36.e4 Black takes over with 36...c3! 37.b4 Nxb4
38.axb4 Bxa2 39.bxc5 c2) 36...Nxc3 37.bxc3 b4 and Black draws easily) 35...Nd7! 36.Rxb5 (in the event of 36.Rb7
Ne5 37.Ne4 Bd5 Black would be very active) 36...Rxb5 37.Nxb5 Nc5 38.Nc7 Na4 and Black is not worse.

34.Rd4

Karpov stops ...b5-b4.

34...Kf8 35.Be2 Rxd4

Kasparov could have replaced the rook somewhere on the c-file. Both options are slightly worse for him.

36.exd4 Ke7

All black pawns are on the colour of Karpov’s bishop – not a particularly happy arrangement. However, the position is
rather closed, so Black should be able to stop all penetration attempts.
Geller recommends 36...Ne4!? – a move in the style of Karpov, who usually starts endgames by improving his pieces
before bringing out the king. Geller evaluates the endgame after 37.Na2 Nd6 as equal.
37.Na2 Bc8 38.Nb4 Kd6 39.f3 Ng8 40.h4 Nh6 41.Kf2 Nf5 42.Nc2

42...f6?!

Geller doesn’t like Kasparov’s sealed move. It weakens the pawn structure slightly. However, Efim Petrovich still gives
the position a plus/equal evaluation. One alternative was 42...Bd7!?.

43.Bd3 g5

Keeping the knight on with 43...Ne7 was also reasonable.

44.Bxf5 Bxf5 45.Ne3

We have arrived at one of the most extensively analysed endgames ever. Many excellent commentators have worked on
this position. To name a few: Averbakh with Taimanov, Dvoretsky, King, Marin, Mikhalchishin and Gipslis, Karsten
Müller, Timman, and Tisdall.

45...Bb1

Mikhalchishin analysed 45...Ke6!, bringing the king to the part of the board where White intends to play. After 46.b4
Bg6 47.g4 hxg4 48.hxg5 (48.Nxg4 gxh4 49.Kg2 Bf5 50.Ne3 Bd3 51.Kh3 Be2! and Black prevents all penetrations)
48...gxf3 49.gxf6 Be4 50.Ng4 Kf7 51.Kg3 Ke6 52.Ne5 Kxf6 53.Nd7+ Kf5 54.Nc5 f2 55.Kxf2 Kf4 56.Nxa6 Bd3
57.Nc5 Bc4 White is in zugzwang, therefore Black is not worse.

46.b4
46...gxh4?

Kasparov and his team did not anticipate one of the most amazing moves in the history of over-the-board endgames.
A) Taimanov and Averbakh analysed 46...Bg6 first, and now:
A1) 47.hxg5 fxg5 48.f4 g4 (48...gxf4 49.gxf4 Ke6 50.Kg3 Kf7 51.Kh4 Kf6 52.f5 Be8 53.Nxd5+ Kxf5 54.Nc7 Bf7
55.Nxa6 Ke4= King) 49.f5 Bh7 50.Ng2 Ke7 (King) 51.Nf4 Kf6=;
A2) 47.g4! Ke6 48.Kg3 Bf7 49.hxg5 (in case of 49.f4 gxh4+ 50.Kxh4 hxg4 51.Kxg4 Be8 White’s invasion is
prevented) 49...fxg5 50.gxh5 Bxh5 51.f4 gxf4+ 52.Kxf4

analysis diagram
52...Kf6!! (stopping White’s invasion on the kingside) 53.Nxd5+ Ke6 54.Nc7+ Kd6 55.Nxa6 and White cannot win
despite the extra material as Dvoretsky points out.
B) 46...Ke6! 47.g4 (47.hxg5 fxg5 48.f4 gxf4 49.gxf4 Bg6 50.Kg3 Bf7 51.Kh4 Kf6=) 47...hxg4 48.hxg5 (after
48.Nxg4 gxh4 49.Kg2 Kf5 50.Ne3+ Kg5 51.Nxd5 Bf5 Black is doing all right) 48...gxf3 (Nunn) 49.gxf6 (49.Kxf3
fxg5 50.Kg4 Kf6 51.Nxd5+ Kg6=) 49...Be4 50.Ng4 Kf7 51.Kg3 Ke6 52.Kf4.
It looks like Black is in zugzwang but he can still move: 52...Kf7 53.Ke5

analysis diagram

53...Bg6! (thus, Black can sacrifice the d5-pawn under circumstances when he can hold) 54.Kxd5 Bh5 55.Nf2 Kxf6
and now:
B1) If White starts the race with 56.Kd6, Black just manages to draw: 56...Kf5 57.d5 Kf4 58.Kc7 Ke3 59.d6 Be8
60.Ng4+ Kd4! (Black must go after White’s queenside pawns) 61.d7 Bxd7 62.Kxd7 Kc3 63.Kc6 Kb3 64.Kb6 Kxa3
65.Ka5 Kb3 66.Nf2 Kc4 (66...Ka3?? 67.Ne4) with a draw, as White cannot make any progress;
B2) 56.Ke4 (trying to win the f3-pawn) 56...Ke6 57.Nh3 Kd6 58.Ke3 Kc6 59.Nf4 Bg4 60.Kf2 Kd6 61.Kg3
analysis diagram

61...Kc6!=. Thanks to the fact that the f3-pawn is close to promotion, Black holds this endgame as Dvoretsky has
pointed out on the Russian website www.e3e5.com.

47.Ng2!!

This is an absolutely marvellous move, one of the few that have become a part of chess culture! There was a rumour that
it was Dolmatov who found this fantastic idea after the game was adjourned, and this precious information had
somehow reached Karpov’s team. As it appears, the source has never been identified in print.
The idea of the text is that in this particular case, opening the position is in the interest of the side with the knight, not
the bishop, as the white pieces can invade.
Kasparov’s team must have hoped for 47.gxh4 Bg6 48.Kg3 Ke6 49.Ng2 Kd6 50.Nf4 Bf7 51.Nd3 Be6 52.Nc5 Bc8.
Black’s fortress withstands the pressure in this line.

47...hxg3+

If 47...h3 48.Nf4.

48.Kxg3 Ke6

Marin reported a convincing win against 48...Bg6 in his Learn from the Legends book. It requires skill from White to
prepare a way for penetration: 49.Nf4 Bf7 50.Kh4 Ke7 51.Nxh5 Ke6 (51...Bxh5 52.Kxh5 Ke6 53.Kg6+–) 52.Ng7+
Ke7 53.Kg4 Kd6 (in case of 53...Bg6 54.Nh5 Bd3 (54...Bf7 55.Nf4) 55.Nf4 Bc4 56.Kh5 Kd6 57.Kg6 Ke7 58.Kg7,
White penetrates) 54.Nh5 Be6+ (54...Ke6 55.Nf4+ Kd6 56.Kf5 Ke7 57.Ng6++–) 55.Kh4 Ke7 56.Nf4 Bf7 57.Kg4
Bg8 58.Kh5 Bf7+ 59.Kh6 Bg8 60.Kg7 Bf7 61.Ng6+ Ke6 62.Nf8+ Ke7 63.Nh7 f5 64.f4 and White wins.

49.Nf4+ Kf5 50.Nxh5 Ke6

As Dvoretsky explains, this defends against the threat of Ng7-Ne8-Nc7.

51.Nf4+ Kd6 52.Kg4 Bc2 53.Kh5 Bd1 54.Kg6 Ke7!

55.Nxd5+?

Karpov errs by taking this pawn. White will be ahead in material, but allows the black king to ‘make an impression’ on
White’s kingside pawns. Most surprisingly, Karpov does not examine this position in detail in his 2006 book. He adds
no question mark. Remember that Karpov did not hurry to take material against Ljubojevic in the 1981 game in
Linares, nor in the game with Larsen in Linares 1983, where he sacrificed a pawn to open files. This underlines that
Karpov knew this kind of idea very well! Of course it is possible that this objectively better concept did not come to his
mind.
But there is something else. Karpov and Kasparov already knew each other well; maybe Kasparov was already tense
and Karpov realized this. He may have taken the pawn to humiliate his opponent, who was in an apparently unfamiliar
situation. Even Karpov could not have been sure that the subtle invasion alternative would work. For an upset
opponent, that alternative line would have been easier to play, whereas in the position after the text move, his task is
more difficult. One should never forget that a regular over-the-board game is a battle that you want to win. That is quite
a different thing from analysing a position without any tension and trying to find the truth. Of course, it is also possible
that Karpov simply thought that capturing on d5 would win.
Excellent commentators have been analysing this game for two decades now, and yet in some lines improvements can
still be found. But it looks as though Karpov chose the weaker continuation. 55.Nh5!! was stronger. This is a very
subtle move. White takes the f-pawn but leaves the d-pawn be, as the latter limits the scope of the already passive
bishop. The king has no really active role to perform, either: 55...Bxf3 (after 55...Bc2+ 56.Kg7 Ke6 57.Nxf6 Kf5
58.Nxd5 Bb3 59.Ne7+ White wins, as Marin points out) 56.Nxf6 Be4+! 57.Kg5 Bd3!.

analysis diagram

Once again, taking on d5 would provide Black with counterplay too early. It would require too much effort to hold the
black king off, leaving White no ‘energy’ to make progress: 58.Nxd5+? Kd6 59.Nc3 (59.Ne3 and 59.Nf4 are both
equal) 59...Bf1! 60.Kf4 Bg2 61.Ke3 Bh3 62.Ne4+ Kd5 63.Nc5 Bc8 64.Kd3 Bf5+ 65.Kc3 Bc8 and White cannot do
anything with the extra d-pawn. This motif makes the position hard to play for White. Actually this is what misguided
Karpov in the game.
Marin shows how White can invade: 58.Ng4! Bf1! 59.Ne5 Bh3 60.Kg6! (Marin keeps coming up with lovely moves)
60...Ke6
analysis diagram

61.Nc6! (thus White reveals his intentions: he goes after the a6-pawn) 61...Kd6 (61...Bf5+ 62.Kg5 Bh3 loses to
63.Nb8) 62.Na5 Ke7 63.Nb3 Bd7 64.Nc5 Bc8 65.Kg7 (65.Kg5? Kf7 – now Black is in zugzwang) 65...Bf5 66.Nxa6
(the fruit of Marin’s remarkable and accurately executed idea falls into White’s hands – he has won the a-pawn.
However, Black’s position still doesn’t fall apart by itself) 66...Bd3 67.Nb8 Bc2 68.Nc6+ Ke6

analysis diagram

A) 69.Ne5 Kf5 (69...Bb1 70.Kf8; 69...Ke7 70.Ng4) 70.Kf7 Ke4 71.Nc6 Bd1 72.Ke6 Bh5 73.Kd7! (73.Kd6? Be8
74.Ne7 Kxd4 75.Nxd5 Kc4=) 73...Be2 (73...Kd3 74.a4; or 73...Bg6 74.a4! bxa4 75.b5 a3 76.Nb4 and White
promotes) 74.Kd6 Bc4 75.Kc5 and White wins, viz. 75...Kd3 76.Nd8 Kc3 77.Ne6 Kb3 78.Nc7 Kxa3 79.Nxb5+;
B) 69.Kf8 – this is the simplest way to invade: 69...Bg6 70.Na7 Bd3 71.Ke8 Be2 72.Nc6 Kd6 73.Ne7! Ke6 74.Kd8
Kd6 75.Nf5+ Ke6 76.Ne3 Kd6 77.Kc8 Kc6 78.Kb8 and White wins. These nice lines represent the cooperative work
of Dvoretsky and Marin, according to Dvoretsky.
Back to the game.

55...Ke6

With 55...Kd6!? 56.Nxf6 (after 56.Nc3 Bxf3 57.Kxf6 Bg2 White cannot win according to Dvoretsky) 56...Bxf3 57.Kf5
Bg2 58.Kf4 Bb7 59.Ke3 Bc6 60.Kd3 Bb7 61.Ne4+ Kd5 62.Nc3+ (62.Nc5 Bc8=) 62...Kd6, Black could hold.

56.Nc7+ Kd7?

Now 56...Kd6 would have been answered by 57.Ne8+ (57.Nxa6 Bxf3 58.Kxf6 Kd5 59.Nc7+ Kxd4 60.Nxb5+ Kc4=).
Garry must have been worried about this check, but after 57...Ke7 (57...Kd5 58.f4) 58.Nxf6 Bxf3 59.Kf5 Kd6 60.Kf4
Bg2, once again Black has managed to reach a defendable position.

57.Nxa6 Bxf3 58.Kxf6

White is two pawns up, but the game is far from over as Black’s king is more active and White has only one passed
pawn, which is well-blocked to boot.

58...Kd6 59.Kf5 Kd5 60.Kf4 Bh1

The further away the bishop stands the better, as this way attacks by the knight are avoided.

61.Ke3 Kc4 62.Nc5 Bc6!

Kasparov had to be careful where he placed his bishop. The following line, for example, given by Marin, loses quickly:
62...Bg2? 63.Nd3 Kb3 and now 64.Nf4! wins the vital tempo: 64...Bb7 65.Kd3 Kxa3 66.Kc3 and White wins.

63.Nd3 Bg2
Averbakh and Taimanov came up with the idea of 63...Be8!?. According to Dvoretsky, Black may hold because of his
active king, for example 64.Ne5+ Kd5!.

64.Ne5+ Kc3 65.Ng6 Kc4 66.Ne7

66...Bb7?

Kasparov retreats the bishop to an unfortunate square, missing a study-like riposte. Again no mention by Karpov in his
most recent book.
No good was 66...Kb3? 67.d5 Kxa3 68.d6 Bh3 69.Nd5 and White wins according to Averbakh and Taimanov.
Best was 66...Bh1!! according to Gipslis and Mikhalchishin, who came up with this remarkable move. It is hard to spot
the difference between their move and Kasparov’s:
A) 67.Nc8 Kd5 68.Kd3 Be4+ 69.Kc3 Kc6 and White has not accomplished anything;
B) 67.d5 Bxd5 68.Nxd5 Kxd5 69.Kd3 (if White’s a-pawn were on a2, this would win as White could lose a tempo
with a triangulation) 69...Ke5 70.Kc3 Kd5 71.Kb3 Kc6 72.Kb2 Kd6 73.Kc2 Kc6 and Black is in zugzwang, but he
would lose only if White’s a-pawn were on its starting square: 74.Kb3 Kb6 75.Kc3 Kc6 76.Kd4 Kd6 and now White
doesn’t have a spare move;
C) 67.Nf5 and now:
analysis diagram

C1) 67...Kc3? 68.Kf4 Kb3 69.Ne7 Kxa3 70.d5 Kxb4 71.d6+–;

C2) 67...Bc6? 68.Kf4 Kb3 69.Ke5 Kxa3 70.Kd6 Be4 71.Ng3 and White wins, as Speelman and Tisdall have pointed
out;
C3) 67...Kd5! – now we can see the difference that the bishop on h1 would have made: 68.Kd3 would be prevented
because of the check on e4:
C31) 68.Ng3 Bg2:
C311) Here, Timman came up with the idea 69.Kf2 to exploit the limited scope of the bishop. There follows 69...Bh3
70.Ne2 and now:
C3111) 70...Kc4 71.Kf3 (71.Ke3 Bg2; 71.Nf4 Bc8 72.Ke3 Bb7=) 71...Bf1? (71...Kb3! draws according to Dvoretsky)
72.Ng3 Bd3 73.Ke3 Bg6 74.Ne2 wins as Timman points out, e.g. 74...Kb3 75.d5 Kxa3 76.Nf4 Be8 77.Nd3 and Black
is lost;
C3112) 70...Bg4! (Marin spotted this hole in the Dutch grandmaster’s analysis) 71.Ke3 Bxe2 72.Kxe2 Kxd4 73.Kd2
Kc4 74.Kc2 Kd5 and Black draws.
C312) 69.Kd3 was analysed by Dvoretsky: 69...Bf3 (69...Bh3? loses to 70.Nh5 Bf5+ 71.Ke3 Kc4? 72.Nf6 Kb3
73.Kd2!) 70.Nf1 Be4+ 71.Kc3 Bf3 72.Ne3+ Ke4 73.Nc4 Kd5 (73...bxc4? 74.Kxc4+–) 74.Nb2 (White wants to create
two passed pawns by exchanging the a-pawn for the b-pawn) 74...Kc6 75.a4 bxa4
analysis diagram

76.Nxa4. White has achieved his aim, but the position is not winning for him.
Back to the position after 67...Kd5! from the previous diagram.
C32) Marin came up with the witty 68.Kf4. This stops 68...Kc4 for the moment in view of 69.Ke5. Now:
C321) 68...Ke6 69.Ne3 Bb7 70.Kg5 Be4 71.Nd1 Kd5 (in case of 71...Bf3 72.Nc3 Bc6 73.Kg6 Be8+ 74.Kg7 Bc6
75.Kf8 Kd6 76.d5 Bd7 77.Kf7 Black loses by zugzwang) 72.Nc3+ Kxd4 73.Nxe4! (note that Black can draw after
73.Nxb5+? Kd3) 73...Kxe4 74.Kf6 and White wins;
C322) Black must prevent the knight reaching e3 with 68...Be4! 69.Nd6 and now:
C3221) 69...Bg6? 70.Nxb5 Kc4 71.Nc7! (Marin misses this move in his otherwise impressive analysis) 71...Kxd4
(71...Kb3 72.d5) 72.a4 and White wins;
C3222) 69...Bd3 70.Ke3 Bf1 71.Ne4 Bg2 72.Nd2 (72.Nf6+ Kc4) 72...Bh3 73.Kd3 Bf5+ 74.Kc3 Bd7 and White
cannot win;
C3223) 69...Bc2! (this wins back the material at once) 70.Nxb5 Kc4 71.Ke5 (71.Nc7 Kxd4) 71...Kxb5 72.d5 Bb3!
and Black holds.
Back to the game.

67.Nf5

Karpov comes up with a threat that is easy to see, but hard to counter. 67.d5 would not work for White.
67...Bg2?

In this game Garry had no luck with the g2-square. It seems a strange mistake from Kasparov – although his previous
losses must have left their mark. He possesses fabulous calculating abilities, which deserted him at this moment,
making him miss a primitive win for White. One of the alternatives would have prolonged the resistance.
A) Not 67...Kb3? as Black cannot go after the a3-pawn: White wins with 68.Kd3 Bg2 69.Ne3 Kxa3 70.Kc3;
B) Nor does stepping out of the check on d6 with 67...Kc3? work: 68.Kf4 Kb3 69.Ne7 Kxa3 70.d5 and White wins, as
Speelman and Tisdall point out;
C) 67...Bc6? defends b5, but this doesn’t help: 68.Kf4 Kb3 69.Ke5 Kxa3 70.Kd6! Be4 71.Ng3 wins;
D) Best is 67...Kd5!, which stops the d6-check and defends the e5-square as well: 68.Kd3 Ke6 and now:
analysis diagram

D1) 69.Ng3?! Bg2 70.Ne4 (70.Ne2 Bf1) 70...Bf1+ 71.Ke3 Kd5! 72.Nc3+ Kc4 and Black is rather active;
D2) 69.Ne3 Bf3! (this is Karsten Müller’s move. Dvoretsky explains in his superb analysis that Black has to stop the
manoeuvre Nd1-c3. Black can cope with that only if his king gets to c4, but he cannot accomplish that here):
D21) 70.d5+ Ke5! (70...Bxd5? 71.Kd4 Bf3 72.Kc5);
D22) 70.Kc3 (White tries to play Kb3 followed by a3-a4 and taking back with the king – we have seen that recapturing
with the knight is not enough) 70...Kd6 71.Kb3 Bc6! (after 71...Kc6? 72.a4 bxa4+ 73.Kxa4, with his better-placed
king White is already winning) 72.d5 Be8 73.Kc3 Ke5! and Black holds;
D23) 70.Kd2. I thought that this king march would bring down Karsten Müller’s defensive plan thanks to a later
finesse, but Müller found a ‘counter-finesse’: 70...Kd6 (70...Kf6? loses to 71.Nd1 Bb7 72.Nc3 Bc6 73.Ke3 Ke6
74.d5+ Bxd5 75.Nxb5) 71.Ke1 Ke6 72.Kf2 Be4 73.Kg3 Ba8 74.Kf4 Bb7 75.Kg5 and now:
D231) 75...Be4? 76.Nd1 Bf3 (76...Kd5 77.Nc3+ Kxd4 78.Nxe4 Kxe4 79.Kf6+–) 77.Nc3 Bc6 78.Kg6 (in case of
78.d5+? Bxd5 79.Nxb5 Bc4 Black draws) 78...Be8+ 79.Kg7 Bc6 80.Kf8 Kd6 (80...Bd7 81.d5+ Kd6 82.Ne4++–)
81.d5 Bd7 82.Ne4+ (or 82.Kf7+–) wins – Dvoretsky;
D232) 75...Bf3! 76.Kg6 Be4+ 77.Kg7 Bf3 78.Kf8 Bh5. Dvoretsky claims that no progress can be made here. I have
found another idea which sets Black a new problem but turns out not to break his defence.

analysis diagram

79.Ng2! (this knight move keeps reappearing in this game! 79.Kg8!? is another try) and now:
D2321) 79...Kd5 80.Nf4+ Kxd4 81.Nxh5 Kc3 82.Ke7 (or also 82.Nf4 Kb3 83.Nd5) 82...Kb3 83.Kd6 Kxa3 and now
84.Kc5 wins;
D2322) 79...Kf5 80.Ke7 Ke4 81.d5! and White wins;
D2323) 79...Kd6 80.Nf4 Bf3?! (according to Dvoretsky, 80...Bg4 is even simpler as it makes it easier to avoid the later
zugzwang possibility) 81.Ke8 Be4 (again 81...Bg4!?) 82.Kd8 Bf5 (after 82...Bb7 83.d5 Ba6 84.Ne6 Bb7 85.Nc7
Black is in zugzwang) 83.Ng2!? (one wonders if Garry would have appreciated the amusing side of the repeated Ng2-
move. Ultimately, the g2-square is the location for the final mistake of the game. Instead, 83.Ne2 Kd5 84.Nc3+ Kxd4
85.Nxb5+ wins) 83...Be4 (83...Be6 84.Ne3 Bd7 85.Nd5 wins prettily for White; 84...Bh3 85.d5 is zugzwang, and
85...Bd7 loses to 86.Nd1) 84.Ne3 Bb7 85.d5 Ba6 (85...Ba8 86.Kc8) 86.Nf5+! Kxd5 87.Kc7 and Black’s bishop is
trapped;
D2324) 79...Bf3! 80.Nf4+ Kd6! (80...Kf5? 81.d5! Kxf4 82.d6 Bc6 83.Ke7 and White wins) 81.Ke8 Bd1 82.Kd8 Bg4
D2324a) 83.Ng2 Bf5! (83...Be6? 84.Ne3) 84.Ne3 Be6 85.Nd1 Kd5 86.Nc3+ (White cannot win by 86.Ke7 Bg8
87.Nc3+ Kxd4 88.Nxb5+) 86...Kxd4 87.Nxb5+ Kc4 – with the bishop on f5, Nd6 would win, but now it is a draw;
D2324b) 83.d5 Bf5 (83...Ke5? 84.Kc7 Kxf4 85.d6 Ke5 86.d7 Bxd7 87.Kxd7 Kd5 88.Kc7) 84.Ng2 (84.Ke8 Be4
(84...Ke5? 85.Ke7) 85.Kf7 Bh1 (85...Bxd5+ 86.Nxd5) 86.Kf6 Be4 and White cannot make any progress) 84...Bg4
(84...Be4? 85.Ne3 Bf3 86.Kc8) 85.Ne3 Bh3 86.Ke8 Bc8 87.Kf7 Bb7 (87...Bh3 88.Kf6 Bd7) 88.Kf6 Ba8 (88...Bc8;
88...Bxd5? 89.Nxd5 Kxd5 90.Ke7 Kc4 91.Kd6 Kb3 92.Kc5) 89.Kf5 Bxd5. Now the white king is too far away from
the b-pawn, so Black is safe.
Back to the analysis diagram after 67...Kd5! 68.Kd3 Ke6.
D3) Marin found the stylish 69.Ng7+!!, which seems to be the only way to win:
D31) If 69...Kf7 70.Nh5;
D32) After 69...Kd6, 70.Ne8+ Ke7 (70...Kd7 loses to 71.Nf6+ Ke6 72.Ne4) 71.Nc7 Bc6 72.d5 spells the end for
Black;
D33) 69...Kd7 70.Nh5 Bg2 71.Nf4 (71.Ng3 Ke6 72.Ne4 Bf1+ 73.Ke3 Kd5=) 71...Bf1+ 72.Ke4 Kd6

analysis diagram

D331) 73.d5 Bc4 74.Kd4 Ba2 is equal;

D332) 73.Ke3! and Black is in zugzwang. This is something of a surprise: 73...Kc6 (73...Bc4 loses to 74.Ne2 Kd5
75.Nc3+ Kd6 76.Ke4) 74.d5+ Kd6 75.Kd4 Bc4 76.a4 and White wins.

68.Nd6+ Kb3 69.Nxb5 Ka4 70.Nd6 1-0

 GAME 83
Karpov, Anatoly
Kasparov, Garry
Moscow World Championship 1984 (27)

1.Nf3 d5 2.d4 Nf6 3.c4 e6 4.Nc3 Be7 5.Bg5 h6 6.Bxf6 Bxf6 7.e3 0-0 8.Qc2 c5 9.dxc5 dxc4 10.Bxc4 Qa5 11.0-0
Bxc3 12.Qxc3 Qxc3 13.bxc3 Nd7 14.c6 bxc6 15.Rab1 Nb6 16.Be2 c5

17.Rfc1!!

In a way this is an incredible move, and in a way it is just a harmless novelty. It is hard to understand why Karpov
places his rook behind his pawn rather than on the open file, which would be the classical way to play. The next ten
moves are somewhat reminiscent of the Franco-Russian war in 1812, when Napoleon encircled Moscow. Kutuzov
avoided the fight, thus winning the war, and Karpov’s rook avoids a confrontation on the d-file. It prepares to defend
the c-pawn if necessary. It doesn’t pose Black unsolvable problems, but it is hard to appreciate what Karpov really
meant here. According to the database this move has been played 16 more times and Black drew every time.
Alternatively, 17.Ne5 Bb7 18.Nd7 Rfc8 19.Nxb6 axb6 is equal.

