Karolyi T., Aplin N. - Kasparov. How His Predecessors Misled Him About Chess (2009)

First published in the United Kingdom in 2009 by
Batsford
The Old Magistrates Court
10 South combe Street
London
W14 ORA
An imprint of Anova Books Company Ltd
Copyright © B T Batsford 2009
Text copyright © Tibor Karolyi, Nick Aplin
The moral right of the authors have been asserted.
All rights reserved. No part of this publication may be reproduced, stored

in a retrieval system, or transmitted in any fonn or by any means,
electronic, mechanical, photocopying, recording or otherwise, without the
prior written pennission of the copyright owner.
ISBN: 9781906388263
A CIP catalogue record for this book is available from the British Library.
17 16 15 14 13 12 11 10 09
10987654321
Reproduction by Spectrum Colour Ltd, Ipswich
Printed and bound by Athenaeum Press Ltd., Gateshead, Tyne & Wear
This book can be ordered direct from the publisher at the website
www.anovabooks.com
Or try your local bookshop
2
Contents
Page
Authors' Preface 5
Anatoly Karpov the 12th 9
Robert James Fischer the 11 th 54
Boris Spassky the 10th 81
Tigran Petrosian the 9th 94
Mikhail Tal the 8th 123
Vassily Smyslov the 7th 143
Mikhail Botvinnik the 6th 169
Max Euwe the 5th 192
Alexander Alekhine the 4th 202
Jose Raul Capablanca the 3rd 218
Emanuel Lasker the 2nd 235
Wilhelm Steinitz the 1st 254
3
Authors' Preface
This book is a unique reaction to a that we briefly considered the title The
unique collection of creative work. Great Successor would be appropriate.
When Garry Kasparov, the most This present book now provides us
successful world champion, retired, he with an obvious opportunity to
published a series of books under the introduce some humour, particularly as
title My Great Predecessors and it was Kasparov - we think - subconsciously
the stimulation from reading his favoured some teasing of the great
excellent series that prompted the players and former champions more so
present work. than others. By doing so he invited
others to have a joke at his expense too.
After writing two books on Humour in chess - sometimes a rare
Kasparov's astonishing career, cover- commodity - needs to take its rightful
ing his [mal period of active play from place.
1993 to 2005, we realised that there The Hungarian half of our
were similarities between Garry's co-authorship played in tournaments
games and some of his predecessors - with Garry and even faced him across
and this has opened the door for a little the board in 1980 and 1981. The 1980
bit of friendly leg-pulling! World Junior Championship was
Our original idea to write an article particularly memorable for Tibor,
for the satirical chess magazine because of the leisure hours he spent
Kingpin took on greater proportions with Garry himself - during which time
as we found more and more games the future world champion revealed his
resembling those of past champions. keen sense of humour.
This characteristic is something that
The increasing number of examples has not been reflected in his interviews
changed the single article into a series in recent times, although there were
of articles. We had originally intended glimpses of it when he came to write
to look only at the post-World War II his My Great Predecessors books.
champions but then discovered so
It goes without saying that games
many comparable games from earlier
played by world champions can be
times that we were encouraged to write
especially interesting, entertaining and
a whole book!
instructive. But it is also well worth
There was a stage during the writing looking at them from a new angle - and
of Kasparov s Fighting Chess 1 & 2 with a lighter touch.
5
Authors' Preface
The temptation is also there to look at contact him at all and the words are
some of Kasparov's losses - which are ours!! We just put our ideas into his
in fact well worth analysing. Anyone mouth in the following way.
who manages to force resignation from
the most successful chess player ever, * * * *
clearly deserves due recognition for
their triumph. My series on the world champions is
entering its final phase. In these books,
In no way does the present book try
I have covered the development of
to erode the tremendous respect
chess culture. Thank God they sold like
Kasparov has rightfully earned
hot cakes. I wrote nice things about all
with his stunning and breathtaking
the 12 champions, which is what they
performances. It just reminds everyone
justly deserved, but I only showed the
emphatically what a great game chess
rosier side of their chess.
is and that even the greatest players
make mistakes and do lose By now most of the books have been
sometimes! The royal game is just so sold, so it is time to tell the rest of the
complicated ... story. My career has been the best a
Also we consider that the My Great chessplayer has ever had and, all things
Predecessors books represent a superb considered, I am satisfied with how
contribution to chess culture and things went. On the other hand, I am
warmly recommend that both non- convinced I did not achieve everything
professionals and serious players read that I could have done: for example, I
the whole series, as Garry's chess lost more games than was necessary.
genius shines brightly through his deep And in the present work I reveal for the
analysis. first time how I came to lose quite a
few important games simply because I
One of the intentions of the present copied the world champions. It's a pity
book is to take a look at some lesser- that I didn't gain a fuller appreciation
known masterpieces of the champions, of their methods.
as well as presenting the better-known
examples, with short explanations. We Almost all chessplayers read books
hope you enjoy and learn from these on the world champions. I did so as
games. well and in my childhood I even went
through their games in great detail. In
It is great that Garry wrote his series,
fact I frequently tried to memorise
but if I were him I would have
their games, but it is more likely that
produced another version for reading
they planted themselves in the
on New Year's Eve!
subconscious part of my brain. Their
Our book is designed to be light- games were praised so many times and
hearted. So before we allow Garry to in so many places that I came to trust
speak, let us emphasise that we did not them implicitly.
6
Authors' Preface
Of course, I must also take some soften, but not erase, the negative effect
responsibility for my losses, but you they had on me.
will see that for the particular defeats
shown here the world champions are So as to underline the fact that this is
not a totally serious book, I do not lay
mostly to blame because they misled
out the material in the conventional
me - sometimes seriously. After all, it way. Instead of starting from the distant
was they who demonstrated the ideas in past and working my way towards the
the first place. present day, I adopt a different plan
based on the fact that the closer a
Can you imagine how hard it has champion was to me in time, the more
been for me to hold back my true energy I spent on examining his play.
opinions for so long? But now I cannot
remain silent any longer and must show So I look at the champions in reverse
how the champions really played. order, starting with Anatoly Karpov,
Though I have to admit that their games who was crowned before me as the 12th
are very entertaining, that can only world champion.
7
Anatoly Karpov the 12th
Anatoly Karpov was my immediate effect on my style. Of course I learned
predecessor. He held the title from to play simple positions - there were
1975 until 1985 and certainly had an many of them - and I improved my
immense effect on my chess. I played technique in this area.
him 23 times in regular tournaments.
There is nothing special about that In this book I would like to
but the 144 games in the five world concentrate on the negative effects
championship matches we contested is that I experienced from the world
unique in the history of chess. champions - effects which prevented
Despite this large number of games, me from becoming even more
you might think they had little negative devastating in my play.
One idea I picked up from Karpov was to push the a- or h-pawns all the way
- and win. Below are positions from Karpov's games illustrating this theme and
then positions from my own games where I followed his plan.
S.Sazontiev - A.Karpov A.Karpov - G.Kasparov
A.Karpov - P.Markland A.Karpov - G.Kasparov
Readers note: throughout the text you will read the words see diagram with
a reference to a certain page. It's the diagrams in the frames to which we refer.
9
First let me show you some games 26 "xg7+ ~xg7 27 hxg6 hxg6
where Karpov employed one of his 28 :bl ltab8!
favourite concepts. Now Black turns his attention to the
side where he is stronger.
S.Sazontiev - A.Karpov 29 i.a6 ':c7 30 liJfdl It)fd7
Vladimir 1964 31 llh3?!
1 d4 liJf6 2 liJo e6 3 i.g5 d5 4 c4 White wants to transfer the rook to
i.e7 5 liJc3 0-0 6 e3 liJbd7 7 .i.d3 lte8 the queenside.
8 0-0 liJf8 9 tDe5 c6 10 f4 tD6d7 On the other hand 31 .i.e2! would
11 i.xe7 'ii'xe7 12 :0 f6 13 liJg4 tDb6 have kept Black rather busy on the
14 c5 liJbd7 15 l:i.g3 ~h8 16 liJn e5 kings ide and he would not then have
17 'it'h5 e4 18 .i.e2 g6 19 -.h6 b6 had such a free hand for his queenside
operations.
20 b4
20 ... a5!
Karpov starts pushing his a-pawn. It 31. .. i.a4!
looks like it merely undennines Karpov starts exchanging on the
White's pawn chain but in fact this queenside so as to prepare an invasion.
move represents its debut perfonnance He follows up this plan with his
in a very important role. customary and distinctive purpose-
21 b5 i.b7 22 cxb6 liJxb6 23 bxc6 fulness.
i.xc6 32 l:lh2 .i.xdl 33 ~xdl ~a4!
See diagram on page 9. 34 :al liJc3! 35 g3 liJxdl 36 lbdl
:'c3 37 lIel
24 h4 :ec8 25 h5
25 ...'ii'g7! 37 ... a4!