17...Bb7?!

Ever since this game, players have developed the bishop to d7 and followed up with ...Rfd8 and ...Kf8. Black is safe.

18.Kf1 Bd5?!

Garry still doesn’t fully appreciate the danger. It probably has to do with the fact that the previous 17 games of the
match had ended in a draw. Then it is easy to think that all games will end that way.
Black could still have transferred the bishop to the better diagonal with 18...Bc6!?. Then, after 19.Ne5 Ba4 20.Bb5
Bxb5+ 21.Rxb5 Rfc8 the loss of a tempo doesn’t cause Black any trouble. He can still hold.
19.Rb5!

To be able to inject energy into the position was the point of the mysterious rook move.

19...Nd7?!

It is logical to defend the c-pawn with the least possible effort, but on d7 the knight stands in the way of the bishop.
A) Not 19...Bxa2?, when 20.c4 wins;
B) Best was 19...Rfc8! 20.c4 Bb7 (after 20...Bxf3?! 21.Bxf3 Rab8 22.Ke2 Rc7 23.a4 Rd8 24.a5 White has an edge)
21.Ne5 Rc7 22.Nd3 Rac8 and Black seems to be able to cope with White’s slight pressure;
C) 19...c4!? – even though the pawn wanders away from camp, maybe it can be defended.

20.Ra5

Rather like Kutuzov, the other rook also avoids contact on the open (b-)file. The rook exerts pressure on both of Black’s
queenside pawns, causing many problems. The fact that he played for this type of continuation in advance shows
Karpov’s sharp vision.

20...Rfb8?

Garry is acting like Napoleon. He sees the chance to invade and he doesn’t miss it. Garry seeks play on the b-file and
the second rank, but it turns out that the rook cannot do enough there and will be repulsed.
Black could still consider protecting the c5-pawn with the quiet 20...Rfc8, for example 21.c4 Bc6 22.Ne1 Rc7 (or
22...Be4 23.f3 Bg6 24.Nd3 Rc7) 23.Nd3 Bb7 24.f3 a6 25.Bd1 Rac8 26.Rb1. In both cases Black has chances to get
away with just defending passively. He may soon be able to centralize his king.

21.c4 Bc6
22.Ne1!

Going backward just like Kutuzov, Karpov is closing in on the target.

22...Rb4

Just like Napoleon, penetrating deeper into the Motherland. Not 22...Rb2? 23.f3! (23.Nd3 Bxg2+) and White follows up
with Nd3. After 22...Be4 23.f3 Bg6 24.e4 Rb7 25.Nd3 Rc8 Black is passive, but he doesn’t lose material.

23.Bd1!

Just like in 1812 the defence is retreating. Anatoly stops Garry exchanging the strong a5-rook from a4.

23...Rb7

Just like the French army, he quickly abandons what has been achieved.
Maybe he should have tried to exchange a rook anyway with 23...Be4, but Garry no longer has pleasant options. After
24.f3 Rb1 25.Rxb1 Bxb1 26.f4 White is somewhat better.
24.f3!

Limiting the scope of Garry’s pieces further.

24...Rd8

Giving up the pawn with 24...Kf8 leads to a depressing ending where Black is a pawn down as well: 25.Nd3 Ke7
26.Nxc5 Nxc5 27.Rxc5.
In case of 24...Rc7 25.Nd3 Bb7 26.Ba4 Nb6 27.Rxc5 Rxc5 28.Nxc5 Bxf3 29.gxf3 Rc8 30.Nxe6 Nxa4 31.Nf4 White is
better according to Tisdall.

25.Nd3 g5 26.Bb3!

In the spirit of 1812, he is still avoiding direct contact. The text is a nice subtle move. White postpones winning the
pawn until he has closed off the b-file. Maybe Garry was hoping for 26.Nxc5? Nxc5 27.Rxc5 Rb2 28.Rxc6 Rdd2 after
which Black’s rooks on the second rank secure a draw.

26...Kf8
27.Nxc5

Kasparov has lost a smaller portion of his army than Napoleon. However, his future in the game is equally doomed. Ten
moves earlier Anatoly has started to play against c5 – now he harvests. His position is winning with the extra pawn, but
the win still requires the necessary technique.

27...Nxc5 28.Rxc5 Rd6 29.Ke2 Ke7 30.Rd1 Rxd1 31.Kxd1 Kd6 32.Ra5 f5 33.Ke2 h5

Garry still goes forward; at this point it was hard to just sit and wait. Maybe 33...Ke7 was the best waiting move, but it
is ugly. After 34.Kd3 Kf6 35.Ra6 Rc7 36.Kd4, White is much better.
Tisdall recommends 33...Rg7, but this is no better than the game move: 34.Kd3! g4 35.c5+ Ke5 36.f4+ Kf6 37.g3 and
White is a pawn up for nothing.
34.e4!?

With the help of the rook on a5, Karpov opens up the position.
Geller recommends 34.Kd3 in his Chess Informant analysis, to prepare e3-e4. This would have given excellent winning
chances as well. But Karpov is eyeing Black’s kingside pawns.

34...fxe4 35.fxe4 Bxe4 36.Rxg5 Bf5 37.Ke3

The quiet 37.g3 looks attractive as well.

37...h4 38.Kd4 e5+ 39.Kc3

Karpov signals that he is going to play on the queenside again.

39...Bb1

Aiming to reduce the material with 39...Ke6 40.Rh5 Be4 was another option. Regardless, Black’s position is hard to
play.

40.a3 Re7
41.Rg4

Tal preferred 41.h3.

41...h3

Garry hopes that White having the h-pawn with the wrongly-coloured bishop may help him. Geller mentions 41...e4, but
White can easily block this pawn: 42.Bd1 Rf7 43.Kd4 Rf2 44.c5+ Ke7 45.Ke3 (this is more convincing than Geller’s
line 45.Bb3 Rd2+ 46.Kc3 e3!) 45...Rf1 46.Be2 Rc1 47.Rxh4 Rxc5 48.Rh7+ and White wins.

42.g3

Karpov prefers to keep the g-pawn rather than win another pawn. The black rook cannot go to the f-file directly.

42...Re8 43.Rg7!

Karpov bases his play on his queenside pawns and lets Kasparov have his own passed pawn.

43...Rf8 44.Rxa7 Rf2

45.Kb4! Rxh2

Garry takes the h-pawn, but Karpov can create a mate threat.
Kasparov could have simplified to a rook ending with just a small number of pawns with 45...Rb2. This must have been
checked carefully and the team probably discovered it was lost: 46.c5+ Kc6 47.Kc4 (Karpov can hunt the king in the
endgame like no other) 47...Ba2 (47...Bc2 48.Ra6+ Kc7 49.Bxc2 Rxc2+ 50.Kd5 Rxh2 51.Rh6 Rh1 52.Rh7+ transposes)
48.Bxa2 Rxa2 49.Ra6+ Kb7 50.Rb6+ Kc7 (unlike in the game, Black’s king can stay in front of the c-pawn, but it
strays too far from the white g-pawn) 51.Rh6 Rxh2 52.Kd5 Rh1 53.Rh7+ and now:

analysis diagram
A) 53...Kc8 54.Kd6! h2 55.Rh8+ Kb7 56.c6+ Kb6 57.c7 Rd1+ 58.Kxe5 Kxc7 59.Rxh2 and White wins;
B) 53...Kb8 54.Ke4! (a nice switch):
B1) 54...Kc8 55.Kf3 Kd8 56.a4 h2 57.Kg2 Ra1 58.Kxh2 Rxa4 59.Rh5 Re4 (59...e4 60.Re5; or 60...Rc4 61.g4 e3
62.Kg3 e2 63.Rxe2 Rxc5 64.Kh4) 60.Kg2 Kd7 61.Kf3 Re1

analysis diagram

62.Kf2!. Quite wittily White ‘traps’ the black rook: 62...Re4 and now 63.Rh4! Rd4 64.Rxd4+ exd4 65.g4 and wins;
B2) 54...h2 55.Kf3 Ra1 (55...Rc1 56.Rxh2 Rxc5 57.Ke4 Rc3 58.Rg2 Rxa3 59.Kxe5 and White wins) 56.Rxh2 Rxa3+
57.Kg4 Re3 (57...Rc3 58.Re2 Rxc5 59.Kf5 – the g-pawn can only rarely be stopped if the defending king is far away;
this is winning as well) 58.Rh7 Kc8 59.c6 Kd8 60.Rd7+ Kc8 61.Rg7 Re1 62.Kf5 e4 63.c7 e3.
analysis diagram

Now White wins with 64.Ke4 e2 65.Ke3.

Let’s go back to the game after 45...Rxh2.

46.c5+ Kc6 47.Ba4+ Kd5

48.Rd7+!

Now Garry’s king is cut off from White’s strong c-pawn.

48...Ke4

On 48...Ke6 49.c6 Rb2+ 50.Bb3+ Rxb3+ 51.Kxb3 Be4 52.Rd8 Bxc6 53.Rh8 Bg2 54.a4 Kf5 55.Rh4 White wins, as
Karpov points out.

49.c6 Rb2+ 50.Ka5! Rb8 51.c7

The c-pawn decides the issue.

51...Rc8 52.Kb6 Ke3 53.Bc6 h2 54.g4 Rh8

If 54...e4, 55.Rh7 wins.

55.Rd1 Ba2 56.Re1+

Now even the e-pawn falls.

56...Kf4 57.Re4+ Kg3 58.Rxe5 Kxg4 59.Re2 1-0

 GAME 84
Karpov, Anatoly
Miles, Anthony
Amsterdam OHRA 1985 (4)

1.e4 c6 2.d4 d5 3.Nd2 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.h4 h6 7.Nf3 Nd7 8.h5 Bh7 9.Bd3 Bxd3 10.Qxd3 Qc7 11.Bd2
e6 12.0-0-0 Ngf6 13.Ne4 0-0-0 14.g3 Nc5 15.Nxc5 Bxc5 16.Qe2 Bb6 17.Bf4 Qe7 18.c4 Rhe8 19.Qe5 Bc7 20.Qxc7+
Qxc7 21.Bxc7 Kxc7 22.Rh4 Re7 23.Ne5 Nd7 24.Rf4 Rf8 25.Kc2 Nxe5 26.dxe5 Rd7 27.Rxd7+ Kxd7 28.c5 Ke7
29.Ra4 Ra8 30.Kc3 Kf8 31.Rd4 Ke7 32.Rb4 Rb8 33.Ra4 Ra8 34.Ra3 Ke8 35.Rb3 Rb8 36.Ra3 Ra8 37.Kc2 Ke7
38.Rb3 Rb8 39.Ra3 Ra8 40.Ra4 Ke8 41.Rb4 Rb8 42.Ra4 Ra8 43.Kc3 Ke7 44.Kc4 Ke8 45.Ra3 Ke7 46.Rb3 Rb8
47.Ra3 Ra8

48.f4

Karpov has obtained a slightly better ending out of the opening, but he has not made a pawn move for twenty moves.
Now he opts for a change of structure.

48...g6

Miles wants to recover some ground. Doing nothing would offer chances of survival; however, the better option seems
to be to try and break out. Karpov honours the text with an exclamation mark. Black had another move in the same
spirit as the game continuation – it is 48...f6!?.

49.hxg6 fxg6 50.Kd4 h5?!

Karpov recommends 50...a6! 51.Ke4 Rd8 as a strategy that might hold: 52.Rb3 (after 52.g4 Black has better chances of
drawing than White has of winning) 52...Rd7 53.g4 Ke8 54.Rh3 (54.f5 exf5+ 55.gxf5 gxf5+ 56.Kxf5 Rf7+ 57.Kg6 Re7
is equal) 54...Rh7 55.f5 exf5+ 56.gxf5 gxf5+ 57.Kxf5 Rf7+ 58.Kg6 Ke7 59.Rxh6 Ke6 and he evaluates this as equal.

51.Ke4 Kf7 52.Rb3 b6

Waiting passively with 52...Rb8 was tough on Black as well: 53.Kf3 Ke7 54.g4 Kf7 55.Kg3 and White gets decent
winning chances by marching his king to g5.

53.Rd3!!

White wastes no time improving his rook with this subtle glide back to the d-file.

53...Ke7 54.Rd6 bxc5

Waiting passively with 54...Rc8 gives little hope because Black’s rook as well as his king are tied up: 55.Kd4 Rc7 56.b4
Rc8 (in case of 56...b5 57.Ke4 the white king will invade on the kingside) 57.a4 Rc7 58.Kc4 Rc8 59.cxb6 axb6 60.a5
and White wins.

55.Rxc6 Rb8 56.Rxc5!

The simplest.

56...Rxb2 57.Rc7+ Kf8 58.Rxa7 Rc2

59.Kf3!

White defends g3, after which pushing the a-pawn forward will decide.

59...Kg8 60.a4 Ra2 61.a5 Kf8 62.a6 Kg8 63.Ra8+ Kg7 64.a7

Normally one must be careful when pushing the pawn this far when the rook is in front of it. If the king cannot create
another passed pawn it is a draw. Moreover, the new passed pawn may not be a g- or an h-pawn.

64...Ra3+ 65.Ke4

And Miles resigned, as the king simply marches over to collect the e6-pawn.

GAME 85
Karpov, Anatoly
Kasparov, Garry
Moscow World Championship 1985 (22)

1.d4 d5 2.c4 e6 3.Nc3 Be7 4.cxd5 exd5 5.Bf4 Nf6 6.e3 0-0 7.Nf3 Bf5 8.h3 c6 9.g4 Bg6 10.Ne5 Nfd7 11.Nxg6 fxg6
12.Bg2 Nb6 13.0-0 Kh8 14.Ne2 g5 15.Bg3 Bd6 16.Qd3 Na6 17.b3 Qe7 18.Bxd6 Qxd6 19.f4 gxf4 20.exf4 Rae8
21.f5 Nc7 22.Rf2 Nd7 23.g5 Qe7 24.h4 Qe3 25.Rd1 Nb5 26.Qxe3 Rxe3
27.Kh2!

One of those fine-tuning moves.

27...Nb6

Garry wants to transfer this knight to d6 and keep the other on b5 to keep an eye on d4. He could have put the other
knight on d6 and kept its d7-compatriot where it was.

28.Ng3 Nc8 29.Nf1 Re7

Better was 29...Rc3! – the exclamation mark is borrowed from Karpov – where the rook is stronger, e.g. 30.Ng3 Ncd6
and Black does better than in the game.
30.Rd3!

As usual, Karpov improves whenever he can.

30...Ncd6 31.Ng3 Ne4?

Garry forces matters, but the exchange helps White. With fewer pieces on the board it is easier to improve the king.
Karpov mentions 31...Kg8 and attaches a plus/equal mark. Maybe he would have improved with something like 32.Kh3
Re1 33.Rf4!? (in case of 33.Kg4 Ne4 Black gets some play) 33...Kf7 34.Kg4. White is slightly better with his initiative
on the kingside.

32.Bxe4

After 32.Nxe4 dxe4 33.Re3 Nd6 34.Bh3 Ref7 the d6-knight is well placed, giving Black a good game.

32...dxe4 33.Re3 Nxd4

A knight on d6 is less effective now. Karpov evaluates 33...Nd6 34.Kh3 as clearly better for White. His plan is Kg4, h4
and Rf4. Black can do little but wait and see how White develops his plan.
34.Kh3!

White could have started action with 34.f6, but why not improve the king first? It is hard to judge whether the
immediate push is strong or whether it fails in the end: 34...gxf6 35.gxf6 Re6 (after 35...Rd7 36.Nxe4 b6 37.b4 White
has only a small initiative, but it is hard to get rid of it) 36.Rxe4 c5! 37.Rxe6 Nxe6 38.Ne4 Kg8 39.Rd2 (this is a bit
frightening for Black) 39...Rd8!? 40.Rxd8+ Nxd8 41.Nxc5 Kf7 42.Ne4 h6 43.Kg3 Ke6 44.Kg4 Ke5 and it looks as
though Black just holds.

34...Re5

Black could have held back the white king with 34...h5?! but he would give up the g5-square in return – according to
Karpov this is not worth it: 35.f6 gxf6 36.gxf6 Re6 37.Rxe4 Rxe4 38.Nxe4 and, as Karpov mentions, Ng5 follows. That
will be strong in combination with Kg3-Kf4.

35.Kg4 h5+

After 35...Kg8 36.Rxe4 Rxe4+ 37.Nxe4 b6 38.f6 c5 39.h5 White’s space advantage on the kingside will stifle Black.
After 35...c5 36.Rxe4 Nc6 37.Rfe2 Rd5 38.h5 Black does not have a leg to stand on.
36.Kxh5!

Now the white king invades with decisive force.

After 36.Kf4 Rexf5+ 37.Nxf5 Nxf5 38.Kxe4 Nd6+ 39.Ke5 Rxf2 40.Kxd6 Kh7 Black would still be alive.

36...Nxf5 37.Rxf5 Rfxf5 38.Nxf5 Rxf5 39.Rxe4

Suddenly Black is dead lost.

39...Kh7

In case of 39...Rf7 40.Kg6 Kg8 41.Re8+ Rf8 42.Rxf8+ (this is the most instructive win; 42.Re7 wins easily as well)
42...Kxf8 43.Kh7 c5 (43...Kf7 44.h5 c5 45.g6+) 44.h5 b5 45.g6 c4 46.h6 White wins the race by a huge margin.

40.Re7 b5

If 40...Rb5, then 41.a4.

41.Rxa7 b4 42.Kg4

Kasparov resigned. A powerful performance by Karpov.

Chapter 4
Later Years

Luckily for chess, Karpov did not lean back after the loss of his title and decided to try to raise his level and beat
Kasparov. His games against his successor had improved his own play, even his endgames.
Karpov’s play had sharpened and his openings had improved. He did not aim straight for the endgame any more. In the
return match with Kasparov in London and Leningrad, the most uncompromising of all their matches, Karpov played
for a win also with black. Again he lost with the narrowest of margins: 11½-12½. Their fourth match, in Sevilla the
next year, ended in a tie and Kasparov retained his title. In 1990, in New York and Lyon, Kasparov was already the
better player. Again the final score was close (11½-12½), but Karpov was defending and won a game only when he had
already lost the match.

Challenging Kasparov (1986-1987)

In 1986 Karpov scored several big tournament successes. In the strong SWIFT tournament in Brussels, he left runner-up
Kortchnoi two points behind. In Bugojno he finished one point ahead of Andrey Sokolov, despite losing ½-1½ to him.
In the next year he beat Sokolov in the Candidates’ final convincingly with 7½-3½. From this match we have taken two
endgames. From the fourth Kasparov match we have taken a brilliant concept against Kasparov’s passed a-pawn in
Game 88, and a sharp endgame in Game 89. From Sevilla we give one strong win (Game 95).

GAME 86
Karpov, Anatoly
Timman, Jan
Brussels SWIFT 1986 (2)

1.Nf3 Nf6 2.c4 b6 3.g3 c5 4.Bg2 Bb7 5.0-0 g6 6.d4 Bg7 7.Nc3 Ne4 8.Nxe4 Bxe4 9.d5 e5 10.Qb3 0-0 11.Bh3 Bxf3
12.Qxf3 f5 13.e4 f4 14.Qd1 d6 15.Bd2 a5 16.Qa4 Bh6 17.Rac1 Bg5 18.Rc2 Kh8 19.a3 Ra7 20.Be1 Qe8 21.Qxe8
Rxe8 22.Bd2 Na6 23.Rb1 Rf8 24.Rc3 Kg7 25.Bg4 Raf7 26.Rb3 Rb8 27.Bd1 h5 28.h4 Bh6 29.g4 hxg4 30.Bxg4 Kf6
31.Kf1 Rfb7 32.Ke1 Bf8 33.Rh3 Rh7 34.Ke2 Be7 35.h5 gxh5 36.Rxh5 Rxh5 37.Bxh5 Rg8 38.Bf3 Rh8 39.Bg4 Nc7
40.Bd7 Kf7 41.Ra1 Bg5 42.Bc3 Ke7 43.Bf5 Bf6 44.Rd1 Ra8 45.Kd3 Rh8 46.Be1 Ra8 47.Bd2 Rh8 48.Rc1 Ra8
49.Rc3 Rh8 50.Rb3 Na8 51.Ke2 Bg5 52.Be1 Rh1 53.Bc8 Bh4 54.Bd2 Rh2 55.Be1 Rh1 56.Bb7 Nc7 57.Bc6 Kd8
58.Bd2 Rh2 59.Be1 Rh1 60.Ba4 Kc8 61.Bc3 Rh2 62.Rxb6 Rxf2+ 63.Kd3 Rf3+ 64.Kc2 Rg3 65.Rxd6 f3 66.Rh6 f2
67.Rh8+ Kb7 68.Bc6+ Ka6 69.Rf8 Rg8 70.Rf3 Rg3 71.Rf5 Rg5 72.Rf7 Rg7 73.Rf3 Rg3 74.Rf5 Rg5 75.Rf8 Rg8
76.Rf7 Rg7
77.Rxg7

Timman has kept attacking the white rook for quite some time. Karpov now decides to play for a win with a rook, a
piece and a pawn against the queen.

77...f1=Q 78.Rxc7 Qxc4

A) Maybe 78...a4 is the simplest. Black can make use of the time it takes White to capture the pawn:
A1) In case of 79.Bb5+ Kb6 80.Rc6+ (80.d6 Qe2+) 80...Kb7 81.Bd2 Qe2 82.Kc1 Be1 White cannot find shelter for
his king;
A2) After 79.Bxa4 Qxc4 80.Bc6 Qxe4+ 81.Kb3 Be1 White’s king is too exposed to win this endgame.
B) 78...Qe2+!? (Timman gives very interesting lines with unusual material configurations) 79.Kb3 a4+ (79...Qd1+
80.Ka2 a4 81.Rb7 Qd3 holds as well) 80.Kxa4 (80.Bxa4 Qd1+ 81.Ka2 Qxa4 82.Rxc5 Qc2 83.Bxe5 Qxe4 84.Bc3 Be1
and most probably White cannot win) 80...Qc2+ 81.b3 Qxc3 82.Bb5+ Kb6 83.Rc6+ Kb7 84.Rxc5 Bd8 and in these
lines by Timman Black keeps the balance.

79.a4! Qxe4+ 80.Kb3 Be1?

Timman is rebuilding White’s shelter. Maybe objectively this is not the losing move, but Black is playing with fire.
A) Also to be considered was 80...c4+!? 81.Ka2 Kb6!? and Black can wait and see: 82.Rb7+ (82.Rc8 Be1 83.d6 Qc2
and White cannot escape the coming checks) 82...Ka6 83.Rb8 Bf2 84.Ra8+ (84.Re8 Bd4) 84...Ba7 and White is also
tied up – Black holds;
B) 80...Qb1 is Timman’s recommendation, and now:
B1) 81.d6 Qd1+! – this check is the point of Timman’s move. It makes the difference compared with the game;
B2) 81.Bb5+ Kb6 82.Rc6+ Kb7 83.d6 Be1 84.Bxe1 (84.d7 Qd1+) 84...Qxe1 85.d7 Qd1+ 86.Ka2 Qd5+ 87.Ka3 Qf3+
and Black holds.

81.d6 Qh4?
This removes the queen from the scene and loosens any grip he had on the enemy king.
A) Black could have encaged the king in the corner with 81...c4+ 82.Ka2 Qc2 83.Bb5+ Kb6 84.Rc6+ (84.Rxc4 Bxc3
85.Rxc3 Qg2 86.d7 (86.Rd3 e4!) 86...Qd5+) 84...Kb7 85.Rxc4 Bxc3 86.Rxc3 Qg2 87.d7 Qd5+. This certainly doesn’t
give White an easy way to proceed; 88.Kb1 Qh1+ 89.Rc1 Qe4+ 90.Ka1 Qd4 91.Rc5 Qd1+ 92.Ka2 Qd4 93.Rxe5 Kc7
and it looks as though Black can survive;
B) 81...Qb1!? – it is not natural to change strategy, but this time it was the most practical choice: 82.d7 c4+ (after
82...Qd1+ 83.Ka2 Bxc3 84.Rc8 White wins according to Timman) 83.Kxc4 Bh4! (Timman doesn’t mention this
attempt to stop the d-pawn with the bishop. Now Black’s position is a hard nut to crack) 84.Bb5+ (84.Rc8 Qa2+ 85.b3
Kb6 86.Bb5 Qe2+ 87.Kd5 with a perpetual) 84...Kb6 85.Rc6+ Kb7 86.Rd6 Bd8 and Black should be able to withstand
the pressure.

82.Bb5+!

Karpov exploits his tactical chances with remarkable precision.

82...Kb6 83.Rc6+ Kb7 84.d7 Bxc3 85.bxc3 c4+ 86.Rxc4 Qe1

Trying to force a perpetual with 86...Qf6 does not work: 87.Rc8 Qf5 88.Bc6+ Ka6 89.Ra8+ wins.
87.d8=N+!

Relinquishing the new queen in favour of a knight and a tempo. It is possible, yet unlikely, that the endgame composer
Timman missed this underpromotion when he played 81...Qh4. However, he may have missed that Karpov’s accurate
move sequence provides an escape for his king and Timman has no perpetual.

87...Ka7! 88.Rc7+!

If 88.Nc6+ Kb6.

88...Kb8 89.Rd7! Qb1+ 90.Kc4 Qf1+ 91.Kd5 Qf3+ 92.Kd6! Qf8+ 93.Ke6 Qh6+ 94.Kxe5 Qe3+ 95.Kd6 Qf4+
96.Kc5 Qe3+ 97.Kc4 Kc8 98.Nf7 Qe4+ 99.Kc5 Qe3+
100.Kc6

Now he gives the c3-pawn, but it is not charity. The king invades and sets up a mating net.

100...Qxc3+ 101.Kb6 Qe3+ 102.Ka6 Qe6+ 103.Nd6+ Kb8 104.Rd8+ Kc7 105.Rc8+ 1-0

GAME 87
Karpov, Anatoly
Timman, Jan
Bugojno 1986 (10)

1.Nf3 Nf6 2.c4 g6 3.d4 Bg7 4.g3 0-0 5.Bg2 c6 6.Nc3 d5 7.cxd5 cxd5 8.Ne5 e6 9.Bg5 Qb6 10.Qd2 Nfd7 11.Nf3 Nc6
12.Rd1 Nf6 13.0-0 Bd7 14.Bxf6 Bxf6 15.e4 Qa5 16.Qf4 Bg7 17.Rfe1 Rad8 18.exd5 exd5 19.Ne5 Be6 20.a3 Qb6
21.b4 a5 22.b5 Ne7 23.Bf1 Qd6 24.Na4 Bc8 25.Qc1 Nf5 26.Qc5 Rfe8 27.Qxd6 Rxd6 28.f4 g5 29.Bh3 gxf4 30.gxf4
Nh4 31.Bxc8 Rxc8 32.Rc1 Re8 33.Kf2 Rf6 34.Nd3 Rxe1 35.Rc8+ Bf8 36.Kxe1 Kg7 37.Rb8 Nf3+ 38.Kd1 Nxd4
39.Rxb7 Bxa3 40.b6 Rg6 41.Ne5 Rf6 42.Rd7 Nc6 43.b7 Nb8 44.Rd8 Rxf4 45.Nc3 Be7 46.Rxb8 Bd6 47.Re8 Rb4
48.Nf3 Rxb7 49.Rd8 Bb4 50.Nxd5 a4 51.Nxb4 Rxb4 52.Ra8 Kg6 53.Ke2 f6 54.Ra5 h5 55.Ke3 h4
56.Nd4!?

Karpov places the knight in the centre where it exerts control in both directions. Thus he restricts the movement of the
black rook, and the f-pawn is not moving either. Of course Black has chances to draw, as only the h-pawn has to be
eliminated.

56...Rc4?

This is a witty move with which Black intends to attract the king to the queenside to improve the rook’s chances of
hunting down the h-pawn. However, time and again Karpov calculates with great precision.
Karpov suggests 56...Rb2!? instead of the game continuation. This looks better – even 56...h3 deserved consideration.
After 56...Rb2!? 57.Ne2 Rb3+ 58.Ke4 (58.Kd4 Rb4+ 59.Kc3 Rb3+ 60.Kc4 or 60.Kc2 Rf3 and Black soon gets the h-
pawn) 58...f5+! (this little tactic is an improvement on Karpov’s line 58...Rb4+ 59.Kf3 which the champion evaluates as
plus/minus) 59.Kd4 Rb2 60.Ke3 (60.Nf4+ Kg5 – here the f-pawn provides cover, which is the point of 58...f5)
60...Rb3+ Black has decent drawing chances.

57.Kd3 Rb4 58.Kc3 Rb1 59.Rxa4 Rf1

60.Kc4!

In typical fashion Karpov launches his attack against Black’s king. This may have been the result of home analysis.
After 60.Ra2 Kg5 61.Rg2+ Kf4 Karpov evaluates the position as just slightly better. It is hard to see how Black can
make further improvements and White may be able to do something here.

60...Kg5

After 60...Rf2 61.h3 Rg2 62.Ra3 Rg3 (62...Kg5 63.Rf3 Rg3 64.Ne6+ wins) 63.Rxg3+ hxg3 64.Kd3 Kg5 65.Nf3+ Kf4
66.Ke2 wins as Karpov points out: 66...f5 67.Ng1.

61.Ra3 Rf2 62.Kd5!

This is not difficult to see, but it is attractive nonetheless.

62...Kg4 63.h3+ Kf4

64.Ke6! Rd2 65.Nf5 Kg5 66.Rf3

Karpov weaves his mating net. As always he goes about it quite purposefully.

66...Ra2 67.Ne7 Ra6+

If 67...Ra5 68.Rf5+ wins.

68.Kf7 Ra5 69.Rxf6 Ra3

It looks like Black finally wins the pawn.

70.Rf5+! 1-0

But that won’t happen as Karpov checkmates first (70...Kh6 71.Ng8+). Karpov has kept going after the enemy king
with lethal precision.

GAME 88
Karpov, Anatoly
Kasparov, Garry
London/Leningrad Wch-m 1986 (5)

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Bf4 Bg7 5.e3 c5 6.dxc5 Qa5 7.Rc1 Ne4 8.cxd5 Nxc3 9.Qd2 Qxa2 10.bxc3 Qxd2+
11.Kxd2 Nd7 12.Bb5 0-0 13.Bxd7 Bxd7 14.e4 f5 15.e5 e6 16.c4 Rfc8 17.c6 bxc6 18.d6 c5 19.h4 h6

20.Nh3!! a5 21.f3 a4 22.Rhe1! a3 23.Nf2 a2 24.Nd3 Ra3 25.Ra1 g5 26.hxg5 hxg5 27.Bxg5 Kf7 28.Bf4 Rb8 29.Rec1
Bc6 30.Rc3 Ra5 31.Rc2 Rba8 32.Nc1 1-0

GAME 89
Kasparov, Garry
Karpov, Anatoly
London/Leningrad Wch-m 1986 (18)

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.Nc3 Bb4 5.Bg5 Bb7 6.e3 h6 7.Bh4 Bxc3+ 8.bxc3 d6 9.Nd2 g5 10.Bg3 Qe7 11.a4 a5
12.h4 Rg8 13.hxg5 hxg5 14.Qb3 Na6 15.Rb1 Kf8 16.Qd1 Bc6 17.Rh2 Kg7 18.c5 bxc5 19.Bb5 Nb8 20.dxc5 d5
21.Be5 Kf8 22.Rh6 Ne8 23.Qh5 f6 24.Rh7 Ng7 25.Qf3 Kf7 26.Qh5+ Kf8 27.Qf3 Kf7 28.Rh6 Ne8 29.e4 g4 30.Qf4
Bxb5 31.Rxb5 Nd7 32.Bxc7 Nxc5 33.Qe3 Nxe4 34.Nxe4 dxe4 35.Bxa5 f5 36.Bb4 Qd7 37.Qd4 Ra7 38.Rh7+ Ng7
39.a5 Kg6 40.Qxd7 Rxd7
41.Rh4

Garry sealed this move. While he was thinking he gradually became more optimistic.

41...Rgd8!

Karpov and Kasparov played extremely exciting, almost wild chess in many of the games of these matches. This was
the only game in which the dust settled with no indication of a clear-cut endgame. Karpov improves his passive rook
and starts going after Garry’s king.
Kasparov writes that his team analysed the game until 6 in the morning. 41...Nh5 is the other move they had considered:
42.g3 e3 43.Rb6! exf2+ (43...Kg5 44.Bc5!) 44.Kxf2 Kg5 45.Bd6 Rgd8 46.Rxh5+! Kxh5 47.Bf4 and according to
Garry, White must be able to hold.

42.c4

White must prevent the rook from coming to d2.

42...Rd1+ 43.Ke2
43...Rc1!?

Karpov keeps control very well when his opponent has passed pawns. Being in a must-win situation, he had to make a
move like this.
Karpov was 10 minutes late for the resumption of the game, as he had been looking for a win with 43...Ra1 44.Bc3! Rc1
(44...Ra2+ 45.Rb2; 44...Ra3 45.Bb2 Ra2 46.Rh1 f4 47.Rh4!) 45.Be5 and now:
A) 45...Rxc4 46.Rb6 Rc2+ 47.Ke1 and White gets pressure on e6:
A1) 47...Rd5 48.Bf4 Nh5? 49.Rxh5 Kxh5 50.Rxe6 and White wins;
A2) 47...Ra2!? (Garry doesn’t mention this) 48.Bxg7 Kxg7 49.Rxe6 Rc8 (49...Rb8 50.Rb6) 50.Kd1 is no fun for White,
but Kasparov’s team probably felt that this position could be held;
A3) 47...e3 48.fxe3 Rxg2 49.Bf4 Nh5 50.Rxe6+ Kf7 51.Re5 Rc8 52.Rxf5+ Ke6 53.Re5+ Kf6 54.Kd1 and White is out
of danger.
B) 45...Rc2+ 46.Ke1 Rxc4 (46...Rdd2 47.Rh6+!! Kg5 48.Rh7 Re2+ 49.Kd1 Ne8 50.Rb8 and Black has no more than a
perpetual) 47.Rb6 e3 48.fxe3 Re4 49.Rd6 (49.Bxg7? loses to 49...Rxe3+ 50.Kf2 Rdd3!!) 49...Rxe3+ (49...Ra8 50.Bd4
Rxa5 51.Rh8 Nh5 52.Rg8+ Kf7 53.Rdd8 and White has enough activity to draw) 50.Kd2 Rxe5 51.Rxd8 and White
holds.
This is Kasparov’s analysis from his Two Matches book.

44.a6?

Maybe White can still stay in the game after this push, but it does not make his defence easy and there was a better
move.
After 44.Bc5! Nh5 (44...Rxc4 45.Be3 Nh5 46.g3) 45.g3 (after 45.Be3 Rxc4 46.a6 Ra4 47.Rb6 f4 48.Bc1 f3+ Black
wins) 45...Rxc4 46.Be3 Ra4 47.Rb6 White escapes according to the later World Champion.

44...Rc2+!
Karpov forces Garry’s king to the back rank. According to Kasparov this was the best option, although Garry and his
team had analysed other possibilities as well:
A) 44...Rxc4 45.a7 Ra8 46.Rb7 Rxb4 47.Rxb4 Rxa7 48.g3 and White holds;
B) 44...Nh5 45.Rxh5 Kxh5 46.a7 Ra8 47.Bc5 Ra1 48.Rb7 and White is safe;
C) 44...Ra1 (in this line we can find mistakes with the aid of computer programs) 45.Bd2 (Garry considers no other
moves here. He wants to improve the bishop. After 45.Ba5!? Rd7 (45...Ra8 46.Bd2) 46.a7 Rxa7 47.Bd2 White’s pieces
are active; or 45.Rb6!? Ra2+ 46.Ke1 (46.Ke3 Rd3+ 47.Kf4 Rxf2+ 48.Ke5 e3 49.a7 Ra2 50.Rb7 e2 51.Rh1 Nh5 gives
Black decent chances) 46...Rd3 47.Bc5 – it is not easy to break open White’s stronghold. For example: 47...g3 48.fxg3
Rc3 49.Rd6 and White is still alive) 45...Ra2 46.Rh6+ Kf7

analysis diagram

C1) Kasparov only considers the natural 47.Rb7+, and now: 47...Kg8 48.Rg6 Raxd2+ 49.Ke3 R2d7 50.c5
analysis diagram

C11) 50...e5 51.Rd6 Rxd6 52.cxd6 Ne6 53.a7 f4+ 54.Kd2 Rxd6+ 55.Ke1 and White holds according to Kasparov;
C12) 50...Kh7!! (this intermediate move wins. Garry’s brigade (this word is borrowed from Kasparov) did not notice it.
Please bear in mind that they had no computer and had to cover a lot of ground) 51.Rf6 g3 (51...e5 52.Rxd7 Rxd7
53.Rd6 Ra7) 52.fxg3 (52.Rxd7 gxf2 53.Kxf2 Rxd7) 52...Rd3+ 53.Kf4

analysis diagram

53...e5+ 54.Kxe5 Re8+ 55.Kf4 e3 56.Rff7 Re4+ 57.Kg5 Rg4+ 58.Kf6 Rg6+ and Black wins.
C2) 47.Ke3!! and amazingly White can create counterplay with his queenside pawns and king: 47...Raxd2 (47...e5
48.Rd5 Rxd5 49.cxd5 f4+ 50.Kxe4 Rxd2 51.Kxe5 and White gets rid of Black’s pawns) 48.Rb7+ Kg8 (in case of
48...R2d7 49.c5 Re7 50.c6 White’s passed pawns keep him in the game, thanks to Black’s awkwardly placed knight)

analysis diagram

49.Kf4 R2d7 (after 49...Rxf2+ 50.Kg5 Rfd2 51.a7 Ne8 52.Rxe6 White survives) 50.c5 e5+ (in case of 50...Rf7 51.c6
Ne8 52.Rxe6 White holds as Black’s pawns are vulnerable and the knight can’t really find its place) 51.Kxe5 g3
52.fxg3 e3 53.Rxd7 Rxd7 54.Rh1 and White can defend.

45.Ke1 Ra2 46.Rb6?

Garry doesn’t like this move, but he thinks his position was troublesome anyway. He says that after he had spoiled his
middlegame advantage he had resigned himself to losing this game. I think even in that kind of mood he would have
saved the game against many ordinary players.
Kasparov mentions 46.Bc5 as a slight improvement. But there was a genuine improvement in 46.Ba5! Ra1+ (in case of
46...Ra8 47.Bd2 Nh5 48.Rb6 R2xa6 49.c5 White resists) 47.Ke2 Rd7 48.a7 Rxa7 49.Bd2 R1a2 (49...Nh5 50.Rb6)
50.Rh6+ Kf7 51.Ke1 and it is not easy to brush White aside.

46...Rd3!

Both Makarichev and Kasparov thought this was the start of a decisive attack.

47.c5

After 47.a7!? Black has real threats indeed: 47...e3! 48.fxe3 Rxe3+ 49.Kd1 Rb3 50.Kc1 (50.Bc5 Rd3+ 51.Ke1 Rxa7
52.Rxe6+ Nxe6 53.Bxa7 Nf4 and Black wins) 50...Rxa7 51.Rh6+ Kxh6 52.Bd2+ Kg6 53.Rxb3 and Black has winning
chances.
Alternatively, after 47.Bc5 g3! (47...f4 48.Rb1) 48.fxg3 Rxg3 49.Rd6 Rgxg2 50.a7 e3 51.Rh1 Nh5 52.Rd8 Nf4 53.a8=Q
Rxa8 54.Rxa8 Re2+! 55.Kd1 Rd2+ 56.Kc1 Nd3+ 57.Kb1 Nxc5 Black also wins.

47...Ra1+ 48.Ke2 Ra2+ 49.Ke1

49...g3?!

Kasparov’s recommendation 49...Kg5! is clearly better than the text. The great stakes take their toll, even on a fabulous
endgame player like Karpov. Let’s examine:
A) 50.Rh7 Nh5 51.Rxh5+ (51.g3 e3) 51...Kxh5 52.c6 (52.Rxe6 loses to 52...Rb3 53.Bd2 f4) 52...g3 53.fxg3 Rb3
54.Bc5 Rc3 and Black wins;
B) 50.c6 Kxh4 51.c7 f4 (this is Kasparov’s line. Black can win with another motif as well: 51...Ra1+ 52.Ke2 f4 53.f3
Ra2+ and mate follows) 52.g3+ Kh3 53.gxf4 Nf5 and White is caught, as Kasparov points out.

50.fxg3 Rxg3 51.Kf1!

Makarichev shows the win against 51.Rh2. There follows 51...f4 52.Kf1 and now:
A) 52...f3 53.a7 Nf5 wins;
B) 52...Rd3 53.Rd6 (53.Be1 Rd1) 53...Nf5 54.Rxd3 exd3 55.Ke1 Ne3 scores the point as well.

51...Rgxg2

Karpov liked to double rooks on the seventh rank even in his childhood.

52.Be1 Rgc2 53.c6 Ra1 54.Rh3 f4

54...Nh5 55.c7 Rxc7 56.Rxe6+ Kg5 wins as well.

55.Rb4

If 55.Rhb3 Nf5.

55...Kf5

According to Kasparov 55...Nf5 was simpler: 56.Rxe4 Ng3+ 57.Rxg3+ fxg3 and Black wins.
56.Rb5+ e5 57.Ra5

57...Rd1?!

Makarichev says this squanders the win. Garry is right; it just forces White to find more good moves.
57...Rac1! was more careful, as Black threatens ...e3 and ...Rf2+ and White cannot resist.

58.a7?

Kasparov goes down without putting up a fight.

58.c7 would have offered some practical chances, even though it would also lose: 58...e3 59.Rh2!! and now:
A) 59...Rxh2. Commentators at the time thought that White could escape after this. However, today, computer-
equipped annotators can spot holes in old analyses: 60.c8=Q+ Ne6 61.Qc3 Rf2+! (this intermediate check frees the e2-
square for the black king. After 61...Nd4? 62.Qc8+ Ke4 63.Qb7+ White has a perpetual check at his disposal) 62.Kg1
Nd4 63.Qc8+ Ke4 64.Qb7+ Kd3 65.Ra3+ Ke2 and Black wins;
B) 59...Rcc1! (the simpler way) and now:
B1) 60.c8=Q+ Rxc8 61.a7
analysis diagram

61...f3! 62.Re2 (62.a8=Q Rxe1+! is a lovely mate) 62...fxe2+ 63.Kxe2 Rdd8 and Black wins;
B2) 60.Re2 Ne6 61.a7 Nxc7 62.a8=Q Nxa8 63.Rxa8 Ke4 64.Ra3 Rc8 and Black wins according to Kasparov.

58...e3

Kasparov resigned as he will be checkmated now: 59.Rf3 Nh5 60.a8=Q Ng3+ 61.Rxg3 Rf2+ 62.Kg1 Rxe1 mate.

GAME 90
Portisch, Lajos
Karpov, Anatoly
Tilburg 1986 (2)

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Bb7 5.Bg2 Be7 6.0-0 0-0 7.Nc3 Ne4 8.Nxe4 Bxe4 9.Nh4 Bxg2 10.Nxg2 d5 11.Qa4
dxc4 12.Qxc4 c5 13.Be3 cxd4 14.Bxd4 Qc8 15.Rac1 Qxc4 16.Rxc4 Rd8 17.Be3 Nd7 18.Rfc1
18...Kf8

White controls the c-file but he cannot invade. Karpov finds no pieces to improve, so the other useful plan is to bring the
king closer to the centre. It is hard to imagine that this innocentlooking position can be won by Black.

19.Ne1 Bd6

According to Karpov this was a novelty. However, these two opponents had already played a game like this. Six years
earlier they had drawn after 19...Nf6. Maybe Karpov was slightly worried about a deviation that would have given
Portisch a position in which he could press. It is also possible that he was already toying with the idea of giving it a try
to win.
Portisch-Karpov, Tilburg 1980, had continued (19...Nf6) 20.Nf3 Nd5 21.Bd4 f6 22.e4 Nb4 23.a3 Nd3 24.R1c2 e5
25.Be3 draw.

20.Nf3

Interestingly, Karpov attaches an exclamation mark to this move – we are in the opening and in the endgame at the same
time.
Preferable was 20.f4!?, which prevents ...Ne5 and gains some space: 20...Nf6 21.Nd3 Nd5 22.Bd4 and objectively the
position is equal. However, Black is the one who has less opportunities to go astray.

20...Ne5

Slightly inconvenient for Black is 20...Ke7?! 21.Nd4.

21.Ra4

Portisch doesn’t reduce the position to a dead draw by taking on e5.

21...Nxf3+

Karpov has achieved something by doubling the white pawns. Of course it is still far from anything real – just
something.

22.exf3 Rdb8!?

Interestingly, Karpov calls this an ‘only’ move. He probably means that this was the only active move here.

23.Rc6

After 23.Kg2 Ke7 24.f4 Rb7!? (24...a5 25.Rc6!?) 25.Kf3 a5 26.Rac4 (26.Rc6 Rab8) 26...b5 27.Rc6 Kd7 Black has
nothing to worry about.
Likewise, after 23.Ra6 Rb7 24.Rc4 Rd8 25.Rca4 Rdd7 Karpov claims the position is equal.

23...Ke7 24.b4

Portisch exerts pressure, but he is taking some risks as well. This pawn might become a target, especially as Karpov’s
king is close by.

24...Kd7 25.b5 e5 26.Rc2?!

Portisch makes sure Karpov cannot play against the c6-square with ...a7-a6. However, this costs time and now Karpov
can solve all his problems on the queenside.
After 26.f4!? Black can kill the position with 26...a6 (after 26...Rd8 27.f5 Karpov prefers White) 27.fxe5 Bxe5 28.Rxb6
axb5 29.Rxa8 Rxa8 30.Rxb5 Ke6 31.Rb6+ Kf5 32.Rb7 Rxa2 33.Rxf7+ Ke6. White is a pawn up and he can play on,
but he cannot win.
Dramatic would be 26.Ra6 Bb4 when Black can trap the rook in an unusual way. After 27.Rc4 Ba5 Black must have a
good game.
26...Rb7!

A small improving move by Karpov.

27.Rd2 Ke6 28.f4 exf4 29.gxf4?!

Portisch still pins his hopes on his grip on the queenside. He voluntarily accepts the doubled pawns for this.
After 29.Bxf4 Bxf4 30.Rxf4 Rc8 the four-rook endgame would have been equal.

29...f5!

This fixes the f-pawn on the bishop’s colour.

30.Rad4 Rd8 31.Kf1 Rc7 32.Re2?!

The battery is a paper tiger. There is no useful discovered check by the bishop. According to Karpov, Portisch should
have reserved this square for his king.

32...Rdd7 33.a4?

White has made moves that can be criticized; still up until this point the position has remained equal. Portisch now
allows the exchange of a pair of rooks, after which Karpov has better chances to use his king.
It would have been hard for Black to make progress after 33.Rd3!? Kf7 34.Red2.
33...Bc5!

After the exchange of the bishops the f4-pawn will grow weaker.

34.Rxd7 Kxd7 35.Rd2+ Ke6 36.Rc2 Kd5

Normally, with the better king, Black would like to exchange rooks. But 36...Bd6?! would have helped the weaker side
here: 37.Rxc7! (37.Rd2? Rc4; 37.Re2 Kd7) 37...Bxc7 38.Ke2 Kd5 39.Kd3 – whoever stands worse in same-coloured
bishop endings usually loses, but here Black cannot invade.

37.Kg2 Re7 38.Kf3 g6 39.h4 Rc7

Karpov plays for the exchange of rooks as White’s king on f3 doesn’t bother him. Black could have started playing on
the queenside with 39...Re4 40.Rd2+ Ke6 41.Rc2 (after 41.Bxc5? bxc5 42.Re2 Rxe2 43.Kxe2 Kd5 Black wins)
41...Bd6 42.a5 bxa5 43.Rc6 Kd5 44.Ra6 a4 45.Rxa7 a3 with an advantage.

40.Rc1 Bd6 41.Rd1+ Ke6 42.h5

White must take action on the kingside, otherwise Black’s rook would invade on the c-file, e.g. after 42.Rd4 Rc3.

42...gxh5 43.Rh1 Bc5?!

This will lose a tempo.

44.Rxh5 Bf8!?

For the first time we see Karpov changing his mind. One of his most remarkable character traits in the endgame is that
he almost invariably picks up a plan and sticks to it. This time he wastes a tempo before he starts executing his plan.

45.Rh1 h6 46.Ra1
46...Bg7!

Karpov knocks Portisch off-balance. The white rook stands well on a1, from where it prepares a4-a5 and can return to
the h-file quickly. Now Portisch has to abandon one of these options.
The black rook could have invaded with 46...Rc3, but this would have beeb too hasty: 47.a5 Bc5 48.axb6 axb6 49.Rh1
and White holds.

47.Rg1

After 47.Ra2 Karpov suggests 47...Bc3 to stop a4-a5; he gives the assessment minus/plus. This is worth a closer look.
White must do something about the king’s invasion on the c-file:
A) 48.Ke2 Kd5 49.Kd3 Bb4 50.Ra1 Rc3+ (50...h5 51.Rg1 gives White counterplay) 51.Ke2 Kc4 52.Rh1 Ra3 53.Rxh6
Rxa4 54.Rf6 Kxb5 55.Rxf5+ Kc4 – an unusual race. Black should be better;
B) 48.Rc2! Kd5 and now:
B1) 49.a5? bxa5 50.b6 (50.Bxa7 Rxa7 51.Rxc3 Rb7 loses) 50...axb6 51.Bxb6 Rc6 and Black manages to keep one pawn
on the queenside. This wins according to Karpov;
B2) 49.Bd2 Kd4 (in case of 49...Be5 50.Ra2 Bd4 51.a5 (51.Be3 Bc3) 51...Rc5 52.axb6 axb6 53.Ra7 Rxb5 54.Be3
White has decent chances of survival) 50.Rc1 Rc4 (50...Kd3 51.Bxc3 Rxc3 52.Rh1) 51.Be3+ Kd3 52.Rh1 and Black is
pressing, but his win is not certain;
B3) 49.Rc1 (surprisingly, White can even wait and see. It is not easy for Black to make progress) 49...Rc4 (49...h5
50.Kg3; 49...Kc4 50.a5) 50.a5 Bxa5 (after 50...bxa5 51.Bxa7 a4 52.Rd1+ Ke6 53.Rc1 a3 54.b6 a2 55.b7 White holds)
51.Rd1+ Ke6 52.Rh1 Bc3 53.Rxh6+ Bf6 and White is still worse, but he has chances to escape.

47...Kf7 48.Rd1
48...Bf6!

Not a sparkling move, yet a strong one. It looks like just another move, but Karpov closes the sixth rank.

49.Rb1?

Karpov attaches a question mark to this move, but suggests no alternative. Porstisch hopes to survive. White cannot
force matters by playing 49.Rd6 as a prelude to a4-a5. After 49...Ke7! 50.Rd1 Rc4 51.a5 bxa5 52.Bxa7 a4 53.b6 Rb4
Black is better.
Going after the h-pawn with 49.Rh1 looks like White’s best chance: 49...Rc4 (after 49...Kg7 50.Rg1+ Kh7 51.Rd1! h5
52.Rd6 White is rather active) 50.Rxh6 Rxa4 – Karpov would probably have settled for this slightly better ending.

49...Rc4 50.a5 bxa5 51.Bxa7

51...Rb4!

Karpov has finely assessed the ensuing bishop ending.

52.Rxb4

If 52.Rh1 a4!.

52...axb4 53.Ke3

Portisch probably overestimated the power of his king.

53...Be7 54.Kd4 h5!

Karpov pushes the h-pawn quickly, before White can achieve something with his king. Portisch probably did not fully
appreciate the danger of the h-pawn.
55.b6 b3

Karpov wins in the most straightforward manner. The champion mentions 55...h4. Interestingly, when he plays
endgames he hardly makes any mistakes, but when it comes to analysing the story is a little different: 56.b7
A) 56.Ke5 Bc5–+;
B) 56.Kd5 and now:
B1) 56...Bd8? 57.b7 Bc7 – Karpov stops here.

58.Kc6!! (the active king forces promotion) 58...Bxf4 59.Bb6 and now Black has several ways to draw: 59...Bb8
60.Bc7 Ba7 61.Kb5 b3 62.Ka6 b2 63.Kxa7 b1=Q 64.b8=Q Qa2+ 65.Kb7 Qxf2 and White loses his last pawn, or
59...h3 60.Bc7 Bxc7 61.Kxc7 h2 62.b8=Q h1=Q;
B2) 56...h3! (the champion misses this simple win in his Chess Informant analysis) 57.b7 h2 58.b8=Q h1=Q+ 59.Kc4
Qe4+ 60.Kb3 (60.Kb5 loses to 60...b3) 60...Qd3+ 61.Ka4 Qa3+ 62.Kb5 b3 and White is lost.
C) 56.b7 Bd6 57.Kd5 Bc7 (57...Bxf4?? 58.Kc4! h3 59.b8=Q) 58.Bc5?? (here is another strange mistake in the
winner’s analysis of this game. 58.Kc6 would transpose to the variation 56.Kc6 Bd8 which leads to a draw) and now:
C1) Karpov evaluates 58...h3 59.Bd6 h2 60.Kc6! Bb8! 61.Bxb8 h1=Q+ 62.Kb6 as unclear. However, White loses by
force: 62...Qd5 63.Be5 (or 63.Bc7 b3 64.b8=Q b2) 63...Qd8+ 64.Kb5 (in the event of 64.Ka7 Qa5+ 65.Kb8 Ke7
66.Bc7 Qa2 wins) 64...b3;
C2) 58...b3! (it was not Karpov’s day when he made this analysis. He missed this relatively easy win as well) 59.Bd6
b2 60.Bxc7 b1=Q 61.b8=Q Qxb8 62.Bxb8 h3 and Black’s h-pawn is not to be stopped.
One may ask how it is possible that a genius of the endgame can produce this level of analysis. Well, the players at the
very top are millionaires and the money they receive for written analysis is minimal. Another possibility is that Karpov
just accepted analysis that was undertaken on his behalf.
With 55...Bd6 Black even had another win here: 56.Kd5 b3 (56...Ke7 57.b7 Kd7 also wins) 57.Kxd6 b2 and Black will
promote to two queens, of which one survives.

56.b7 Bd6 57.Kc3

After 57.Bc5 Bb8 58.Kc3 h4 59.f3 h3 wins.

57...h4 58.Bd4

58...Bb8!!

When it comes to accuracy on the board, Karpov is without equal.

59.Be5 Ba7

After 59...b2 60.Kxb2 Ba7 61.Kc3 (or 61.Kc2 h3 62.Bd4 Bb8! etc.) 61...h3 62.Bd4 Bb8 63.Be5 h2 64.Bxb8 h1=Q
wins.
60.Bd4 b2 61.Kc2

61...Bb8!

Portisch resigned.
The motif is nice: Black sacrifices the bishop to block the route of the pawn. The tempo that White has to lose does the
trick. Had Portisch played on, Karpov would have won another tempo with a well-known finesse: 62.Be5 h3 63.Bxb8
h2 64.Be5 b1=Q+ (here is the second tempo) 65.Kxb1 h1=Q+.
 GAME 91
Karpov, Anatoly
Kindermann, Stefan
Vienna Open 1986 (2)

1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 d6 6.g4 h6 7.h4 a6 8.Rg1 d5 9.exd5 Nxd5 10.Nxd5 Qxd5 11.Bg2 Qc4
12.c3 Be7 13.g5 Nd7 14.Qe2 Qxe2+ 15.Kxe2 Nb6 16.Bf3 hxg5 17.hxg5 Rh4 18.Rh1 Rxh1 19.Bxh1 e5 20.Nc2
Bg4+ 21.Bf3 Bxf3+ 22.Kxf3 0-0-0 23.Ke4 Nd7 24.Be3 Nc5+ 25.Kxe5 Nd3+ 26.Ke4 Nxb2 27.f4 Nc4 28.Rh1 g6
29.Rh7 Nd6+ 30.Kf3 Kd7 31.Bd4 Ke6

32.Be5

Karpov has obtained little out of the opening. Now he wants to bring out his knight.

32...Nf5

Black prepares to eliminate the knight.

With 32...Rc8 Stefan could have put pressure on Karpov’s weakness. After 33.Nd4+ Kd5 34.Nb3 Nc4 35.Bd4 Ke6
Black is not at all worse.

33.Nd4+ Nxd4+ 34.cxd4 b5?

The German player pins his hopes on pushing his queenside pawns. Karpov elegantly reveals the drawback of this plan.
Better was 34...Rd7! when Black can keep White’s king away from the centre. In case of 35.Rh8 (35.Ke4 Bxg5!!)
35...Kd5 Black has a healthy game. If 35...f6 36.f5+!?.

35.Ke4!

Karpov’s king becomes surprisingly powerful. Actually it decides the outcome, as Black can probably no longer save
the game.
35...b4

Not 35...Rd5? as 36.Rh6! prepares f4-f5+.

36.Rh2!

Neatly changing the direction of his attack. Now the rook heads for the c-file.

36...Kd7

After 36...Rc8 37.d5+ Kd7 38.Rh7 Rf8 39.f5 gxf5+ 40.Kxf5 a5 41.Rh6 Black is struggling.

37.d5 Bc5?!

Kinderman closes the c-file, but this loses without forcing White to calculate accurately.
A) 37...Rc8 38.Rh7 Rf8 39.f5 gxf5+ 40.Kxf5 leads to the position we have seen;
B) 37...a5, pushing the pawns, was slow as well: 38.Rc2 Rc8 and now:
B1) 38...a4 39.Rc7+ Ke8 40.Rxe7+! (this simplification is decisive) 40...Kxe7 41.Bf6+ Ke8 42.Bxd8 Kxd8 43.Kd4
Kd7 44.Kc4 b3 (or 44...Kd6 45.Kxb4 Kxd5 46.Kxa4) 45.axb3 a3 46.Kc3 Kd6 47.b4 Kxd5 48.Kb3 Ke4 49.b5 and
White wins;
B2) 39.Rxc8 Kxc8 40.d6 Bd8 (40...Bf8 41.Kd5 Kd7 42.Kc5)
analysis diagram

41.Kd5. This is a nice example of domination by the king. It wins by just one tempo: 41...Kd7 42.Kc5 Ke6 43.Kc6 a4
44.d7 b3 45.axb3 axb3 46.Bf6 Ba5 47.Bc3 Bd8 48.Be5! Ba5 (48...Be7 49.Kc7) 49.Bc7 Bxc7 50.Kxc7 b2 51.d8=Q
b1=Q 52.Qd7 mate, or 46...Bxf6 47.gxf6 b2 48.d8=Q b1=Q 49.Qd5+ Kxf6 50.Qe5 mate. In both lines White
checkmates just in time.

38.Rh7!

The c-file is closed, but now the seventh rank is weakened. Karpov wastes no time in reverting to the earlier position.

38...Rf8 39.f5! gxf5+ 40.Kxf5 Ke8 41.Rh6! 1-0

GAME 92
Karpov, Anatoly
Sokolov, Andrey
Linares Candidates’ Final 1987 (6)

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.b3 Bb4+ 6.Bd2 Be7 7.Nc3 d5 8.cxd5 Nxd5 9.Nxd5 exd5 10.Bg2 Nd7 11.0-0
0-0 12.Rc1 Re8 13.Re1 c5 14.Be3 Bb7 15.Bh3 cxd4 16.Bxd4 Nf6 17.Rc2 Bb4 18.Rf1 Ba6 19.Nh4 Bf8 20.Nf5 Ne4
21.Ne3 Qd6 22.Qc1 Rad8 23.Rd1 Qh6 24.Bg2 Ng5 25.Qb2 Nh3+ 26.Kf1 Ng5 27.Kg1 Nh3+ 28.Kf1 Ng5 29.h4
Ne4 30.Ng4 Qe6 31.Bh3 Ba3 32.Qxa3 Bxe2+ 33.Rxe2 Nxg3+ 34.Kg2 Qxe2 35.Qc1 Nh5 36.Kh2 Rd6 37.Qd2 Qf3
38.Ne5 Qf4+ 39.Qxf4 Nxf4
40.Bd7!

Karpov improves his bishop and makes Andrey’s rook passive by driving it away from the e-file. White is better. He has
two pieces for a rook and two pawns, whereas Black’s rooks are not active and he has no passed pawns. White can
improve his pieces.

40...Rd8

Not 40...Re7? 41.Rc1!.

41.Bb5 Ne6

Maybe Black should have occupied the c-file at once with 41...Rc8!?.

42.Bb2 a6

Karpov likes this move. He won’t try to refute it, so he transfers the bishop to the other diagonal, improving it.

43.Bf1 Rc8 44.Rd2 d4 45.Bg2 Nf4 46.Bf3 d3 47.Nc4 Rh6?!

This just helps White to bring his king closer. After 47...Rdd8 48.Kg3 Ne6 White would have had to work to get at the
d-pawn.

48.Kg3 Ne2+ 49.Bxe2 dxe2

Sokolov gives back a pawn to open up the position for his rooks.

50.Rxe2 b5 51.Ne3 Re6

After 51...Rd8 52.Nf5 Re6 53.Rc2! White keeps his edge.

52.Rd2!

Of course Karpov occupies the open file.

52...h5 53.Kf4!

This is an unpleasant move. It improves the king and forces Black to reckon with Nf5 with an attack on his king.

53...Rc7 54.Rd5!

Provoking a pawn move, but the text has another aim as well.

54...f6 55.Ba3

Kasparov points out that it would not have been smart to take the h5-pawn, as Black’s e-rook would then occupy the d-
file.

55...g6

This prevents the enemy king from penetrating on the kingside.

56.Bc5!

Now we can see the other point of White’s 54th move. Karpov improves his bishop and makes the black c-rook passive.
At the age of 15 Karpov accomplished a similar bishop improvement similar against this pawn formation in a game
with Drizgalovich (see Game 7).

56...Kf7 57.b4!

Defending the bishop and fixing the a-pawn.

57...Ke8 58.Rd1!

Karpov changes the target.

58...Rd7
59.Ra1!

He spots and uncovers a new weakness.

59...Rd2 60.Kf3 Kf7 61.a4 bxa4

The a6-pawn is less vulnerable than the b5-pawn would have been.

62.Rxa4 g5?!

Sokolov looks for counterplay on the kingside. Maybe he hopes to exchange enough pawns to be able to draw.
Doing nothing would have run into the manoeuvre Ne3-c4-b6, winning a6. However, with 62...Rd3!? Black could have
tied White’s hands on the third rank: 63.Kf4 (63.Kg2 Ke8!?) 63...Rd2 (Black can try to keep waiting with 63...Kg7;
after 64.Nc4 Rh3 things become unclear) 64.Nc4 Rd5 65.Nb6 Rf5+ 66.Kg3 Re4 and Black is alive. After 67.f3 Ref4
68.Ra3 g5 it is not clear whether White is still better.
63.Ra3!

Karpov prevents the rook from causing problems on the third rank. By waiting with hxg5 he improves his chances of
keeping the f-pawn.

63...Kg6 64.Kg2 gxh4?

Karpov analyses this game with Igor Zaitsev in the Chess Informant. They attach a ?! here, suggesting that Black should
have done nothing but wait and try to keep his position together. Their evaluation of plus/equal suggests that this was
feasible. Kasparov also thinks that Black should be able to live with his position.

65.Nf1 Rc2 66.Nh2 Rc4 67.Nf3 Kf7 68.Kh2

Karpov keeps the tension. With 68.Nd4 Re5 69.Nc6 White could have forced matters at once, but after 69...Rg5+
70.Kh2 Rgxc5 71.bxc5 Rxc5 72.Rxa6 Rc2 73.Kg1 h3, according to Kasparov White is unlikely to retain his last pawn.

68...Rf4 69.Kg2 Ree4

Players rarely analyse their lost games. It would be interesting to know why Andrey did not repeat moves here with
69...Rc4.

70.Nd2

Not 70.Rxa6? as the a6-pawn is poisoned: 70...h3+ 71.Kg3 h4+ wins.

70...Re6 71.Nf1 Rg4+ 72.Kh3 Rf4

73.Ra2 Rf3+

Andrey decided to sit and wait after taking the h-pawn, and now he changes his mind. He prefers his rook to be on the
third rank. He could have held the rook on the fourth rank for longer with 73...Kg6!?.

74.Kg2 Rd3 75.Ne3 Kg6 76.Ra1 f5

This move earned another ‘?!’ from Karpov and Zaitsev. The young grandmaster wants to attain the draw with active
play. One possible waiting move was 76...Rd7.

77.Kh2!
Karpov always recognizes his chances against the opponent’s king. He opens the g-file for his rook. On f3 the king
would be less effectively positioned.

77...f4 78.Rg1+ Kf7 79.Ng2! Rc6?

Andrey voluntarily gives up the f-pawn. He will be able to exchange his a-pawn, but still he will lose.
A) After 79...Rg6 comes 80.Ra1! (White has better chances with his rook on the board, as this allows him to pose the
black king problems. In case of 80.Nxf4 Rxg1 81.Kxg1 Rc3 Black would have real chances of survival) 80...Rg4
(80...Rf6 81.Nxh4) 81.Rxa6 Rb3 82.Rh6 f3 83.Ne3;
B) Keeping the f-pawn on with 79...Rf6 would probably have put up more resistance, e.g. 80.Nxh4 Rd5 81.Ra1 Rd3.

80.Nxf4 Rf3 81.Nh3 Kf6

Black cannot carry out ...a6-a5 with 81...Rf5. There follows 82.Ng5+ Kf6 83.Ne4+ Ke5 (83...Ke6 84.f3) 84.Nd6 Rf6
85.Nc4+ Kd5 86.Na5 and White should gradually win by capturing the h4-pawn and pushing his f-pawn.

82.Re1!

Karpov uses his rook to create threats with great effect.

82...Rf5?!

It was not Andrey’s day. He should not have kept his king on the dark squares.
A) Stepping aside with 82...Kg6 gave practical chances:
A1) 83.Re4 a5 84.Be3! (otherwise Black gets a dangerous a-pawn with ...a5-a4) 84...a4 85.Kg2! (85.b5? Rc2!) 85...Rf5
86.b5 Rxb5 87.Rxa4 and White wins;
A2) 83.Re5! Rf5 84.Re4 a5 85.Be3 a4 86.b5 Rxb5 87.Rxa4 – White wins;
B) With 82...Kf5!? Black could prevent the white rook to get to the h4-pawn. After 83.Rd1 Ke5 White still has to work
for the win.
83.Re4 a5 84.Be3 axb4

If 84...a4 85.b5.

85.Rxb4

This looks like a difficult technical win. But with a few very strong moves Karpov makes it look easy.

85...Ra6 86.Rxh4 Raa5 87.Rc4 Rfb5 88.Kg3!

A nice employment of the king.

88...Ra8
89.Kh4! Rg8 1-0

The game was adjourned here. Sokolov resigned only after the seventh game of the match.

GAME 93
Karpov, Anatoly
Sokolov, Andrey
Linares Candidates’ Final 1987 (10)

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.b3 Bb4+ 6.Bd2 Be7 7.Nc3 0-0 8.e4 d5 9.cxd5 Bxf1 10.Kxf1 exd5 11.e5 Ne4
12.Qe2 Nxc3 13.Bxc3 Qd7 14.Kg2 Nc6 15.Rhe1 Nd8 16.Ng1 c5 17.f4 cxd4 18.Bxd4 Qf5 19.Rad1 Bb4 20.Rf1 Ne6
21.Qd3 Qxd3 22.Rxd3 Rac8 23.Nf3 Rc2+ 24.Rf2 Rfc8 25.f5 Nxd4 26.Nxd4 Rxf2+ 27.Kxf2 Rc1 28.g4 Kf8 29.Kf3
Rf1+ 30.Kg3 Rc1 31.Kf4 h6 32.h4 Ke8 33.Nf3 Rc2 34.a4 Rb2 35.Nd4 Be7
36.h5!?

Karpov felt that this was a moment that required responsible decision-making. He had used up much of his time. It
looks as though the text was not the best option. By pushing the h-pawn it can in certain situations – as in the Caro-
Kann – become a dangerous passed pawn.
According to Podgaets, who worked for Karpov, 36.g5!?, gaining space with f5-f6 later on, deserved consideration as
well. This position caught Kasparov’s imagination and he analysed it deeply: 36...hxg5+ 37.hxg5 and now:
A) 37...Rb1 38.g6!? (after 38.f6 Kasparov found a remarkable way for Black to generate counterplay: 38...gxf6 39.gxf6
Bc5 40.Nf3 a5 41.Kf5 Rf1 42.e6 Bd4!) 38...fxg6 39.fxg6 Rg1 40.Nb5 – Black has chances to hold this difficult
endgame because of the limited material on the board, as Kasparov points out;
B) 37...a6 38.g6!? (after 38.f6 gxf6 39.gxf6 Bc5 40.Nf3 Ba3 41.Kf5 Rf2 42.e6 Bb2 Black can hold, according to
Kasparov) 38...fxg6 39.fxg6 Rg2 40.Ne6 and Black is still suffering here.

36...a6
37.Kf3!!

It was not so easy to anticipate this witty move. Karpov energizes the position, with little time left on his clock. It may
not be enough for an advantage objectively, but it was enough to defeat his opponent. In over-the-board chess this
approach is practical.
Why is Karpov retreating his king? It is a square-clearing move.

37...Bc5?

Kasparov found that Black’s best option here was to change the target with 37...Rh2!, but this move is really hard to
come up with over the board, when there is another sensible move. After 38.Nc6 Rh3+ 39.Ke2 Rh2+ 40.Ke3 Rh3+
41.Kd2 Rh2+ 42.Kc3 Rg2 the result of the game would still be unclear according to Kasparov.

38.Ne2!

Maybe objectively this is not the best, but it follows the theme and it’s a very unpleasant move to face in time-trouble.
Karpov, also in time-trouble (which became an increasing concern for him in the mid-1980s), perhaps missed 38.Nc6!!
– a lovely move suggested by Kasparov. Perhaps Karpov saw no point in sacrificing material. There can follow:
A) 38...d4 39.g5 hxg5 40.b4 Be7 (40...Bxb4 loses to 41.Rxd4 Be7 42.h6) 41.h6 gxh6 42.Rxd4 and White’s central
domination is too much for Black;
B) 38...Be7 39.b4 Bxb4 40.g5 Be1 41.Rxd5 Rd2 42.Rxd2 Bxd2 43.g6 and White wins, as Kasparov points out;
C) 38...Rf2+ 39.Kg3 Rb2 (after 39...Rc2 40.Rxd5 Bf2+ 41.Kf3 Rxc6 42.Kxf2 Rc3 43.Rd6 Rxb3 44.f6 White wins
according to Kasparov) 40.g5 hxg5 41.Rxd5 Rxb3+ 42.Kg4 Be7 43.h6 gxh6 44.f6 and White wins. In all Kasparov’s
lines White’s h-pawn plays an important role.

38...d4?!

Black still should have repositioned the rook to cause problems on the kingside with 38...Rb1!. This was also
recommended in Kasparov remarkable analysis, e.g. 39.Nf4 Rf1+ 40.Kg3 Rg1+ and the rook is a nuisance.
39.Nf4

In time-trouble one doesn’t like to make such a committing decision as 39.f6 gxf6 40.exf6 and now:
A) 40...Rb1 41.g5 hxg5 42.h6 g4+;
B) The exemplary improvement 40...Kd7! saves the game according to the 13th World Champion:
B1) The breakthrough 41.g5 is not decisive: 41...hxg5 42.h6 g4+ 43.Kf2 Rb1! (43...g3+? 44.Kf3 g2 45.Ng1) 44.Ng3
Rb2+ 45.Kg1 Rb1+ 46.Nf1 Re1 and the active rook keeps Black in the game;
B2) 41.Nxd4 Bxd4 42.Rxd4+ Ke6 43.Rb4 Kxf6 (or 43...b5 44.axb5 axb5 45.Rxb5 Kxf6) 44.Rxb6+ Kg5 45.a5 Ra2
46.b4 Ra4 (Kasparov calls this a miracle draw) 47.Ke4 Kxg4 and Black won’t lose the race.

39...Kd7?!

This loses a tempo, although it gains a move in order to reach the time control. 39...a5 would yield a tempo, but it would
not solve Black’s problem.

40.e6+ Ke8 41.Ke4!

Now Karpov invades with his king.

41...a5

42.Rf3!!

This sealed move was the result of a 20-minute think. It is another sensational move that is hard to anticipate and also
hard to understand. Karpov plays against the enemy king and tries to force matters with his passed pawn. Very original
play by Anatoly, who found it pleasing and rated the text as the best move after the game.

42...Rb1 43.Nd5 Rg1?

Sokolov attacks when he should have defended against Karpov’s threat. 43...Kf8! was the only move to stay in the
game, for example:
A) 44.Nc7!?. An attractive option; this might even be better than the move given by Karpov’s second in B).
Interestingly, the analysis depicted below was not confirmed until Karpov’s 2006 book – he had not mentioned it
before: 44...Re1+ (in the event of 44...Rg1 45.f6 Rxg4+ 46.Kd3 gxf6 47.Rxf6 Rg7 48.Nd5 White is clearly better)
45.Kd5!? (Karpov only mentions 45.Kd3, after which 45...Re3 holds) 45...Re3 46.Rf1 Ke7 (46...Rxb3? loses to 47.f6)
47.Kc6! fxe6 (after 47...Rxb3 48.Nd5+ Ke8 49.f6 Black’s position is difficult) 48.fxe6 Rxe6+ 49.Nxe6 Kxe6 50.Rf5
and White has winning chances;
B) Mikhail Podgaets gives a remarkable line with a stunning king march: 44.e7+ Bxe7 45.Kxd4 Rg1

analysis diagram

46.Kc4!! (a surprising, lovely king invasion) 46...Rxg4+ 47.Kb5. Here Podgaets stops and calls the position plus/minus.
Let’s take a closer look:
B1) 47...Bd8 48.Rc3 (or 48.Kc6! Rd4!? 49.Rc3 and White is slightly better) 48...Rh4 49.Kc6 Rxh5 50.Kd7 Rxf5
51.Nxb6 g5 52.Kxd8 g4 53.Nc4 h5 and Black is not at a disadvantage in the race, according to Kasparov;
B2) 47...Rd4 48.Nxb6 Bb4 and Black is still in the game. According to Kasparov, the blockade is impregnable.

44.Kd3! Rxg4

After 44...Kf8 45.Rf4 Rd1+ (or 45...Rg3+ 46.Kc4 d3 47.Kc3 Ke8 48.e7 and White is close to winning – Kasparov)
46.Ke2 Rb1 47.Re4! White wins.
45.f6!

The strong pawns decide the outcome. Karpov not only builds up the position well, he also controls the struggle when
the pieces come into contact.

45...Bd6

Forestalling the check on c7. In case of 45...fxe6, 46.f7+ Kd7 47.Nxb6+ Kc6 48.Nc4 wins.

46.Nxb6!

The strong pawns allow White time to grab a pawn on the queenside.

46...Rg5 47.fxg7 Rxg7

47...fxe6 loses to 48.Rf6.

48.Nc4 Bb4 49.exf7+ Rxf7 50.Rxf7 Kxf7 51.Ne5+ Kf6

52.Nc6

White is faster in the race.

52...Be1 53.Nxd4 Bb4 54.Nc6 Be1 55.Ke2 Bc3 56.Kd3 Be1 57.Kc4 Kg5

58.Nxa5!

This is a cute shot to clear the way for his pawns.

58...Bxa5 59.b4 Bd8 60.a5 Kxh5 61.Kb5 Bg5 62.a6 Be3 63.Kc6 1-0
Because of its imaginativeness, this is a special masterpiece, even by Karpov’s standards. A superb build-up rounded off
with very sharp tactical shots.

GAME 94
Meulders, Richard
Karpov, Anatoly
Brussels SWIFT 1987 (3)

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 c5 5.Bd3 Nc6 6.Ne2 d5 7.cxd5 Nxd5 8.a3 cxd4 9.axb4 dxc3 10.bxc3 0-0 11.e4
Nb6 12.Bf4 e5 13.Be3 Be6 14.Bc5 Re8 15.Bb5 Qc7 16.0-0 Nc4 17.Qc2 b6 18.Bxc4 Bxc4 19.Be3 Nd8 20.Rfd1 Nb7
21.Ng3 Nd6 22.Nf5 Nb5 23.Bd2 Qc6 24.Ne3 Rec8 25.Be1 f6 26.f3 Be6 27.Rdc1 Nd6 28.Qa4 Qxa4 29.Rxa4 Kf7
30.Raa1 Rc7 31.Kf2 Rac8 32.Bd2 Bc4 33.Ke1 Ke6 34.Kf2 Ra8 35.Be1 a5 36.bxa5 bxa5 37.Rcb1 Ra6 38.Rb8 a4
39.Nc2 Raa7 40.Rb6 Rab7 41.Rxb7 Rxb7 42.Nb4 Ra7 43.Nc2 Bb3 44.Na3 Nc4 45.Nxc4 Bxc4 46.Ra3

46...Kd6!

Karpov marches his king to defend the a-pawn. Thereby, he frees the rook and keeps the bishop on c4, where it cuts off
White’s king. Black has a winning position mainly because of his distant passed pawn.

47.Bd2 Kc6 48.Bc1 Kb5 49.Ra1 Rd7!

Of course he occupies the only open file.

50.Ba3

White has to allow the invasion on the second rank.

After 50.Ke1 Bb3 51.Ba3 Kc4 52.Bb4 and now:
A) 52...Bc2 53.Ra3 Rd1+ 54.Ke2 Rg1 (54...Rh1 55.Bf8) 55.Kf2 Rb1 56.Bf8 Rb7 57.Ke2 (57.Bb4? Rxb4!) – of course
Black has not spoiled anything, but he has not been able to force a win either;
B) 52...Kd3!? 53.Rc1 Ra7 54.Bc5 (after 54.Ba3 Bc2 55.c4 Kxc4 56.Rxc2+ Kb3 Black wins) 54...Ra8 55.Ba3 h5
56.Kf2 h4 57.h3 (57.Ke1 h3) 57...Be6 58.Rb1 Ra7 59.Bc5 Ra6 Black wins.

50...Rd2+ 51.Kg3 h5! 52.h4 Bb3 53.Bf8 Rd8!

Karpov often makes use of little tactics in endgames.

54.Bb4

54...Kc4!

Bringing the king closer to enable his a-pawn to move.

Marin points out that it would have been premature to exchange rooks with 54...Rd1 because of 55.Rxd1 Bxd1 56.Kf2
Kc4 57.Ke1 Bc2 58.Kd2 Bb1 59.Kc1 Bd3 60.Kb2 (60.Bf8 Kb3) 60...Bf1 61.Bf8 and White holds.

55.Rf1 Rd1 56.Rf2

Compared to the previous variation, the extra tempo provides an easy win after 56.Rxd1, e.g. 56...Bxd1 57.Kf2 Bc2!
(after 57...Kb3?! 58.Ke1 Bc2 59.Kd2 Bb1 60.Ba5 a3 61.c4 Kxc4 62.Bc3 White still resists) 58.Ke2 Bd3+ 59.Kd2 Bf1
60.g3 Kb3 61.Kc1 Bg2 62.Kb1 Bxf3 63.Bf8 Bxe4+ 64.Ka1 Bd5 65.Bxg7 e4 and with two passed pawns Black wins,
according to Marin.

56...Rd3 57.Bf8 Rd8! 58.Bb4 Rb8 59.Ba3?!

White waits passively. He should have tried to give up material on the queenside by 59.f4 and exchange pawns on the
kingside: 59...exf4+ 60.Kxf4 Ra8 61.Ba3 Kxc3 62.e5 fxe5+ 63.Kxe5 Bd1, although Black should win in this case as
well.

59...Rd8 60.Bb4
60...Rd7!

This waiting move is stronger than the one Karpov played on the 58th move, when they had the same position.

61.Rf1 Bd1!

Preparing the king’s invasion to b3.

62.f4 Rd3+ 0-1

GAME 95
Kasparov, Garry
Karpov, Anatoly
Seville World Championship 1987 (16)

1.c4 e5 2.Nc3 Nf6 3.Nf3 Nc6 4.g3 Bb4 5.Bg2 0-0 6.0-0 Re8 7.d3 Bxc3 8.bxc3 e4 9.Nd4 h6 10.dxe4 Nxe4 11.Qc2 d5
12.cxd5 Qxd5 13.e3 Na5 14.f3 Nd6 15.e4 Qc5 16.Be3 Ndc4 17.Bf2 Qe7 18.Rad1 Bd7 19.f4 Rad8 20.e5 Bg4 21.Nf5
Qe6 22.Rxd8 Rxd8 23.Nd4 Qc8 24.f5 c5 25.Qe4 cxd4 26.Qxg4 Nxe5 27.Qe2 Nec6 28.cxd4 Nxd4
29.Bxd4?

Kasparov commits a mistake. Of course he knows that being behind in material he is not supposed to exchange. He
probably thought his counterplay on the kingside would come in time. He must have missed something.
A) After 29.Qe4 Qc4 30.Re1 Nac6 31.f6 Black’s task to create something with the extra pawn is not easy;
B) With 29.Qg4 Nac6 30.Be3 Rd6 31.Qe4 White might have kept his position together.

29...Rxd4 30.f6

30...Qe6!
Karpov brings the queen into the centre with gain of tempo.

31.Qb2?!

After 31.Qf2 Qxf6 32.Qxf6 gxf6 33.Rxf6 Kg7 White is just a pawn down. According to Karpov, 31.Qxe6 was
relatively best. After 31...fxe6 32.Bh3 Rd6 Karpov evaluates the position as slightly better for Black. The small
advantage might have been big enough for Karpov to win.

31...Qe3+!

Centralizing his queen. Karpov plays very strongly until the end of the game.

32.Kh1 b6 33.fxg7 Nc4 34.Qc2 Kxg7 35.Bd5 Nd6

Centralizing all his pieces.

36.Qb2 Qe5! 37.Bb3 a5

Karpov wants to drive the bishop from the diagonal.

38.Qf2 f5!

The black king is not in danger at all as his pieces are quite dominant in the centre.

39.Qb2 b5 40.a3

40...Kg6!

By gradually improving his position with every move, Black renders White defenceless against his dominant pieces in
the centre.

41.Qf2
Kasparov resigned before the resumption. 41...Rd3 would follow.
Fruitful Years (1988-1990)

In 1988, for the first time since 1983 Karpov had no matches with Kasparov and he could think of other goals again. He
used this opportunity to play more tournaments than ever before, making this the year in which he played the highest
number of regular games. Objectively, he played better than ever. 1988 was an especially fruitful year with many good
endgames. Karpov performed well in the new World Cup series, which culminated in a shared win with Kasparov in
Skelleftea. In the overall standings Karpov came second, with a total score of 81 points, two points less than Garry.
In the new FIDE cycle of 1989, Karpov first crushed Johann Hjartarson 3½-1½, but Artur Yusupov made him look like
a human. This match was very close – Karpov won by 4½-3½. In the final he defeated Jan Timman 6½-2½, qualifying
for his fifth match with Kasparov. This was their last duel for the world title, won by Garry with 12½-11½ again.

GAME 96
Karpov, Anatoly
Timman, Jan
Amsterdam Euwe Memorial 1988 (4)

1.c4 g6 2.d4 Nf6 3.Nf3 Bg7 4.g3 c5 5.Bg2 Qa5+ 6.Nc3 Ne4 7.Bd2 Nxd2 8.Qxd2 cxd4 9.Nxd4 Nc6 10.e3 0-0 11.0-0
Rb8 12.Rfd1 e6 13.a3 a6 14.Rac1 Ne5 15.Ne4 Qxd2 16.Rxd2 b5 17.cxb5 axb5

18.b3!

Karpov stops ...Nc4 and keeps the option open to play either Nc5 or Nd6.

18...Rb6

After 18...d5 19.Nc5 White is better because of the weak b5-pawn. Bf1 is in the air.

19.Rc7 Bb7 20.h3!?

A typical Karpov move which now permits f2-f4.

20...Bxe4 21.Bxe4 d5

After 21...Rd8 White plays 22.Bg2!, allowing no opportunity for ...f5 and a knight transfer to d6 via f7. In case of 22...f5
23.Bf1, White is better.

22.Bg2 g5 23.Bf1!

Increasing the pressure on the queenside.

23...Rfb8

In case of 23...Ra8 24.f4 gxf4 25.gxf4 Ng6 26.Nxb5 d4 27.exd4 Nxf4 28.a4 White is a pawn up.

24.Rdc2 Bf8 25.Ra7 Bd6

26.Ra5!

Creating a passed pawn.

26...b4 27.a4 Nd7 28.Ra7 R6b7 29.Ra6

Also promising was 29.Rxb7 Rxb7 30.a5.

29...Bf8 30.Rac6 Ra7 31.Bb5

Tying up Black’s queenside.

31...h5 32.Be2 g4

Timman imaginatively sacrifices a pawn to free himself.

33.hxg4 hxg4 34.Rc7 Rxc7 35.Rxc7 Nc5 36.Bxg4 e5

37.Nc6?

Karpov was probably short of time, as he rarely makes mistakes like this. He makes gestures at the enemy king but he
will not catch it. The knight will find itself far removed from Black’s passed pawn.
A) After 37.Nf3!? Nxb3 (or 37...Bd6 38.Rc6 Rd8 39.Nd2, defending the b-pawn) 38.Bh5 Bd6 39.Rd7 Nc5 (39...Rb6
loses to 40.Bxf7+ Kf8 41.Bxd5 Nc5 42.Rf7+ Ke8 43.Nd2 b3 44.a5 Rb4 45.Nxb3 Nxb3 46.a6) 40.Rxd6 b3 41.Nd2 b2
42.Rxd5 Nxa4 43.Kg2 Nc3 44.Rd7 White wins;
B) 37.Nf5 is the move Karpov and Zaitsev suggest in their Chess Informant analysis: 37...Nxb3 38.Bh5 and now:
B1) 38...Nd2 39.Bxf7+ Kh8 40.Kg2 b3
analysis diagram

41.Rc1!! Rb6 42.Rh1+ Bh6 43.a5 wins;

B2) 38...Nc5 39.Bxf7+ Kh8 40.Bxd5 (alternatively, 40.Nh4 Bd6 (40...b3! 41.Bxd5 Rb6 transposes to 40.Bxd5) 41.Ra7
Rb7 (41...b3? loses to 42.Bxd5 with a mate threat on g6) 42.Rxb7 Nxb7 43.Bxd5 Nc5 44.a5 also wins) 40...b3 41.Nh4
Rb6 42.Rc8 Nd7 and now:

analysis diagram

B21) Karpov and Zaitsev mention only 43.Rc1, which allows Black a dangerous pawn on b2. Probably this is what
Karpov had calculated during the game. We can understand that he did not like to live with a knife on his throat if it
could be avoided: 43...b2 44.Rb1 Ba3;
B22) 43.a5!! (this intermediate move is decisive) 43...b2 (43...Rb4 loses to 44.Bxb3! Rxb3 45.a6 Ra3 46.Rd8) 44.Be4
b1=Q+ 45.Bxb1 Rxb1+ 46.Kg2 Rb8 47.Rxb8 (47.Rc7 Rd8 48.a6 should win as well) 47...Nxb8 48.Ng6+ Kg7 49.Nxf8
Kxf8 50.Kf3 Ke7 51.Ke4 Ke6 52.f4 and White wins.

37...Bd6 38.Ra7

Objectively 38.Nxb8!? was still better. After 38...Bxc7 39.Nc6 Nxb3 40.Nxb4 d4 41.Bd1 dxe3 42.Nd5 Ba5 43.Nxe3
Nc5 White is a pawn up and he would have more winning chances than Black drawing chances.

38...Rb6 39.Nd8

The knight operates originally. However, it is too far removed from the b-pawn.

39...Nxb3!

Of course this is trivial now, but Timman had foreseen his chances with the b-pawn several moves earlier when he
sacrificed the g-pawn.

40.Nxf7

Or 40.Bh5 Bb8; in case of 40.a5? Nxa5 41.Rxa5 b3 42.Ra1 Bc7 Black traps the knight.

40...Bb8 41.Rd7 Nc5 42.Rxd5 b3 43.Nxe5 b2 44.Bf5 Nxa4 45.Ng4 b1=Q+ 46.Bxb1 Rxb1+
47.Kg2

So White has three connected passed pawns for a piece. Three pawns against bishop or knight would be a draw, as long
as they do not get much closer to promotion. With the rooks unlikely to disappear soon, the story is different.

47...Kf7 48.f4 Nc3 49.Rd3 Ne4 50.Kf3 Rb4 51.Nf2 Nxf2

Also unpleasant-looking would be 51...Nc5 52.Rd5 Rc4 53.g4.

52.Kxf2 Ke6 53.Kf3 Bd6 54.g4 Ra4 55.g5

55...Ra3!

The Dutch grandmaster plays actively. He does not just wait, but pins the e-pawn. Timman’s job to hold this position is
no fun at all. He probably guessed that it was a draw, but in practice such positions seldom occur.

56.Rd4

After 56.Rxa3 Bxa3 57.Ke4 Bc1 White cannot push his pawns. He must force the pawns closer to their promotion
squares first. If all three pawns reach the fifth rank, White is likely to win.

56...Bb8

Karpov mentions 56...Ra1, attaching a plus/equal sign.

57.Rb4 Bd6

Trying to tie White to the e3-pawn with 57...Ba7 does not solve Black’s problems. After 58.Re4+ Kf5 59.Re5+ Kg6
60.Re6+ Kg7 61.f5 it is not clear whether he can hold.

58.Rb5 Rc3

Timman keeps pinning the e-pawn.

Karpov and Zaitsev suggest the active 58...Kd7. This indeed forces matters:
A) 59.f5 Kc6 60.Rb1 Bc5 61.Re1 Kd5 62.g6 and now:
A1) 62...Ke5? 63.g7 Ra8

analysis diagram

64.Ra1!! (winning the vital tempo – 64.Rg1 Rg8) 64...Rg8 (64...Rc8 65.Ra5) 65.Ra5 Kxf5 66.Rxc5+ Kf6 67.Rc7 and
White wins;
A2) 62...Bd4 63.Kf4 Be5+ (63...Ra1 also holds, but this rescue would not be available if White’s rook had gone to e2
via b2) 64.Kg5 Ra2 65.Rg1 Rf2 66.Rg4 Rf3 67.e4+ Kd4 and Black holds.
B) The reply 59.g6!? is not mentioned. This push aims to exploit the fact that the g-pawn is unguarded. It is very scary
for Black, but he should just be able to hold: 59...Ke7 60.Rb7+ Ke6 61.Rb6 (in case of 61.g7 Ra8 62.e4 Rg8 63.f5+
Kf6 64.Rb6 Ke7 65.f6+ Kxf6 66.Rxd6+ Kxg7 Black defends easily) 61...Rd3 62.Ra6 Rc3 63.f5+ Ke5 64.f6 Kxf6
65.Rxd6+ Kg7 and in spite of the two-pawn deficit Black can draw: 66.Kf4 Rc1! (if White’s pawn were on e4 he
would be winning here) 67.e4 Rg1 68.Kf5 Rf1+ 69.Ke5

analysis diagram

For the draw an effort is still required, as Black cannot allow e4-e5:
B1) 69...Re1 70.Kd5 Rd1+! 71.Ke6 Ra1! and Black barely draws. His last two moves are ‘only’ moves. Black must
occupy the a-file to be able to attack successfully from the side.
Other moves are not simple either:
B2) 69...Rf2!? (remarkably, this rook move to the second rank draws while moving to f3 loses) 70.Kd5 Rd2+! 71.Ke6
Ra2! (the rook must be this far removed to attack successfully from the side) 72.Kf5 Rf2+! and Black holds after a
series of ‘only’ moves;
B3) 69...Rf3? 70.Kd5 Rd3+ 71.Ke6 Ra3 72.Kf5 Rf3+ 73.Kg5! Rg3+ 74.Kh5 Rh3+ 75.Kg4 and White will be able to
carry out e5.

59.f5+ Ke7
60.Rb7+!

Karpov revitalizes the position.

60...Rc7?

Timman releases the pin on the e-pawn for a moment and Karpov makes him pay dearly for it.
After 60...Bc7! 61.Ke4 Rc1 62.Kd4 Rd1+ Black is likely to hold.

61.Rb6!

Attacking the bishop and preventing the rook from pinning the e-pawn again.

61...Rc5

After 61...Rc3 62.f6+ Ke6 (62...Kd7 63.g6)

analysis diagram

63.f7!! is the way to win the bishop. It is not too difficult to see. The alternative 63.g6? was worse, leading to the
difficult draw we have seen after 63...Kxf6 64.Rxd6+ Kg7. Karpov and Zaitsev do not give any other move and only
evaluate this as plus/minus. There follows 63...Kxf7 (or 63...Rc8 64.g6 Rf8 65.Kg4) 64.Rxd6 and White wins, as the g-
pawn stands much better on g5.

62.Kg4!

Karpov now invades decisively with the king.

62...Rd5
After 62...Re5 63.f6+ Ke6 64.g6 Kxf6 (or 64...Rxe3 65.Rxd6+ Kxd6 66.g7) 65.Rxd6+ Kg7 66.Kf4 White will push his
e-pawn, ensuring the win.

63.Rb7+ Kf8 64.f6! Ra5 65.Rd7 Be5 66.Kf5 Bb2+ 67.Kg6 Ke8 68.Rb7 Be5 69.Re7+ Kd8 70.Kf7 Bb2 71.g6 Ra6
72.Re6 Ra7+ 73.Kf8 1-0

GAME 97
Timman, Jan
Karpov, Anatoly
Tilburg 1988 (3)

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.a3 Bb7 5.Nc3 d5 6.cxd5 Nxd5 7.Qc2 Nxc3 8.bxc3 Be7 9.e3 Qc8 10.Bb2 c5 11.Bb5+
Bc6 12.Bd3 c4 13.Be2 Nd7 14.a4 a6 15.0-0 0-0 16.e4 b5 17.Ba3 Bxa3 18.Rxa3 Qb7 19.Nd2 Nb6 20.a5 Nd7 21.f3
e5 22.d5 Qa7+ 23.Kh1 Bb7 24.Raa1 f5 25.Rad1 f4 26.g4 Rf6 27.Qb2 Re8 28.Rg1 Bc8 29.Rg2 Nf8 30.h4 Rh6 31.Rh2
Qe7 32.h5 g6 33.Qb4 Kg7 34.Rg1 Nd7 35.Qxe7+ Rxe7 36.hxg6 Rxh2+ 37.Kxh2

37...Kxg6!

Karpov wants to keep the option of opening the kingside with ...h7-h5.

38.Kh3?!

And Timman wants to prevent this.

A) After 38.Kg2 Nc5 39.Kf2 Nb7 40.Ra1 White is only a bit worse, according to Zaitsev. He should be able to hold;
B) Zaitsev mentions 38.Rb1!?, 38...Nc5 (Black can play other moves to keep things calm, for example 38...h5)
39.Nxc4 (39.Bxc4 bxc4 40.Rb6+ Kg5 41.Nxc4 Bd7 is unclear according to Zaitsev) 39...bxc4 40.Rb6+ Kg5 41.Rc6
Nd3 42.Rxc8 Rb7 and Black obtains counterplay.
38...Nc5 39.Kh4?

Timman could still have gone back to g2. Karpov wrote in New In Chess that he had quickly calculated that Timman’s
king would get into trouble. Interestingly, he uses the word ‘calculated’, not ‘seen’.

39...Rc7 40.Rb1 Nb7 41.Ra1

According to Karpov 41.Bxc4!? had to be tried. Then, with 41...Nd6! (in case of 41...bxc4 42.Rb6+ Kg7, 43.d6! is
stronger than Karpov’s move 43.Rxa6, when Black is better. If 41...Nxa5 42.Bxb5!?) 42.Bf1 Rxc3 Black can press
without any risk.

41...Nd6

Karpov improves his knight.

42.Nb1?

White believes that the kingside is fixed, so he can manoeuvre on the queenside. This turns out to be very unfortunate.
Interestingly, Karpov doesn’t criticize the text.
White could have tried to stay in the game with 42.Rh1, but it is no longer easy: 42...Rg7 43.Kh3 h5! (in case of
43...Nb7 the white king can still escape from the danger zone: 44.Kg2! Nxa5 45.Ra1 Nb3 46.Nxb3 cxb3 47.Rb1 and
White is safe) 44.Kg2! Rh7 and White has problems because of the weakness on a5.

42...Rg7!!

Almost out of the blue Karpov catches Timman’s king with a powerful regrouping.

43.Na3 Kf6! 44.Rg1

44...Nf7! 45.Kh3

Avoiding the checkmate with ...Rg6-Rh6, but it doesn’t keep White in the game for long.

45...h5 0-1

GAME 98
Karpov, Anatoly
Andersson, Ulf
Thessaloniki Olympiad 1988 (7)

1.d4 Nf6 2.c4 e6 3.Nf3 Bb4+ 4.Bd2 Bxd2+ 5.Qxd2 0-0 6.g3 d5 7.Bg2 Nbd7 8.0-0 c6 9.Rc1 Qe7 10.Qe3 Re8
11.Nbd2 e5 12.dxe5 Nxe5 13.Qxe5 Qxe5 14.Nxe5 Rxe5 15.cxd5 Nxd5 16.e4 Nb6 17.f4 Re7 18.a4 a5
19.Nb3!

Karpov aims to place his knight on c5.

19...f6 20.Ra3!!

Seemingly incongruous, the rook is better here than on a1, although it looks weird to put the rook on an occupied rank.
20.Nc5?! would have been premature because of 20...Nd7.

20...Be6?!

Black can make some preparatory moves before developing the bishop, as Nc5 can always be answered with ...Nd7, for
instance 20...Kf7!? 21.Kf2 (after 21.Rc5 Be6 (21...Rd7!?) 22.Nxa5 Nd7 23.Nxb7 Nxc5 24.Nxc5 Rd8 Black is active)
21...Be6 (or 21...g6 22.Rc5 Be6 23.Nxa5 Rd7 and White must be careful) 22.Nc5 Rd8 23.Rc2 Rd4 with active play.

21.Nc5! Bf7

In case of 21...Rd8 22.Nxe6 Rxe6 23.Rb3 Nxa4 24.Rxb7 White is much better, as his rook is active and Black’s pawns
are vulnerable.
22.Rd3!

This is the point of the subtle 20.Ra3. White now occupies the d-file.

22...Rae8 23.Bh3 Nc8

Karpov recommends 23...g6!?, but White still has the advantage.

24.Bd7 Rd8 25.Rcd1 Nb6 26.Bg4 Rxd3

If 26...Rde8!? 27.Rd4!?.

27.Rxd3 Kf8

In principle this is correct; however, the move has a tactical drawback.

After 27...h5 28.Bh3 g5 29.Bf5 Kg7 30.Kf2 White is pressing.

28.Rd8+ Re8 29.Rd4! Rb8

On 29...Re7, 30.b4! is strong.

30.Rd7!

Exploiting the unfortunate placement of the black king. The rook is strong on the seventh rank.

30...h5 31.Bh3 Ke8 32.Rc7 Na8

A desperate attempt to obtain some counterplay.

After 32...Rd8 33.Nxb7 Rd1+ 34.Kf2 Kf8 35.Nxa5 Rd2+ 36.Kg1 Rxb2 37.Nxc6 Black is in trouble.

33.Rxb7 Rxb7 34.Nxb7 Nb6 35.Nd6+ Ke7 36.Nxf7 Nxa4

Andersson sacrifices a piece to force a passed a-pawn. Karpov refutes this idea by accurate calculation.
37.Nh8! Nxb2 38.e5!

Karpov counts on his own passed pawn.

38...a4 39.Ng6+ Ke8

On 39...Kf7, 40.e6+ Kxg6 41.f5+ wins.

40.Be6

Karpov safely stops the a-pawn and wins with his extra piece.
40...Nd3 41.Bg8 fxe5

On 41...Nb4, 42.e6 Nd5 43.Bf7+ Kd8 44.e7+ wins.

42.Nxe5 Nb4 43.Nc4 h4 44.Bh7 hxg3 45.hxg3 Ke7 46.Be4 Kf6 47.g4 Ke6 48.Kf2 Nd5 49.Kf3 Nb4 50.g5 c5 51.f5+
Ke7 52.Ke3 Kf8 53.Kd2 Ke7 54.Kc3 1-0
 GAME 99
Karpov, Anatoly
Ftacnik, Lubomir
Thessaloniki Olympiad 1988 (14)

1.c4 c5 2.Nf3 Nf6 3.Nc3 e6 4.g3 b6 5.Bg2 Bb7 6.0-0 Be7 7.d4 cxd4 8.Qxd4 d6 9.Bg5 a6 10.Rfd1 Nbd7 11.Qd2 Qc8
12.b3 0-0 13.Rac1 Rd8 14.Qb2 Qc7 15.a4 Rac8 16.Ne1 Bxg2 17.Nxg2 Ne5 18.Be3 Qb7 19.f3 Nc6 20.Nf4 h6 21.g4
Ne5 22.Bf2 Ne8 23.Bg3 g6 24.Qd2 Bf6 25.Ng2 Kh7 26.Qe3 Bg7 27.Kh1 Nc6 28.Bf2 Rb8 29.Qd2 Ne7 30.e4 Nc6
31.h3 Nc7 32.Ne2 Ne8 33.Nd4 Nxd4 34.Bxd4 Bxd4 35.Qxd4 Qc6 36.Ne3 Qc5 37.Kg2 Qxd4 38.Rxd4 Rdc8
39.Rdd1 Kg7 40.f4 Rb7 41.Rc2 Ra8 42.Ra2 Rba7 43.Rda1 Rc8 44.Kf3 Rb7 45.Rd1 b5 46.Rd3 bxa4 47.Rxa4 Rcb8
48.Ra3 Kf6 49.Nc2 Rb6 50.Ke3 g5 51.b4 gxf4+ 52.Kxf4 Rc6 53.Rdc3 Rbc8 54.Ne3 Nc7 55.Rd3 Ne8 56.Ra5 Rxc4
57.Nxc4 Rxc4 58.Rxa6 Rxb4 59.Ra8 Nc7 60.Ra7 Ne8 61.Rda3 Kg6 62.R7a4 Rb1 63.Rd3 Rb6 64.Kf3 Nf6 65.Rad4
d5 66.exd5 Nxd5 67.Kg3

67...Rb4?!

Ftacnik exchanges a pair of rooks. The position is still equal, but this allows Karpov to use his remaining rook freely.
This is a mistake which occurs in many amateur games, by the way. Unless the two rooks can set up an attack against
the king, the endgame with an exchange down tends to be more drawish with rooks or, even more so, with queens on
the board. Then the stronger side cannot use his king freely.

68.Rxb4 Nxb4 69.Rd7 Nd5 70.Kf3 Nf6 71.Ra7 Nh7 72.Kf4 Nf6 73.Ke5

Karpov, of course, improves his king.

73...Nd5 74.Ra3

Defending the h-pawn in advance and hoping to keep his pawns on the board.
74...h5

Ftacnik would have attained a draw by doing nothing, but he wants to force it by means of simplification.

75.gxh5+ Kxh5 76.Rg3!

Fixing the king at the edge of the board.

76...Kh6 77.Kd6 Nf4 78.Ke7 Kh5 79.Kf6

After 79.Kxf7 e5 80.Kf6 Kh4 Black takes White’s last pawn and his king and knight are not separated.

79...Nd5+ 80.Ke5

In case of 80.Kxf7 Kh4 81.Rf3 e5 82.Ke6 e4 (82...Nf4+ 83.Kxe5 Nxh3 84.Kf5 is a bit shaky, but still a draw) 83.Rf8
Nc3 84.Rc8 Nd1 85.Kf5 Kxh3, Black holds.

80...Kh4?

Black should keep his pieces close to each other. This losing mistake separates the king from its pawns and the knight
won’t be able to get close.
81.Rf3!

Winning the pawns and leaving the knight defenceless.

81...Nb6

The knight is driven even further away from the king. A gloomy prospect for Black.

82.Rxf7 Kxh3

82...Nd5 83.Kxe6 loses as well.

83.Kxe6 Nc4

Knights move faster than kings, so it is natural when they approach each other that the former departs. Karpov
recommended 83...Kg3, moving towards the knight. Now 84.Rc7! is the only winning move. Karpov stops after
84...Kf3 (84...Na4 85.Kd5 Nb2 86.Rc1! (White has no other win) 86...Kf3 87.Kd4 – simple, but the only winning
move), evaluating the position as equal. Computer analysis reveals more. If it were Black’s turn here he would hold.
analysis diagram

85.Rc3+!. This nice intermediate check wins. Had Ftacnik played in this fashion, Karpov would doubtlessly have found
this move. After 85...Kf2 (in case of 85...Kf4, 86.Rc6 wins the knight) 86.Rc6 Na4 87.Rc4 the knight falls.

84.Rf3+!

This is not simply a useful intermediate check, it is the only winning move.

84...Kg4

Or 84...Kg2 85.Rc3! Na5 86.Kd5 Nb7 87.Ra3 Nd8 88.Ra7.

85.Rd3!

The rook limits the movements of the knight.

85...Kg5 86.Kd5 Nb6+ 87.Ke5 Nc4+ 88.Ke4 Nb6 89.Rd8!

Preventing any approach towards the king.

89...Nc4 90.Rd4 Nb6

If Black could move one more time, ...Nc8 would draw.

91.Ke5 Nc8 92.Ke6 Na7 93.Kd7 1-0

GAME 100
Karpov, Anatoly
Portisch, Lajos
Rotterdam World Cup 1989 (6)

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.b3 Bb4+ 6.Bd2 Be7 7.Bg2 c6 8.Bc3 d5 9.Ne5 Nfd7 10.Nxd7 Nxd7 11.Nd2 0-
0 12.0-0 Nf6 13.e4 b5 14.Re1 dxe4 15.Rc1 Rc8 16.c5 Nd5 17.Rxe4 Nxc3 18.Rxc3 f5 19.Rxe6 Qxd4 20.Rc1 Bxc5
21.Qe2 Bb6 22.Nf3 Qg4 23.Rd1 f4 24.h3 b4 25.hxg4 Bxe2 26.Rxe2 fxg3

27.Kf1!

Karpov is ready to remove the pin on f2. The king unblocks the pawn and is ready to become active in the centre.

27...Bxf2

The Hungarian grandmaster has aimed for this position, but it is troublesome for him. Black has three pawns but only
one is passed, and it is well blocked. White’s rooks control the open files. However, Black’s position is not an easy nut
to crack.
After 27...gxf2 28.Ne5 Rcd8 29.Rd7 Rfe8 30.Bxc6 White would have won.

28.g5

Opening a diagonal for the bishop.

28...Rc7 29.Re4!

A very strong move, improving the rook and preparing to improve two more pieces.

29...c5

30.Ke2!?

White defends his knight and steps away from the influence of the rook on f8. In his subsequent New In Chess analysis
Karpov preferred 30.Rd3 followed by Bh3 and Kg2.

30...g6 31.Rd6 Kg7 32.Rde6 Rff7

Black stops the check on the seventh rank. He digs in and taunts White: try to win this if you can. He has three pawns
and no weaknesses and moreover, the g3-pawn ties White’s hands.
33.Kd3!!

Karpov is up to the tough task – first he improves his king.

33...Rfd7+

If 33...Rcd7+ 34.Kc4 Rf5 35.Ne5!? (with 35.Re8!? White can try to break through just as in the game) 35...Rc7 (after
35...Rd4+ 36.Rxd4 Bxd4 37.Nf3 Bf2 38.Kd3 Rf7 39.Ke4 White’s pieces are dominating)

analysis diagram

36.Kd5!! – White can improve the king still further, so much so that it wins material by trapping Black’s rook:
36...Rxg5 37.Kd6 Rb7 38.Nd7 h5 39.Re7+ Kh6 40.Kc6 Rxd7 41.Rxd7 and White wins with his extra rook.

34.Kc4

The king is nicely placed on c4.

34...Rf7 35.Re8!

White may attack along the 8th rank.

35...Rfd7?!

Sacrificing a tempo in order to improve the bishop. It is hard for Black to come up with anything constructive.

36.Bh3!

Karpov improves the bishop and also goes after the enemy king.

36...Rf7 37.Be6 Rf8 38.Bd5 Rxe8 39.Rxe8 h6 40.Re2?!

This looks like a move that serves to pass the time control. Best was 40.Ne5!! which gives White a direct assault on the
black king: 40...hxg5 41.Rg8+ Kh7 42.Nxg6 Kh6 43.Nf8 Re7 44.Bf3 and White wins.

40...hxg5?

Portisch thinks he is doing well after undertaking very little in a passive position. Now he decides to open up the game.
Karpov is the beneficiary.
Waiting with 40...a5 was slightly better, but White should win eventually: 41.Re4!? Ra7 (after 41...hxg5 42.Nxg5 Kf6
43.Nh3 g5 44.Re8 Bd4 45.Rg8 Be3 46.Rf8+ Kg6 47.Rf3 White wins the key g3-pawn) 42.gxh6+ Kxh6 43.Rh4+ Kg7
44.Rg4 Kf6 45.Be4 Rg7 46.Rg5 and Black is in big trouble.

41.Nxg5 Rd7?!
Dropping the g3-pawn, as the f-file is not accessible for the rook from the seventh rank. After 41...Rc8 42.Re6! (now
White attacks a new target: the g6-pawn) 42...Kh6 43.Nf7+ Kh7 (43...Kh5? leads to mate: 44.Bf3+ Kh4 45.Re4+ Kh3
46.Ng5+ Kh2 47.Rh4+ Kg1 48.Rh1 mate) 44.Ne5 White wins.

42.Nh3!

Suddenly Karpov shifts his attention and wins the g3-pawn.

42...Bd4 43.Rg2 Be5 44.Ng5 1-0

There is no point in fighting on without the g3-pawn.

GAME 101
Karpov, Anatoly
Seirawan, Yasser
Skelleftea World Cup 1989 (8)

1.d4 d5 2.c4 dxc4 3.e4 e5 4.Nf3 Bb4+ 5.Bd2 Bxd2+ 6.Qxd2 exd4 7.Qxd4 Qxd4 8.Nxd4 Bd7 9.Bxc4 Nc6 10.Nxc6
Bxc6 11.Nc3 Rd8
12.Ke2!

Karpov keeps his king in the centre. He has a small edge, mainly because of the passivity of Black’s bishop on c6.

12...Nf6

In all the games with this opening-endgame, the knight is placed on f6. This is natural in a way, but the knight’s
movements will be limited by White pawns.
A) After 12...Ne7?! 13.Rad1 0-0 14.Ke3 a6 15.f4 White is somewhat better;
B) Risky is 12...f5!?, but it gives Black chances to finish his development, and then he has no problems: 13.exf5 Ne7
14.Be6 (14.Rhe1 Bxg2) 14...Rf8 15.Rhe1 (15.Rad1 Nxf5) 15...Bxg2 16.Rg1 (16.f3 Bh3 17.Nb5 Nxf5) 16...Bc6
17.Rxg7 Nxf5 with active play.

13.f3

Karpov makes two enemy pieces passive with one pawn move.

13...Ke7

There were previous games with this position, so Seirawan had prepared this move. He hoped to be able to hold this
slightly worse position.
After 13...Rd4 14.b3 Ke7 15.Rhd1 Rhd8 (in case of 15...Rxd1 16.Rxd1 White can gain space on both wings) 16.Rxd4
(16.Nd5+?! Nxd5 17.Rxd4 Nf4+ 18.Ke3 Nxg2+=) 16...Rxd4 17.Ke3 Rd8 18.g4 h6 19.h4 White was better and went
on to win in Bagirov-Matulovic, Titovo Uzice 1978.

14.Rhd1 a5

Yasser prevents White from gaining space on the queenside. He could not swap both rooks because the f7-pawn is
hanging.
15.Rac1 Rxd1

After 15...Rhe8 16.Bb5 Bxb5+ 17.Nxb5 c6 18.Nd4 g6 19.Rc5 White still has an edge.

16.Rxd1 Rb8

In case of 16...Nd7 17.Bd5 Bxd5 18.Nxd5+ Kd8 19.f4 Re8 20.Kf3 Black would still have no counterplay whereas
White could build up an initiative on the kingside.

17.Ke3

17...g6

Eventually, this will turn out to be a weakening that Karpov knows how to exploit. It was very hard to choose from the
many possibilities, all of which give White an edge:
A) 17...b5 would also have created a weakness in his own camp. 18.Bf1! is Zaitsev’s move, which keeps the e2-square
free for the knight. After 18...Nd7 19.Rc1 Ne5 20.Kd4 f6 21.f4 and White is better;
B) Improving the knight with 17...Ne8 is not trouble-free either: with 18.e5 (White can also force a doubling of pawns
with 18.Bd5!?) 18...f6 19.Nd5+ Bxd5 20.Rxd5 b6 21.f4 White keeps a pleasant advantage.

18.Bb3

18.h3 was interesting as well.

18...Ra8!

Seirawan notices that the a6-square is free.

19.Rd4!

The fewer pieces on the board, the more important it is to control the only open file. Karpov uses his rook for more then
just occupying the d-file. As usual, he prevents any counterplay of his opponent.

19...Nd7 20.Bd5 Ra6! 21.f4

Gaining space, but this seems a bit hasty. 21.h3!? was another option.

21...f6?!

This looks solid, but Seirawan misses a hidden chance to obtain counterplay. With 21...Nf6!? he could aim to take on d5
and equalize fully, e.g.:
A) 22.Bxc6 Rxc6 23.e5 (after 23.Kf3 Nd7 24.g4 f6 Black has equalized) 23...Ng4+ 24.Kf3 Nxh2+ 25.Kg3 Nf1+
26.Kf2 Nh2 and Black can defend;
B) 22.Bc4!? (controlling the f1-square) 22...Rb6 23.b3 (23.Bb3 Ng4+) and now:
analysis diagram

B1) After 23...Ng4+ 24.Kf3 f5 25.Nd5+ Bxd5 26.Rxd5 Black would be in trouble;
B2) But the intermediate move 23...a4!! would provide counterplay for Black. This is a typical Kasparov-like move. In
the middle of a forced tactical line there is a small tactical finesse, which changes the whole position. After 24.bxa4
Ng4+ 25.Kf3 f5 26.Nd5+ Bxd5 27.Rxd5 Rb4 Black would be back in the game.

22.h4

Karpov gains space.

22...h6?!

This weakens the kingside. Seirawan is a very fine positional player, but even he misses the implications of this
position. Yasser should have exchanged on d5 at once with 22...Bxd5 23.Nxd5+ Kd8 24.g4 (Zaitsev gives this move.
After 24.f5 Rd6 or 24.Nc3 Ke7 Black is just a touch worse) 24...h6 (this is much better than Zaitsev’s 24...Rd6? 25.g5!
– Karpov’s second mentions this fixing move, attaching a plus/minus evaluation) and maybe Karpov would have tried
25.Rd2 here. White is just a fraction better.
23.g3!!

This is a fabulously deep concept – one of Karpov’s best. Essentially he intends to weaken the f5-square in order to
invade! Remarkably, Zaitsev doesn’t even give the text an exclamation mark. This is a special move – it is deep, very
sly and very strong. It also opens the second rank for his rook and prepares h4-h5 with Rd2 and maybe Rh2.

23...Bxd5 24.Nxd5+ Kd8 25.Nc3 Rb6 26.Rd2!

Karpov leaves Seirawan little hope that he will be allowed to exchange pawns on the queenside and he wants the rook
on the second rank to transfer it to the kingside.

26...c6
27.h5!

Five moves ago Black played ...h6, now we can see the drawback of this move. Its exploitation reveals the hand of an
endgame virtuoso.

27...g5

After 27...gxh5 28.b3! (28.Rh2 h4!) 28...h4 (28...Rb4 29.Rh2) 29.gxh4, White’s rook invades on the g-file without
wasting a tempo by going to the h-file first.

28.Kf3!

The white king threatens to invade via g4.

28...Ke7

The knight wants to block on e5 after f4-f5.

29.Ne2! c5

Black is relinquishing control of more and more squares.

30.Nc3! Rd6 31.Rd5 Rxd5?

In principle, pure knight endings provide fewer chances to survive worse positions than rook endings.
31...Rc6! would also have led to a depressing position, but at least it wouldn’t allow a quick invasion. After 32.Kg4 Ke6
33.b3 Black is very much paralysed, but his position doesn’t fall apart on its own, like after 33...c4? 34.Ne2!.

32.Nxd5+ Ke6

After 32...Kf7 33.Ne3 Kg7 34.Kg4 White invades.

33.Ne3!

Finally, as a result of White’s very subtle plan of weakening the f5-square, he is able to capitalize on his opening.

33...b5 34.Nf5 Nb6

Black gets no counterplay for the h6-pawn.

35.Nxh6 Nc4 36.Nf5 Kf7 37.b3 Nd2+ 38.Ke3 Nf1+ 1-0

GAME 102
Karpov, Anatoly
Andersson, Ulf
Skelleftea World Cup 1989 (15)

1.d4 Nf6 2.c4 e6 3.Nf3 Bb4+ 4.Nbd2 b6 5.a3 Bxd2+ 6.Bxd2 Bb7 7.Bg5 d6 8.e3 Nbd7 9.Bh4 c5 10.Bd3 0-0 11.0-0
cxd4 12.exd4 d5 13.Re1 dxc4 14.Bxc4 Qc7 15.Rc1 Rfc8 16.Rc3 Qd6 17.Bg3 Qf8 18.Qd3 a6 19.Ng5 b5 20.Ba2 Rxc3
21.Qxc3 Bd5 22.Bb1 Rc8 23.Qe3 Qd8 24.f3 Qb6 25.Qd2 a5 26.Bf2 b4 27.Rc1 Rxc1+ 28.Qxc1 h6 29.Nh3
29...Qc6?!

Andersson makes the same mistake a number of players have committed against Karpov. Exchanging queens does not
bring Black closer to a draw but simply allows Karpov to use his king. This mistake is not decisive, but Andersson
could have played more astutely. With 29...Qa6 30.axb4 axb4 31.Be1 b3 Black could have kept the b-pawn.
Safest was 29...bxa3!? when Black defends without too much trouble: 30.bxa3 Qa6 31.Nf4 Bc4.

30.Qxc6! Bxc6 31.axb4!

Karpov creates a weakness.

31...axb4 32.Nf4 Nb6 33.b3 Nfd5 34.Nd3 Bb5?!

After this Karpov starts putting pressure on the b4-pawn.

With 34...Nc3!? 35.Bc2 (35.Nxb4 Bxf3) 35...Nbd5 36.Kf1 Kf8 37.Ke1 Ke7 38.Kd2 Bb5 Black coud have kept a
better grip on the position. White can win a pawn by taking on b4, but that simplifies into a drawn opposite-coloured
bishop ending.

35.Bc2 Kf8 36.Be1 Bxd3

With 36...Nc3!? Black could have tried to employ the same defensive set-up:
A) 37.Kf2 Nbd5;
B) 37.Nxb4 Ne2+ 38.Kf2 Nxd4 39.Bd1 and White just has a small, symbolic advantage;
C) Karpov recommends taking: 37.Bxc3 bxc3 38.Nb4, and stops here, assessing this as plus/minus. Indeed, White
threatens to bring the king to e3 and play Bd3: 38...Nd5 39.Nxd5 (39.Na2 f5) 39...exd5 40.Kf2 Ke7 41.Ke3 Kd6
42.Bd3 Kc6 43.Bxb5+ Kxb5 44.Kd3 Kb4 45.Kc2 h5! (it is better to stop White gaining space first) 46.f4 h4 47.f5 g6
48.f6 (48.fxg6 fxg6 49.g4 g5 50.h3 Kb5!) 48...g5 49.h3
analysis diagram

49...Kb5! 50.Kxc3 Ka5 and with this interesting fortress idea Black can draw this pawn endgame in many lines.

37.Bxd3 Ke7 38.Kf2 Kd6 39.Bd2 Nd7

Regrouping with 39...Nc8!? looks adequate for a draw: 40.Bc4 (after 40.Be4 Nce7 41.Ke2 Nc6 42.Kd3 f5 Black is
safe) 40...Nce7 41.Ke2 (or 41.Bxd5 Nxd5 42.Ke2 Kc6 43.Kd3 Kb5 and Black can prevent the invasion via c4)
41...Nf5 42.Kd3 Nh4 43.Bxd5 Kxd5 44.Bxb4 Nxg2 45.Bf8 g6 46.Bxh6 Ne1+ 47.Ke2 Nc2 and Black cannot lose.

40.Bc4!?
Karpov does not want Andersson to put his knight on c6.
In the event of 40.Ke2?! – Karpov attaches these marks – 40...Nb8 41.Bc4 Nc6 42.Kd3 Na5 Karpov thinks Black does
all right.

40...N7b6

A) It would be strange to open the position with 40...e5!? when the opponent has the bishop pair. However, the knights
are controlling the centre very well, and the move would allow Black to get closer to the key b3-pawn:
A1) In case of 41.Be3 Nxe3 42.Kxe3 f5 43.dxe5+ Nxe5 44.Bb5 Kd5 45.Kf4 Ng6+ Black is safe;
A2) 41.dxe5+ Nxe5 42.Be2 (42.Ke2 Nc3+) 42...Nc6 43.Bd3 Nd4 44.Bc4 Nc6 45.g3 Na5 (45...h5!?) 46.Bxd5 Kxd5
47.Bxb4 Nxb3 48.Bf8 g6 49.Bxh6 Ke6 and Black holds.
B) Black can still play 40...Nb8!?, but then he must be prepared to give up a pawn for a fortress: 41.Bxd5 Kxd5
42.Bxb4 g6 43.Ke3 Nc6 44.Bc3 Na7 – Karpov gives a plus/minus to this fortress idea.

41.Ke2 h5 42.Kd3 Kc6 43.g3!?

Karpov makes all the necessary preparations first.

43...g6 44.Bxd5+ Nxd5 45.Kc4 f5 46.h3 Kb6

47.Bxb4

It has taken Karpov 17 moves to conquer the identified weakness.

47...Ne3+ 48.Kd3 Nd5 49.Bd2 Kb5

Andersson leaves the c6-square vacant for his knight. But this means that his king will be further removed from the
kingside.

50.g4 Nf6
After 50...hxg4 51.fxg4 Nf6 52.g5 Nd5 53.h4 Ne7 54.Ke3 Kc6 (54...Nd5+ loses to 55.Kf3 Kc6 56.h5 gxh5 57.g6 Ne7
58.g7) 55.b4 Kd5 (55...Kb5 56.Kf4) 56.Kd3, White wins.

51.Bg5 Nd5 52.gxh5 gxh5 53.Bd2 Nf6 54.Ke3 Nd5+ 55.Kf2 Ne7 56.Bg5 Nc6

A) In case of 56...Ng6 57.Ke3 Kb4 58.d5 e5 59.d6 Nf8 60.f4 e4 61.Kd4 Kxb3 62.Bh4 e3 63.Kxe3 Kc4 64.Be7 Nd7
65.Kf3 White has decent winning chances;
B) Alternatively, 56...Nd5 loses to 57.Kg3 Kb4 58.Kh4 Kxb3 59.Kxh5.

57.Bf6

57...f4?

Andersson blocks the passage with his pawn, but this opens another invasion route. Chess is an interesting game.
Despite a few lapses Black could still have saved himself. Surprisingly, he had time to take the b3-pawn with
57...Na5!! and return with the knight to the kingside in time:
A) 58.b4 Kxb4 (58...Nb3 59.Kg3 Nd2 60.Be7 Kc4 61.Bc5) 59.Kg3 Nb3 60.Kf4 Kc4 61.Kg5 Nd2 (Black would not
lose either after 61...Nxd4 62.f4 Ne2 63.Be5 Kd5 64.Kxh5 Ke4 65.Kg5 Ng1 or 65.Kg6 Nxf4+) 62.Kxh5 Nxf3 63.h4
Kd5 64.Kg6 Nxh4+ and all White’s pawns disappear;
B) 58.Kg3 Nxb3 and now:
B1) 59.Kh4 Nd2 60.f4 (60.Kxh5 Nxf3 61.h4 Kc4=) 60...Ne4 61.Be5 (61.Kxh5? Nxf6+ 62.Kg6 Nd5 63.Kg5 Kc4
64.h4 Ne3 would be going way too far) 61...Nf2 62.d5 exd5 63.Bd4 Nd3 64.Be3 d4 65.Bd2 Nf2 and Black is safe;
B2) 59.Kf4 Nd2 60.Ke5 (60.h4 Kc4 61.Kg5 Nxf3+ 62.Kxh5 Kd5=) 60...Nxf3+ 61.Kxe6 Ng1 62.d5 Nxh3 63.d6
(63.Be5 Nf2 64.Kxf5 Kc5 65.Ke6 Ne4=) 63...Nf4+ 64.Kxf5 Kc6 and Black is again safe.

58.Ke2! Kb4 59.Kd3 Kxb3 60.Ke4 Kc4 61.Be5 Ne7 62.Bxf4 Nc6 63.Be5 Ne7 64.h4 Nd5
65.Bh8!

Setting up a zugzwang,

65...Ne7 66.Bg7 Ng6

After 66...Nd5 67.Ke5 Ne7 68.Kxe6 Ng6 69.Bf6 White also wins.

67.Bf6 Nf8 68.Ke5 1-0

1990 saw Karpov playing the Candidates’ final against Jan Timman. This match became a tragedy for the Dutch
grandmaster. After a blunder in the first game and a missed opportunity in the fourth, he was trailing. With 2-0 the
match was maybe already over in a higher sense. Karpov showed once again that he was very good at taking his
chances when he was offered them.
The following fascinating endgame turned out to be crucial in this match.
 GAME 103
Karpov, Anatoly
Timman, Jan
Kuala Lumpur Candidates’ Final 1990 (4)

1.d4 Nf6 2.c4 g6 3.Nf3 Bg7 4.g3 c6 5.Bg2 d5 6.cxd5 cxd5 7.Ne5 0-0 8.Nc3 e6 9.0-0 Nfd7 10.f4 Nc6 11.Be3 f6
12.Nd3 Nb6 13.b3 Qe7 14.a4 Bd7 15.Bc1 Rfd8 16.e3 Be8 17.Ba3 Qf7 18.Rc1 Bf8 19.Bxf8 Qxf8 20.g4 Qe7 21.Qd2
Rac8 22.Ne2 Rc7 23.Rc5 Nc8 24.f5 g5 25.Ng3 e5 26.Qc1 b6 27.Rc2 e4 28.Nf2 Nd6 29.Qd2 Rdc8 30.Rfc1 a5 31.Bf1
Nb4 32.Rc3 Qd7 33.Nd1 Rc6 34.Rxc6 Rxc6 35.Rxc6 Qxc6 36.Nc3 Kf8 37.Kf2 Ke7 38.Ke1 Kf8 39.Kd1 Qc8
40.Ke1 Kg7 41.Na2 Nxa2 42.Qxa2 Qc7 43.Kf2 Kf8 44.Qb2 Ke7 45.Be2 Kd8 46.Ke1 Kc8 47.Kd2 Kb7 48.Qc1
Qe7 49.Ke1 Bd7 50.Kf2 Ne8 51.Qh1 Qb4 52.h4 Qxb3 53.hxg5 fxg5 54.Qxh7 Qxa4 55.Qe7 Qc6 56.Qxg5 a4

57.Qe7!

In the 1960s Karpov already liked to pin pieces – this remained a useful weapon decades later. In this case the pin
doesn’t stop White having a lost position, yet it puts up the stiffest resistance.

57...Qd6 58.Qxd6 Nxd6 59.Bd1

59...Bb5?

Timman wants to support the a-pawn with his bishop, but the latter achieves nothing on b5 and will soon return. This
mistake costs tempi. The Dutch grandmaster should have pushed the pawn with 59...b5!. Then his edge in the pawn
race would have brought him the win.
White cannot really stop the pawn with both pieces. In his remarkable analysis for New In Chess Timman mentions that
Karpov had planned 60.Ne2. Now: 60...b4 61.Nc1 (61.Nf4 b3 62.Nxd5 Nb5 63.Ke1 Be8 64.Kd2 Bf7 65.Nc3 a3 wins
for Black according to Timman) 61...Kc7 and White is in trouble according to Karpov.
White has chances to push his pawn too, with 60.f6, and now:
A) Black achieves no more than a draw if he just pushes his pawns: 60...b4 61.g5 b3 62.g6 b2 63.Bc2 a3 64.f7 Nxf7
65.gxf7 a2 66.f8=Q b1=Q and the position is even;
B) After 60...Bxg4 Karpov shows some nice lines in which White can stop Black’s pawns: 61.Bxg4! a3 62.Be6 Kc6
63.Ne2! (63.f7? Nxf7 64.Nxe4 loses to 64...Ng5!) 63...a2
analysis diagram

64.Bxd5+! (this nice shot saves White) 64...Kxd5 65.Nc3+ Kc4 66.Nxa2 and now:
B1) 66...Kb3 67.Nc1+ Kb2 68.Ne2 b4 (68...Kc2 69.d5) 69.d5 b3 70.Kg3 Kc2 71.Nd4+ Kc3 72.Nxb3 Kxb3 73.Kf4
and there is no way White can lose;
B2) 66...b4 67.Nxb4 Kxb4 68.Kg3 Kc4 69.Kf4 – as Karpov pointed out White is not worse here: 69...Kd3 (69...Kd5
70.Kg5 Ke6 71.Kg6 Kd7!) 70.Ke5 Nf7+ 71.Ke6 with a draw.
We can conclude that Black has to do something against the white pawns, so 60...Nf7! is the main line. Timman
analyses this very deeply, whereas Karpov doesn’t even mention the move in his Chess Informant analysis.
61.Nh5 (there is no point in bringing the king to the queenside: after 61.Ke1? Kc7 62.Kd2 b4 63.Nh5 Kd6 64.Nf4 Bb5
Black will soon push ...a3 and win) 61...b4 62.Nf4 b3 63.Kg3. The king has to defend the g4-pawn.
analysis diagram

A) 63...Ng5?! (Timman believes this wins; however, White might get away) 64.Kh4 (64.Nxd5 Be6 65.Nc3 a3
66.Bxb3 Bxb3 67.Kf4 Nh7 68.g5 a2 69.Nxa2 Bxa2 70.Kf5 Nf8 71.g6 Nd7 and Black wins) 64...Nf3+ and now:

analysis diagram

A1) 65.Kg3 (Timman doesn’t analyse this retreat) 65...Be8 (after 65...b2 66.f7 b1=Q 67.f8=Q Qxd1 White has a
perpetual with 68.Qb4+) 66.Nxd5 Nd2 67.Nc3 a3 68.Bxb3 Nxb3 69.Kf4 Nd2 70.Ke5.
analysis diagram

A11) Black can go wrong with 70...Nb1? 71.Na2 Bf7 72.d5 Kc7 73.g5 Kd7 74.Kxe4 Kd6 75.Kd3 Bxd5 and now:
A111) Otto Magyar came up with a remarkable idea: 76.g6!! Bxa2

analysis diagram

77.Kc2 – despite Black’s huge material advantage he must be content with a draw because of the wrong-coloured
bishop: 77...Ke6 78.g7 Kf7 79.e4 and White draws;
A112) 76.Nc3 Bb3 77.g6 a2 (or 77...Ke6 78.Ne4! Bd5 79.Nc3 Bf3 80.Kc2=) 78.Nxa2! (I have never seen this motif
before; it saves the game) 78...Bxa2 79.Kc2=.
A12) 70...Bf7?! 71.d5 Kc8 72.Kd4 Kd7 73.Kc5 Nb1 74.Na2 Kc7 75.g5 Kd7 76.Kd4 Kd6? (76...Nd2) 77.Kxe4!
Bxd5+ 78.Kd3 Bxa2 79.Kc2 and Magyar’s idea saves White again;
A13) 70...Kc6!. Black’s best answer. He wants to block the passage of White’s king to the queenside: 71.Ke6 Nc4
72.f7 Bxf7+ 73.Kxf7 Nxe3 74.g5 Nf5 and Black wins.
Back to the diagrammed position on the previous page after 64...Nf3+.
A2) 65.Bxf3 exf3 66.f7 f2 67.f8=Q f1=Q 68.Qd6! (this is stronger than 68.Qe7, after which Timman points out that
Black wins by 68...Qb5 69.Nd3 Kc7) 68...Qb5 69.Nxd5.

analysis diagram

The position looks scary, but Black seems to have a narrow path to the win: 69...b2 70.Qc7+ Ka6 71.Qd6+ Ka7
72.Qc7+ Qb7 73.Qc5+ Ka8 74.Nb6+ (74.Qa5+ Kb8 75.Nc3 Kc8 also gives Black good winning chances) 74...Kb8
75.Nxd7+ Qxd7 76.Qb6+ Qb7 77.Qd8+ Ka7 78.Qa5+ Qa6 79.Qc7+ Ka8 80.Qd8+ Kb7 81.Qe7+ Kc6 82.Qc5+ Kd7
83.Qf5+ Ke7 and it looks like Black escapes – but there is no certainty...;
A3) White turns out to have enough ammunition to draw if he keeps pushing his pawns: 65.Kh5!! Be8+ 66.Kh6 b2
67.Bc2 a3 68.Bb1 Nd2 69.Ba2 (after 69.g5 Nxb1 70.g6 Nc3 71.f7 Bxf7 72.gxf7 b1=Q 73.f8=Q Qb6+ Black wins)
69...b1=Q 70.Bxb1 Nxb1 71.g5 a2 72.g6 a1=Q 73.f7 Bxf7 74.gxf7 Qa3 75.Kg7 Qxe3 76.f8=Q Qxd4+. Timman stops
here, claiming that Black must win, but this is not convincing. White can probably draw:
analysis diagram

77.Qf6! and now:

A31) 77...Qc5 78.Qf7+ Kc6 79.Qe8+ Kb6 80.Qe6+ Kb5 81.Qxd5 Qxd5 82.Nxd5 is a draw;
A32) After 77...Qxf6+ 78.Kxf6 d4 79.Ke5 d3 80.Nd5 White draws;
A33) 77...Qd2 78.Qe7+ Kc6 79.Ne6!! – White’s queen is not closed off and the knight on b1 is unprotected, which is
why White can defend: 79...Nc3 80.Qc5+ Kb7 81.Qb4+ with a perpetual.
Now back to the position after 63.Kg3.
B) 63...Kc6!? (this gives very good winning chances) 64.Kh4 Nd6 65.g5 Be8 66.g6 and now:
analysis diagram

B1) 66...Bxg6 (maybe objectively Timman’s move leads only to a draw but the lines are so beautiful that they are
worthy of examination) 67.Nxg6 Kb5 68.Ne7 Kc4 69.Nc8!. Gschnitzer discovered this lovely move, but Timman
came up with a remarkable retort:

analysis diagram

69...Kd3!! 70.Nxd6 b2 71.f7 b1=Q 72.Bxa4 (72.f8=Q Qxd1=) 72...Qb8 (on 72...Qe1+ White replies 73.Kg4!, not
73.Kh5? Qf2) and now Timman analyses two knight moves:
B11) 73.Nf5 Qd8+ 74.Kg4 (this move by Timman leads to a more instructive draw; 74.Kh5 is simpler: 74...Qf6!
75.Nh6 Kc4 76.Bd7 and White holds) 74...Qf6 75.Be8 Ke2 76.Kf4 Kf2 77.Kg4 Kg2 78.Nh4+ Kf2 79.Nf5 and Black
cannot crack this fortress according to Timman;
B12) 73.Ne8 Qb4 and now:
B121) 74.Kg5? Kxe3 75.Kg6 (75.Ng7? Kd2 76.Ne6 Qe7+!) 75...Kd2! (75...Kxd4 76.Kg7) 76.Kg7 e3 77.f8=Q
Qxf8+ 78.Kxf8 e2 79.Bc6 e1=Q 80.Bxd5 Qf2+ 81.Ke7 Qxd4 and White cannot reach the drawish set-up with the
bishop on b7 and the knight on d5 with his king in the corner;
B122) 74.Kh5! (this is the better way to approach, as it avoids the check on e7. Timman did not examine this)
74...Kxe3 75.Ng7 Kd2 76.Ne6 e3
analysis diagram

77.Bb5!! (quite a nice diversion) 77...Qxb5 78.f8=Q Qe2+ (the underpromotion after 78...e2 79.Ng5 e1=N doesn’t
cause White any problems) 79.Kg6 Qd3+ 80.Kg5 e2 81.Qf2 Qe3+ 82.Qxe3+ Kxe3 83.Nf4 Kd2 84.Ng2! and White
holds.
Back to the position after 69...Kd3!!.

analysis diagram

The Dutch grandmaster did not consider the sacrifice 70.Bxb3!? axb3 (after 70...Nf5+ White shouldn’t lose: 71.Kh3
axb3 72.f7 b2 73.f8=Q b1=Q 74.Qxf5 or 71.Kg5 axb3 72.f7 Nxd4! 73.exd4 b2) 71.Nxd6 b2 72.f7 b1=Q 73.f8=Q
Qe1+ 74.Kh5 and White is not worse.
And then a final look at the position after 66.g6.
B2) Timman did not look at 66...Kb5!!. The final position of my line is tough for Black: 67.Kg5 Kc4 (in case of
67...b2 68.Bc2 Kc4 69.Bb1 Kb3 70.Ne6 a3 71.Nc5+ Kc3 White holds with 72.Na6) 68.Be2+ (68.Ne6 loses to 68...a3
69.Nc5 b2 70.Bc2 Kc3 71.Bb1 Kd2) 68...Kc3 69.Nxd5+ Kd2 70.Bc4 Bxg6 71.Kxg6 Nxc4 72.f7 b2 73.f8=Q b1=Q
74.Qf2+ Kd3 75.Nf4+ Kc3 76.Nd5+ Kb3 and White is struggling. It looks like this wins for Black and if so, that
means that 63...Kc6 is the winning move.
Now we reach the main line of the analysis.
C) Black can force a win by continuing the race aggressively with 63...b2!, another move Timman doesn’t mention. It
wins thanks to a study-like finesse.

analysis diagram

64.Bc2 – now Black has two moves, both leading to the same unusual position:
C1) 64...a3 65.Bb1 Ba4 66.Kh4 Bc2!! (a lovely shot which would also have followed after 66.Ne6. In the event of
66...Bb3? 67.g5 a2 68.Bxa2 Bxa2 69.g6 White would hold – see line C21 below) 67.g5 Nh8! 68.g6 Nxg6+ 69.Nxg6
Bxb1 70.f7 a2 71.f8=Q
analysis diagram

71...a1=Q and Black probably wins as his king can find shelter behind White’s pawns. Experience shows that junior
players often have problems noticing such motifs. For instance: 72.Qb4+ Ka6 73.Qd6+ (or 73.Ne7 Bd3 74.Qd6+ Kb5
75.Qc6+ Ka5 76.Qc5+ Ka4 and Black wins) 73...Kb5 74.Qxd5+ Kb4 75.Qc5+ Kb3 and White will run out of checks;
C2) 64...Bb5 65.Kh4 a3 66.Bb1 and now:

analysis diagram

C21) 66...Bc4 does not win: 67.g5 a2 68.Bxa2 Bxa2 69.g6 Ne5!? (a fancy try. If 69...b1=Q? 70.gxf7 Qb4 71.Ne6
Qe1+ 72.Kh5 Qe2+ 73.Kg6 Qg4+ 74.Ng5 White is even winning. After 74...Qc8 75.Kg7! Qg4 76.f8=Q Qxg5+
77.Kh8 the f-pawn is unstoppable) 70.f7! b1=Q 71.f8=Q Nxg6+ 72.Nxg6 Qe1+ 73.Kg4 Qxe3 74.Qb4+ with
perpetual check;
C22) 66...Bd3!! (clearing the way for the pawn) 67.g5 Nh8! 68.g6 Nxg6+ 69.Nxg6 Bxb1 70.f7 a2 71.f8=Q a1=Q with
the same position as in line C1 above.
Back to the game.

60.Ne2 a3 61.Nc1

After 61.Bb3 Bc4 62.Bxc4 (in case of 62.Nc1 Bxb3 63.Nxb3 a2 64.f6 Kc6 65.g5 Kd7 Black wins the white pawns)
62...dxc4 63.Nc3 b5 64.Na2 Kb6 65.f6 Nf7 Black wins.

61...Kc7

Karpov evaluates 61...Be8 62.Na2 Kc6 63.Bb3 as slightly better for White.
In case of 61...Bd7 62.Bb3! Kc6 63.Kg3 Nc4 64.Kf4 Nb2 65.Kg5! White is better according to Karpov.

62.Kg3 Nc4

Not 62...Kd7? 63.Kf4 Bc4 64.g5 a2 (64...Ke8 loses to 65.f6 Kf8 66.g6 b5 67.g7+) 65.Nxa2 Bxa2 66.g6 Ke7 67.g7
Kf7 68.f6 Bc4 69.Bg4 (69.Ke5 wins as well) 69...Kg8 70.Be6+ Nf7 71.Kf5 Bb5 72.Kg6 and Black ends up in a lethal
zugzwang.

63.Be2 Be8

The alternative was 63...Kd6!? (Karpov gives the ‘!?’ mark) 64.Kf4 (White can still activate his king) and now:
A) 64...Nb2?? (an embarrassing mistake has crept into Karpov’s analysis here) 65.Bxb5 Nd3+. Karpov stops here,
claiming that Black wins, but after 66.Bxd3! exd3 67.Nb3 a2 68.Kf3 Black is hopelessly lost;
B) 64...Be8 65.Bxc4 dxc4 66.g5 b5 67.Na2 c3 68.Kxe4 Bf7 69.Nxc3 Bc4 (after 69...b4 70.Kd3 White holds because
of the wrong colour of the bishop) 70.d5! b4 71.Kd4 b3 72.Ne4+ Kd7 73.Kc3 and White can stop the black pawns.
64.Kf4

Timman mentions that Karpov told him Podgaets had found an amazing line with 64.g5; however, Karpov’s sense of
danger hadn’t let him down: 64...Nxe3 (64...Kd6 is simplest) 65.Kf4 and now:
A) 65...Nc2 66.Ke5 Nb4 67.Kf6 Kd6 68.g6 Nc6 69.Bb5 e3 70.g7 Ne7 71.Bxe8 e2 72.Nxe2 a2 73.Nc3 Ng8+ 74.Kf7
Nh6+ 75.Kf8 a1=Q 76.Nb5 mate!;
B) 65...Ng2+! (this is the hole in Podgaets’s fantastic line) 66.Ke5 Nh4 67.g6 Bxg6 and Black is safe.

64...Nb2

Karpov evaluates 64...b5!? 65.Na2 Nb2! 66.f6 Nd3+ 67.Kf5 b4 68.Bd1 as equal.

65.Kg5

After 65.Na2 Nd3+ (65...Kd6 66.Kg5 Ke7 67.Kh6 Kf6 68.Bf1! b5 69.g5+ Kxf5 70.Bh3 mate) 66.Kg5 b5 67.Bd1 b4
68.Bb3 Kd6 69.Kf6 Nf2 70.Kg5 Nd3 the position is equal according to Timman.

65...Nd3 66.Nb3

Karpov played this quickly and caught Timman unawares.

66...a2?

This natural move is almost losing. Black can only stay in the game with moves that usually feature in studies.
Much better was 66...b5!. This forces White to come up with some moves that are hard to find.
A) Only Timman mentions 67.Na1!? b4 68.f6 a2 69.Bd1 Kd6 and now:
analysis diagram

A1) 70.Kf5 Ne1 71.Bb3 Ng2 72.Bxa2 Nxe3+ 73.Kf4 Ng2+ 74.Kg3 Ne1 75.Kf2 Nd3+ 76.Ke3 Ke6 77.g5 Bf7
78.Bb1 Nc1 79.Kd2 Nd3 80.Bxd3 exd3 81.Kxd3=;
A2) 70.Kh6 Nb2 71.Bb3 Nc4 72.g5 Ke6 73.Bxa2 Nxe3 74.Nb3 Nf5+ 75.Kh7 Nd6 76.Nc5+ Kf5:
A21) 77.Bxd5? e3 78.Kh6 Nf7+! (Timman’s 78...Bf7 is not convincing:

analysis diagram

79.Nb7!! e2 80.Nxd6+ Kf4 81.Nxf7 e1=Q 82.g6 and Black cannot win with his queen) 79.Kg7 e2 80.Nd3 Nxg5
81.Bc4 Ke4 82.Ne1 Ke3 83.Nc2+ Kd2 84.Bxe2 Kxc2 85.Bc4 Kc3 (after 85...b3 86.Bxb3+ Kxb3 87.f7 Nxf7 88.Kf8
with a study-like draw) 86.Ba2 Kxd4–+;
A22) 77.Kh6! Nf7+ 78.Kg7 Nxg5 79.Bxd5 e3 80.Bc4 Bf7 81.Bxf7 Nxf7 82.Kxf7 e2 83.Nd3 Ke4 84.d5! b3
(84...Kxd3 85.d6 e1=Q

analysis diagram

86.d7 – quite remarkable; Black can do nothing with his extra queen) 85.d6 b2 86.d7 b1=Q 87.d8=Q Qxd3 88.Qxd3+
Kxd3 89.Kg8 and Black cannot win.
B) After 67.Kf6 Black can continue the race with:
B1) 67...b4, and now:
B11) 68.Bd1? Ba4 69.g5 a2 70.g6 Bxb3 71.Bxb3 a1=Q 72.g7 Qg1 73.Bxd5 (73.Kf7 Qxe3) 73...Qxe3 74.g8=Q
Qxd4+ and Black wins;
B12) White must reply in spectacular fashion: 68.g5! Ba4 69.g6! Bxb3 70.g7 a2 71.g8=Q a1=Q 72.Ke7!! (a lovely
move – it draws where checks lose) 72...Qa7 73.Qd8+ Kc6+ 74.Ke6 and White is not worse because of his strong f-
pawn and Black’s vulnerable king.
B2) Gschnitzer wrote to New In Chess, asking important questions about 67...a2 68.Na1 b4 69.Bd1 (Timman points out
what is wrong with attacking the bishop by 69.Ke7: 69...Ba4! 70.f6 b3 71.f7 b2 72.f8=Q bxa1=Q and Black wins)
69...Nf2 70.Bb3 Nxg4+ 71.Ke7 Nxe3.
analysis diagram

Here, White can achieve the draw in two ways:

B21) 72.Kxe8?! Nxf5 and now:
B211) 73.Bxa2? (here Gschnitzer’s question came up) 73...e3 74.Bxd5 (74.Nc2 loses to 74...e2 75.Ne1 Kd6) 74...e2
(74...Nxd4 75.Bc4 Kd6 76.Bd3!=) 75.Nc2.

analysis diagram

Here the simple move 75...Nxd4 wins already: 76.Ne1 Kd6 77.Be4 (on 77.Bc4, 77...b3 78.Bxb3 Nxb3 wins) 77...b3
78.Bg6 (78.Kf7 Ke5 79.Bb1 Kf4 80.Kf6 Nf3 also loses) 78...Ke5 79.Bd3 b2 80.Ke7 Kf4.
75...Ne3!! – this lovely winning shot was missed by the Dutch grandmaster: 76.Bf3 b3!! (76...Nxc2 77.Bxe2 b3 78.Bc4
b2 79.Ba2 Nxd4 also wins) 77.Ne1 Nc2 78.Bxe2 and now 78...b2 as well as 78...Nxe1 79.Bc4 b2 80.Ba2 Kd6 81.Kf7
Nc2 82.Kf6 Nxd4 wins for Black;
B212) 73.Bxd5! (the bishop, not the knight must stop the e-pawn) 73...e3 74.Bxa2 Nxd4 75.Bc4 Kd6

analysis diagram

76.Bd3! (White cannot afford the bishop to be attacked with tempo: 76.Kf7 Kc5! (Black must win this tempo, rather
than shoulder the white king. After 76...Ke5 77.Ke7!! (77.Kg6 loses to 77...e2 78.Bxe2 Nxe2) 77...e2 78.Bxe2 Nxe2
79.Nb3! or 77...b3 78.Nxb3 Nxb3 79.Bf1 White holds) 77.Bd3 e2 78.Bxe2 Nxe2 and Black wins) 76...Kc5 (76...Kd5
77.Kd7 e2 78.Bxe2 Nxe2 79.Nb3=) 77.Ke7 e2 (77...b3 78.Nxb3+=) 78.Nb3+! Nxb3 (78...Kd5 79.Bxe2=) 79.Bxe2=.
B22) 72.f6 – this method is relatively easy. Gschnitzer just mentions this move without any further analysis. Let’s take
a look: 72...Bg6 (72...Nf5+ 73.Kxe8 e3 74.Bxa2 Nxd4 75.f7 e2 76.f8=Q is a draw) 73.f7 Bxf7 74.Kxf7 Nf5 and now:
analysis diagram

B221) Timman considers only 75.Bxa2? e3 76.Bxd5 e2 77.Nc2 Ne3!! and Black wins;
B222) Better is 75.Bxd5! e3 76.Bxa2 and White holds in all lines: 76...Nxd4 77.Bc4 Kd6 (or 77...b3 78.Nxb3 Nxb3
79.Ke6) 78.Kf6 e2 (78...Kc5 79.Bd3 e2 (79...b3 80.Nxb3+=) 80.Bxe2 Nxe2 81.Nb3+=) 79.Bxe2 Nxe2 80.Nb3; or
76...e2 77.Nc2 Nxd4 78.Ne1 Kd6 79.Kf6 b3 80.Bxb3 Nxb3 81.Kf5.
Back to the game.

67.Na1 b5 68.Bd1 b4

The World Champion gives a question mark to 68...Kd6? and shows a lovely checkmate. I think Karpov spotted it in
the game and just put it into his analysis. If this is the case, it shows his amazing vision.

analysis diagram

A) 69.Bb3 (Karpov’s move, but it doesn’t win for White) 69...Nc1

A1) 70.Kh6 and now:

analysis diagram

A11) 70...b4 71.f6 (71.Bc2! is possible here – another hole in Karpov’s long checkmating line):
A111) Now Karpov shows a lovely mate against 71...Ke6?? 72.Kg7 Nxb3 73.Nxb3 Ba4 74.Nc5+! Kd6 75.f7 a1=Q
76.f8=Q+ Kc6 77.Qc8+ Kd6 78.Qd8+ Kc6 79.Qd7+ Kb6 80.Qb7+ Ka5 81.Qa6 mate;
A112) However, with 71...Nxb3 Black can take the bishop one move earlier and win, e.g. 72.Nxb3 Kc6 73.g5 Kb5
74.g6 Kc4 75.Na1 b3 76.Nxb3 Kxb3 77.f7 Bxf7 78.gxf7 a1=Q 79.f8=Q Qh1+ 80.Kg6 Qg1+.
A12) Black can even take the bishop still one move earlier: 70...Nxb3!!, suddenly changing direction and winning the
race: 71.Nxb3 b4 72.Kg7 Kc6!! 73.Kf8 Kb5 74.Kxe8 Kc4 75.f6 Kxb3 76.f7 a1=Q 77.f8=Q Qa8+ 78.Ke7 Qxf8+
79.Kxf8 Kc2.
A2) Instead of 70.Kh6, 70.Bxa2!! achieves the draw: 70...Nxa2 71.Kf6 Nb4 (or 71...Nc3 72.g5 Nd1 73.Nc2 b4
74.Nxb4 Nxe3 75.g6 Ng4+ 76.Kg7 e3 77.f6 e2 78.Nd3 Nxf6 79.Kxf6 Bb5 80.Ne1 Bd7=) 72.Kg7 Nc6 73.Nb3! (in
the event of 73.f6 Ke6 74.g5 Bh5 75.g6 Nd8 76.Nc2 Kf5 Black wins) 73...b4 74.f6 Nd8 75.g5 Ne6+ 76.Kh6=.
Instead of Karpov’s 69.Bb3, White does have a win after 68...Kd6?:

analysis diagram

B) 69.Kf6!! (White must push his pawn without wasting any time) 69...Nf2 (69...Nb2 70.Be2 Nc4 71.g5 Nxe3 72.g6
Nxf5 73.Kxf5 Ke7 74.g7 also wins for White) 70.Be2 b4 71.g5 Ba4 72.g6 b3 73.Nxb3 Bxb3 74.g7 a1=Q 75.g8=Q
Bc4 76.Qg3+ Kd7 77.Qxf2! Qa2
analysis diagram

78.Kg7! Qxe2 79.Qxe2 Bxe2 80.f6 Bh5 81.f7, winning the pawn ending.
Back to the game.

69.Bb3 Nc1 70.Bxd5 Kd6 71.Bc4 Bb5?

This is a losing mistake.

A) After 71...Ba4 72.f6 b3 73.Kh6! b2 74.f7 b1=Q (74...bxa1=Q loses to 75.f8=Q+ Kd7 76.Qf5+ Kc6 77.Qd5+ Kb6
78.Qc5+ Kb7 79.Qb4+ Kc7 80.Qa5+) 75.f8=Q+ Kd7 76.Qf5+ Kc7 77.Qa5+ wins – Timman.
B) Best was 71...b3!! 72.Bxb3 Nxb3 73.Nxb3 Kd5 74.Na1 Kc4 and now:
analysis diagram

B1) After 75.Kf6 Kc3 76.Ke7 Bb5 (a nice line by Timman is 76...Kb2 77.Kxe8 Kxa1 78.f6 Kb2 79.f7 a1=Q 80.f8=Q
Qa4+ 81.Kf7 Qb3+ 82.d5! with excellent winning chances) 77.f6 Bc4 78.f7 Bxf7 79.Kxf7 Kb2 80.g5 Kxa1 81.g6
Kb1 82.g7 a1=Q 83.g8=Q Qa2+ 84.Kg7 Qg2+ and Black holds;
B2) Bruno Carlier found 75.d5!!. Timman assesses 75...Kc3 76.d6 Kb2 77.f6 Kxa1 78.f7 Bxf7 79.d7 Kb1 80.d8=Q
a1=Q 81.Qd1+ Ka2 82.Qa4+ Kb2 83.Qd4+ Kb1 84.Qxe4+ Kb2 85.Qd4+ Kb1 86.Qxa1+ as a win, and rightly so.
After 86...Kxa1, 87.Kf6! is the only winning move (after 87.e4 Be6! 88.Kf4 Kb2! Black neatly draws: 87.Kf5 Kb2=).
Black misses the one tempo he needs to draw: 87...Bd5 (87...Bb3 88.e4+-) 88.g5 Kb2

analysis diagram

89.Kf5! (White’s 87th move was simple. This retreat is harder to find, especially when calculating towards this
position) 89...Bb3 90.e4 Kc3 91.e5 Kd4 92.e6 Kc5 93.e7 Ba4 94.g6 and White wins.

72.Bg8 Ke7
73.Kh6?

A mistake which could have been very expensive. It indirectly shows Karpov’s respect for Timman. He was tense, as he
was not certain he would win the match.
Timman points out that shouldering with 73.Kg6! wouldn’t allow the saving finesse found by Sax (see below). Please
play endgames carefully, even if they look boring! Even an expert like Karpov can go wrong (admittedly, very rarely).
You must give it all you’ve got.

73...Kf8?

One feels for the Dutchman. He finally collapses, as his king does not really help to stop White’s pawns. After 73...b3!
74.Bxb3 Nxb3 75.Nxb3 Bc4 76.Na1
analysis diagram

... my countryman, grandmaster Gyula Sax, found 76...Kf6!, which draws.

74.Be6 Bd7

75.g5!

Karpov wins with clinical precision. Again he is about checkmate his opponent in an endgame. The mate did not appear
on the board as Timman resigned.
I feel that this game justifies the effort of going through great older games with the help of the computer. There will turn
out to be mistakes in the existing analysis. This does not in any way mean that the protagonists were not excellent
players – just that they played in a different age.
One possible conclusion is 75...b3 76.g6 Bxe6 77.fxe6 b2 78.g7+ and now:
A) 78...Kg8 79.e7 Kf7 80.g8=Q+ Kxe7 81.Qg5+ Kf7 82.Qd5+ Kf6 83.Qd6+! (not the only move to win, but this
checkmates Black nicely) 83...Kf7 84.Qd7+ Kf6 85.d5 bxa1=Q 86.Qe6 mate;
B) 78...Ke7 79.g8=Q bxa1=Q 80.Qf7+ Kd6 81.Qd7 mate.
 GAME 104
Karpov, Anatoly
Miles, Anthony
Biel 1990 (1)

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.0-0 Nxe4 5.d4 Nd6 6.Bxc6 dxc6 7.dxe5 Nf5 8.Qxd8+ Kxd8 9.Nc3 Ke8 10.b3 h5
11.Rd1 Be7 12.Bg5 Nh6 13.h3 Bf5 14.Bxe7 Kxe7 15.Nd4 Rad8 16.Rd2 Bg6 17.Rad1 h4 18.b4 Nf5 19.Nce2 Nxd4
20.Nxd4 f6 21.exf6+ Kxf6 22.Nb3 Rxd2 23.Rxd2 b6 24.Rd7 Rc8 25.Rd4 Bxc2 26.Rxh4 Re8

27.Rf4+

White has a superior rook and Black has doubled pawns. On the other hand, Black’s king is more active and he has a
bishop that can deal with distant pawns.

27...Ke5

Miles, who was nevertheless a very fine endgame player, allows Karpov to invade with the rook to the seventh rank.
There was no need to let this happen. 27...Ke6 was better. For instance, 28.Rg4 (in case of 28.Nd2 Rd8 29.Nf3 Bg6
30.Rc4 Rd6 31.Rg4 Kf6 Black should also be OK) 28...Bxb3 (after 28...Kf6 29.Nd4 Be4 Black is not worse either)
29.axb3 and now:
A) 29...Re7 30.Re4+ Kd6 31.Rxe7 Kxe7 32.h4 Kd6. This position resembles the famous 24th game of the 1935 Euwe-
Alekhine match, in which Alekhine allowed a similar king move to exchange the doubled pawns. In this case White
cannot prevent it: 33.g4 c5 34.bxc5+ (or 34.b5 c6 35.bxc6 Kxc6 36.h5 Kd5 and again Black is doing all right)
34...Kxc5 35.h5 Kd6 and Black is not worse;
B) 29...Kf6 30.Rc4 Re6 31.h4 Ke5 (31...a6 32.Kh2) 32.Rg4 Re7 33.f3 Kd5 34.Kf2 c5!? 35.Rg5+ Kd4 36.bxc5 bxc5
37.Kg3 c4 38.Rg4+ Kc3 39.bxc4 a5 and Black is in no way worse. This line is not at all forced, but it illustrates that
Black has chances.

28.Rf7 Bxb3?!
When pawns are far away, bishops are usually better than knights. The bishop could have been useful later on, to control
White’s h-pawn.
After 28...Kd6 29.Rxg7 (after 29.Nd4 Bb1 30.a3 Re1+ 31.Kh2 Rd1 32.Nf3 g6 Black is very much in the game)
29...Re1+ (29...Re4 30.Rg4) 30.Kh2 Re2 31.f3 Bb1 32.a3 (in the event of 32.Rg4 Rxa2 33.h4 Rb2 Black will cause
problems) 32...Rb2 33.Nd4 Rd2 34.Rg4 Kd5 35.Nb3 Rd3, the position is unclear.

29.axb3 Kd4

After 29...Kd6 30.Rxg7 Re1+ 31.Kh2 Re4 32.g4 Rxb4 33.h4, White is faster in the pawn race.

30.Rxg7 Kc3 31.Rxc7 Kxb4 32.Rxc6 Kxb3

33.f4!

Karpov pushes the f-pawn; if it arrives at f7 it will win a big tempo.

33...Rf8?

Miles attacks White’s pawn instead of pushing his own. He may have missed wicked idea Karpov had in mind.
This kind of endgame is very complicated and very hard to calculate for a human – indeed, even for the computer.
If Black defends his pawn with 33...Rb8 before he pushes the a-pawn, this will cost a tempo. If the analysis depicted
below is correct, it loses in an unbelievable manner: 34.f5 a5 and now:
A) 35.f6 a4 (35...b5? loses to 36.g4 a4 37.g5 a3 38.f7) 36.f7 – this is pushing things a bit too far: 36...Rf8 37.Rxb6+
Kc3 38.Ra6 Kb3 39.g4 a3 40.g5 Rxf7 41.Kg2 a2 42.Kg3 (42.h4 Rf4) 42...Rf1 43.Kg4 a1=Q 44.Rxa1 Rxa1=;
B) 35.g4 (White pushes his pawns side-by-side) 35...a4 36.f6 a3 37.g5 a2 38.Rc1 and now:
B1) If 38...Rf8:
B11) 39.Kh2 Kb2 40.Rg1 a1=Q 41.Rxa1 Kxa1 42.h4 b5 43.h5 b4 44.g6
analysis diagram

44...Rxf6!! (please bear in mind when you play such endgames: rook and pawn sometimes draw against a queen! 44...b3
would lose to 45.g7 Rb8 46.f7 b2 47.f8=Q b1=Q 48.Qa3++-) 45.g7 Rh6 46.g8=Q Rxh5+ 47.Kg2 b3 and Black can
draw because of 48.Qxb3 Rh2+! with stalemate or perpetual check;
B12) 39.Rf1 b5

analysis diagram

40.Kh2!!. This is a motif that seems unique. The king approaches the pawn, but the key function of the move is
different: 40...b4 41.Rg1! (the king has vacated this square for the rook) 41...Ka3 (41...Kb2 42.g6) 42.g6 b3 43.g7+-.
B2) 38...Kb2 39.Rf1 Rf8 and now:

analysis diagram

B21) 40.h4? b5 41.h5 b4 42.Kh2 Kc3–+;

B22) 40.Kg2 b5 (40...a1=Q 41.Rxa1 Kxa1 42.h4 b5 43.h5 b4 (43...Rh8 44.Kg3!) 44.g6 b3 45.f7 b2 46.g7 b1=Q
47.gxf8=Q and Black has no perpetual checks) 41.h4 b4 42.h5 b3 43.g6 Rxf6–+;
B23) 40.Kh2!! (chess is magic! This time the point of this king move is not at all to vacate the g1-square) 40...b5
(40...a1=Q 41.Rxa1 Kxa1 42.h4 b5 43.h5 b4 44.g6 b3 45.g7 Rb8 46.f7+-) 41.g6 Rxf6 42.Rxf6 a1=Q

analysis diagram
43.g7!!+-. In itself this move is not so great, but the concept is superb. Here we see the point of 40.Kh2 – Black has no
check. However, Black can still resist: 43...Qe1 (after 43...Qa8 44.Rf8 White promotes on the next move as Black has
no checks) 44.Rg6!! (White even has time for this. It’s just like a study) 44...Qe5 45.Rg3 and White wins.
Instead of 33...Rb8? or the game move 33...Rf8?, Black had to push his pawn at once with 33...b5!.

analysis diagram

Karpov gives the solution in study-like fashion:

A) 34.f5 a5 and now:
A1) 35.Rc5? (White wants to win a tempo) 35...Kb4? (35...Ka4! 36.g4 b4 37.g5 b3 38.Rc7 Rb8 and Black wins)
36.Rc1 a4 37.g4 a3 38.f6 a2 39.g5
analysis diagram

A11) 39...Ka3 40.f7 Rc8 41.Rf1 b4 42.g6 b3 43.g7+-;

A12) 39...Rc8 40.Rf1 Rf8 41.Kh2!! (White again clears the g1-square) 41...Kc3 42.Rg1 b4 43.g6 b3 44.g7 Rb8 45.f7+-;
A13) 39...Re5!! (the black rook fights the pawns in an unusual way, keeping them under control from behind) 40.h4
Rf5 41.Kh2! Ka3 42.Rg1 b4 (42...Kb2? 43.g6 Rf2+ 44.Kh1! – a nice shelter, isn’t it? – 44...Rf4 45.Rg2+ Kb3 46.Rxa2
Kxa2 47.f7+-) 43.g6 Rf2+ 44.Kh3 Rf3+ 45.Kg4? (45.Kh2!=) 45...Rxf6 46.h5 Rf8 47.g7 Rb8 48.h6 b3 49.h7 b2
50.h8=Q b1=Q and if 51.Rg3+ Ka4 or 51.Qh3+ Rb3. White has no useful checks.
A2) 35.f6 a4 36.g4 a3 37.g5 a2 38.Ra6 Re5 (38...Rf8 39.Kh2+-) 39.h4 Rf5 40.g6 Rxf6 41.g7=.
B) 34.Ra6

analysis diagram

Black must be careful. These kind of endgames are especially hard to play.
B1) 34...b4 35.Rxa7
B11) 35...Rb8! (the black rook backs the b-pawn, which should be enough for a draw) 36.Kf2 Kc2 37.f5 b3 38.Ke3 (or
38.g4 b2 39.Ra2 Kd3) 38...b2 39.Ra2 Kc3 40.Rxb2 Rxb2 41.g4 with a draw;
B12) 35...Kc3?!. This has been the starting point of some original analysis. It is not the best move, but it leads to some
very instructive lines:
B121) Not 36.Kf2? b3? (36...Rb8! is still fine for Black, e.g. 37.Rc7+ Kd2 38.f5 b3), which loses in an instructive way:
37.Rc7+ Kd3 38.Rb7 Kc2 39.g4 b2 40.f5 b1=Q 41.Rxb1 Kxb1
analysis diagram

42.Kf3! (the king must go forward first to win. Not 42.f6? Kc2 (42...Rg8! draws) 43.g5? Re5!! 44.h4 Kd3 45.Kf3 Rf5+
and Black draws) 42...Kc2 43.Kf4 Kd3 44.f6 and White wins;
B122) 36.Rc7+! Kd3 37.Rb7 Kc3 38.Kf2 b3 39.g4 (39.Kf3? b2 40.f5 Re1 41.g4 b1=Q=) 39...b2 (in case of 39...Re4
40.f5 Rb4 41.Rxb4 Kxb4 42.f6 White promotes with check) 40.f5 Kc2 (40...Re4 loses to 41.Rxb2 Kxb2 42.Kf3)
41.Kf3 b1=Q 42.Rxb1 and White wins.
B2) 34...Ra8!? 35.f5 Kc4 36.f6 b4 37.f7 b3 38.Rxa7 and now:

analysis diagram
B21) 38...Rf8 39.Rb7 Kc3 40.g4 b2 41.Rc7+=;
B22) Karpov shows a remarkable draw: 38...Rxa7!? 39.f8=Q Ra1+ 40.Kh2 b2.
Back to the game.

34.g4!

In the race, the speed of the pawns matters, not the quantity. This witty idea underlines that Karpov has no weaknesses
in the endgame.

34...Rxf4?!

Taking the pawn costs two tempi. Black simply cannot afford that.
Karpov’s path to victory would have been narrower after 34...b5. Then White must push his pawns quickly with 35.f5!
a5 36.f6! a4 37.g5 a3 38.Ra6 a2 39.h4 b4 (39...Kb2 loses to 40.h5 a1=Q+ 41.Rxa1 Kxa1 42.g6 Rxf6 43.g7) 40.h5 Kb2
41.g6 Rxf6 42.Rxa2+ Kxa2 43.g7, winning.

35.g5

The rest was easy to calculate.

35...b5 36.g6 Rf8 37.g7 Rg8 38.Rc7 a5 39.h4 a4 40.h5 a3 41.h6 a2 42.Ra7 1-0

So Karpov faced Kasparov for the fifth time in a World Championship final. After the first half in New York the score
was 6-6, but in the 18th and 20th game Kasparov took a two-point lead. Karpov won the 23rd game, but could not level
the score and lost 12½-11½. This was to be their last duel for the world title. The following game was Karpov’s second
win in the match.

GAME 105
Karpov, Anatoly
Kasparov, Garry
 Lyon/New York Wch-m 1990 (17)

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Be3 c5 8.Qd2 0-0 9.Nf3 Bg4 10.Ng5 cxd4 11.cxd4
Nc6 12.h3 Bd7 13.Rb1 Rc8 14.Nf3 Na5 15.Bd3 Be6 16.0-0 Bc4 17.Rfd1 b5 18.Bg5 a6 19.Rbc1 Bxd3 20.Rxc8 Qxc8
21.Qxd3 Re8 22.Rc1 Qb7 23.d5 Nc4

24.Nd2!

Karpov wants to exchange Black’s best piece.

24...Nxd2

Karpov shows that inserting 24...h6!? would not solve Black’s problems: 25.Bf4 e5 (taking on d2 as in the game would
now mean a loss of a tempo, and 25...g5 26.Bg3 Nxd2 27.Rc7 Qb6 28.Qxd2 would give White an advantage thanks to
the rook on c7) 26.Be3 (26.dxe6 Rxe6 27.Nxc4 Qxe4!) 26...Nxe3 27.Qxe3 Rc8 28.Nb3 Bf8 29.Na5 and White is
better.
After 24...Nd6 25.Rc6 Rc8 26.Qc2 White can keep up the pressure.

25.Bxd2 Rc8
26.Rc6!

Kasparov had apparently missed this tactical finesse.

26...Be5

In case of 26...Rxc6 27.dxc6 Qc7 (27...Qxc6 28.Qd8+) 28.Qd7 Be5 29.Bh6 Qxd7 30.cxd7 Bc7 31.e5! White cages in
Black’s king and wins by bringing his king over to the queenside.

27.Bc3!

Karpov increases his domination of the c-file.

27...Bb8

A) In the event of 27...Rxc6 28.dxc6 Qc7 29.Bxe5 Qxe5 30.Qd8+ Kg7 31.c7 Qa1+ 32.Kh2 Qe5+ 33.g3 Qb2 34.Kg2
Black has no perpetual;
B) Anatoly knew that Garry was unlikely to go for a passive position with 27...Bxc3 28.Qxc3 Rxc6 29.dxc6!? (after
29.Qxc6 Qa7 30.e5 Qd4 Black becomes active) 29...Qc7 30.e5 and, although rather passive, Black has chances to
survive.

28.Qd4! f6
29.Ba5!

This is a subtle positional move. It increases White’s possibilities on the c-file.

29...Bd6 30.Qc3

Karpov could have won a pawn with 30.Qb6, but after 30...Qa8 (30...Rb8 31.Qxb7 Rxb7 32.Rxa6) 31.Rxc8+ Qxc8
32.Qc6 Qf8 33.Qxa6 Qh6 Black would get some play.

30...Re8 31.a3!

Despite the presence of the bishop on d6, Karpov still plays this move. It fixes the queenside. Here Karpov already
knew what he was going to do to the bishop.

31...Kg7 32.g3 Be5 33.Qc5 h5?

Repelling the queen with 33...Bd6!? was better. But White still has every chance to win after 34.Qc1.
34.Bc7!

Karpov takes control of the c7-square and prepares the decisive invasion.

34...Ba1 35.Bf4 Qd7 36.Rc7

Here Garry could have resigned.

36...Qd8 37.d6 g5 38.d7 Rf8 39.Bd2

This is not the only solution. Karpov prevents ...Qa5 to make sure Garry cannot do anything active.

39...Be5
40.Rb7 1-0

In case of 40...h4 (or 40...Kg6 41.Qa7) 41.Ba5! Karpov would demonstrate his nicest threats: 41...Qxa5 42.Qxe7+ Kg6
(42...Rf7 43.Qxf7) 43.Qh7+!. Karpov is enjoying himself in his analysis. He has played so well in the earlier part, he
deserves it. 43...Kxh7 44.d8=Q+ is the end.
Epilogue

Before I started working on Karpov’s endgames, I expected that he had played some beautiful endgames even in his
childhood. Just like many other great players, he can produce, in one particular game, the highest level of creativity.
This happens in music, poetry and chess, and Karpov is no exception. Indeed, in the 1960s he already created some
marvellous endgames which contained some very deep concepts. What was a bit of a surprise – and I had not realized
this before – was that sometimes he produced better games from an artistic point of view against lower-ranked
opponents.
His style went through some slight changes, but nothing substantial. He is an all-round player in endgames and in his
games we can find examples of all elements of endgame skill.
Maybe there has been a slight shift in one respect. In his early days, Karpov had a predilection for pinning his
opponent’s pieces. From the early 1970s onwards, he started to win games by building up an attack on the opponent’s
king and very often he checkmated them.

I believe that any reader who devotes time to play through this selection will experience considerable joy and certainly
learn a lot. I think that many of these games leave a highly artistic impression.
I am grateful to many people for helping me during the process of writing this book. Of course, first of all I am grateful
to Karpov for his unbelievably high quality of play in a great number of endgames. Several excellent commentators
have facilitated my job by analysing some of Karpov’s endings (most notably, in alphabetical order, Dvoretsky,
Kasparov, Marin, Karsten Müller and Timman).
I also owe thanks to several other people. Antonio Palma provided me with a significant amount of material. Attila
Groszpeter and Harold van der Heijden added some valuable analysis and remarks. The book itself was written in
Singapore while I was working for Intchess Asia. Many of my trainer colleagues suggested ideas when I showed them
my work during the process. Woman grandmaster Natalia Edzgveradze and IM Tigran Gyozalyan inspired and
supported me with their encouragement, insisting that it was worth putting so much effort into the book. In a good part
of the book they followed the process almost game by game; I am really grateful to them. Even some of my pupils have
added to this book, for example the Singapore woman’s national team and Ravindran Shanmugan.

January 2007
Tibor Karolyi
Index of Opponents

The numbers refer to pages

The [Gnn] numbers refer to Game Numbers
If the Game Number is bold Karpov’s opponent played with the white pieces.

A
Adorjan 66 [G20]
Andersson 261 [G80], 319 [G98], 328 [G102]

B
Bouaziz 240 [G73]
Browne 93 [G32]

D
Debarnot 156 [G50]
Drizgalovich 33 [G7]

E
Evrosimovski 54 [G17]

F
Ftacnik 320 [G99]

G
Gheorghiu 154 [G49], 248 [G76]
Grigorian 130 [G42]

H
Hort 97 [G34], 186 [G60], 195 [G61]
Hübner 162 [G52], 221 [G67]

I
Ivkov 201 [G63]

K
Kalashnikov 11 [G1]
Karasev 55 [G18]
Kasparov 263 [G81], 268 [G82], 279 [G83], 285 [G85], 291 [G88], 291 [G89], 311 [G95], 347 [G105]
Kindermann 301 [G91]
Kirillov 25 [G3]
Klovans 79 [G27]
Kolishkin 28 [G4]
Kortchnoi 168 [G55], 171 [G56], 181 [G57], 236 [G71]
Krogius 68 [G21]

L
Larsen 165 [G53], 252 [G77]
Ljubojevic 134 [G44], 229 [G70]
Lukin 43 [G11]

M
Markland 89 [G30]
Meulders 310 [G94]
Mikenas 78 [G26]
Miles 166 [G54], 238 [G72], 283 [G84], 343 [G104]
Mukhudulin 22 [G2]

O
O’Kelly de Galway 77 [G25]

P
Parma 74 [G24]
Petrosian 144 [G47]
Pfleger 141 [G46]
Polugaevsky 109 [G39], 259 [G79]
Pomar Salamanca 126 [G41]
Popov 41 [G10]
Portisch 296 [G90], 322 [G100]

Q
Quinteros 222 [G68], 226 [G69], 244 [G75]

R
Rashkovsky 71 [G23]
Ravinsky 37 [G8]
Ribli 105 [G37], 215 [G66]
Romanishin 48 [G14], 256 [G78]

S
Sazontiev 32 [G6]
Seirawan 324 [G101]
Sergievsky 69 [G22]
Smejkal 101 [G36]
Smith 95 [G33]
Sokolov 303 [G92], 306 [G93]
Sosonko 184 [G59]
Spassky 112 [G40], 133 [G43], 183 [G58]
Steinberg 52 [G16]
Sukhanov 45 [G13]

T
Taimanov 84 [G28]
Tal 208 [G64]
Timman 287 [G86], 289 [G87], 313 [G96], 317 [G97], 331 [G103]
Torre 57 [G19], 242 [G74]
Tsamriuk 39 [G9]
Tukmakov 87 [G29]

U
Uhlmann 107 [G38], 139 [G45]
Unzicker 197 [G62]

V
Vaganian 51 [G15], 98 [G35]
Van der Wiel 212 [G65]
Visier Segovia 159 [G51]
Vujakovic 44 [G12]

W
Wirthensohn 92 [G31]

Z
Zakharov 149 [G48]
Ziuliarkin 29 [G5]
Endgame Classification

This classification of Karpov’s endgames may help those players who have problems in certain phases of the endgame.
Is there any chess player who has never spoiled a winning endgame with a material advantage? Karpov’s masterpieces
are helpful in this department.
A word of caution – any one game in the classification scheme we present below may contain multiple elements. For
example, it is possible that there was a zugzwang and also a mating attack, but even control of an open file could have
been an important factor. In these cases the number of the game will be indicated under each of the different headings.
Often the endgames never crystallised into clear rook endings as both sides had retained a bishop or a knight. These
games are not placed under a specific heading. Sometimes there is reference to opposite-coloured bishops when there
were still rooks on the board. Then the reference indicates that the bishops may have had a strong effect on the
outcome.
This classification may give interesting information, even in cases where some motifs only occurred in the analysis.
Karpov applied many different skills in his endgames. However, the fact that there is a balance between his
implementation of these tools shows how versatile he has been.
The numbers below refer to Game Numbers.

Endgame Types

Pawns 48, 63
Knights 23, 29, 32, 46, 48, 76, 101
Bishops 90, 91
Opposite coloured bishops 14, 15, 19, 44, 63, 64, 94
Bishop versus knight 7, 12, 16, 40, 50, 51, 65, 69, 70, 82, 93, 102
Rooks 1, 2, 33, 40, 47, 54, 60, 61, 62, 66, 81, 83, 84, 104
Four rooks 3, 42, 77
Exchange 11, 17, 56, 72, 99
Queens 20, 35, 78
Queens and rooks 24, 60, 71, 75, 80
Queen against pieces 43, 86
Two pieces versus rook 52, 57, 74, 92

Endgame Themes

Open file 1, 4, 5, 6, 7, 9, 10, 14, 18, 21, 22, 23, 24, 27, 28, 38, 40, 41, 45, 48, 52, 58, 64, 66, 67, 68, 74, 75, 77, 80, 84,
92, 98, 105
Attacking the king 3, 7, 14, 21, 22, 28, 36, 39, 40, 41, 42, 45, 57, 60, 64, 66, 67, 71, 74, 75, 77, 78, 83, 86, 87, 89, 91,
92, 95, 96, 97, 100, 103, 105
Passed pawns 3, 5, 9, 10, 15, 20, 23, 24, 27, 30, 32, 33, 34, 35, 36, 40, 44, 50, 52, 56, 60, 62, 62, 63, 65, 66, 72, 80, 81,
82, 83, 84, 89, 92, 93, 94, 96, 98, 100
Converting a material advantage 2, 3, 5, 10, 11, 13, 15, 17, 19, 20, 21, 23, 25, 26, 28, 35, 47, 48, 52, 53, 60, 62, 65,
66, 70, 76, 78, 83, 86, 87, 92, 98, 100, 102
King centralization 3, 11, 13, 16, 18, 27, 28, 31, 40, 46, 49, 50, 51, 58, 59, 63, 65, 76, 77, 82, 91, 93, 101
King manoeuvre 21, 36, 44, 50, 57, 60, 62, 66, 67, 81, 84, 89, 90, 92, 93
Fixing pawns 1, 4, 12, 15, 16, 23, 29, 34, 40, 44, 62, 68, 70, 72, 92, 101, 105
Rooks on the seventh rank 10, 13, 19, 23, 24, 28, 38, 45, 48, 62, 66, 71, 73, 83, 85, 86, 89, 91
Space advantage 4, 7, 29, 31, 33, 37, 39, 46, 54, 60, 70, 72, 85, 91, 93, 101
Zugzwang 1, 16, 34, 40, 43, 47, 55, 56, 58, 78, 81, 82
Exchanging 6, 18, 23, 30, 32, 42, 47, 58, 66, 70, 72, 78, 80
Pawns versus minor pieces 44, 55, 70, 93, 96, 98, 100, 103
Pawns on one side 2, 13, 60, 62, 64, 70, 82
Pin 3, 4, 8, 9, 11, 19, 48, 63, 66, 69
Pawn race 1, 32, 40, 47, 81, 90, 103, 104
Pawn promotion 40, 44, 70, 72, 86, 103, 105
Two bishops 59, 63
Hunting down a piece 12, 21, 27, 99, 100
Building a fortress 16, 17, 43, 70, 82, 102, 103
Battery 19, 65, 69
Cutting off the king 2, 62, 83
Working out a square 7, 16, 18, 47, 48
Pawn as outpost 4, 6, 24, 27, 30, 54, 68
Isolated pawn 18, 50, 71, 102
Stalemate 47, 55
Doubled pawns 27, 28, 58, 90
Shouldering 1, 40, 44, 70