To exchange the most dangerous The pawn is becoming increasingly
white piece in the attack. powerful.
10
38 l::td2 l:tb6 39 iLf1 ~n 40 l:tee2 A.Karpov - P.Markland

Hastings 197111972
1 e4 e6 2 d4 d5 3 ltJe3 iLb4 4 e5 e5
5 a3 iLxe3+ 6 bxe3 'iVe7 7 ttJf3 ttJe7
40 ... a3!
The pawn makes its final stride of a
glorious march. It gets closer to
promotion and takes control of the 8 a4
b2-square. I wouldn't mind betting that Karpov
had not yet seen the final role this
41 ~f2 l:tb4 42 g4ltJb6 43 l:te2 l:txe2
a-pawn had to play. Of course it is a
44 }:txe2 l:tb2!
well-known variation. Since that time
This is a poignant demonstration of
7 'ii'g4 has taken over as the main line.
the strength of the a3-pawn.
8... b6 9 i.b5+ il.d7 10 i.d3 ttJbe6
45 l:te2 ttJa4 46 ~e 1
11 0-0 h6 12 l:tel llJa5 13 'iVd2 l:te8
14 h4
Karpov uses his flank pawns well.
Here he gains space and makes sure
that ... g7-g5 is prevented.
14 ... 0-0 15 'iVf4 f5 16 exf6 lhf6
17 'ii'xe7 lhe7 18 dxe5 bxe5 19 llJe5
iLe8
Here 19 ... c4 looks better. It restricts
the light-squared bishop even though
that in turn grants more freedom to his
dark-squared brother.
46 ... ttJc3 47 %:td2 ttJxa2 0-1
Finally the fixed a-pawn falls, and
Black wins easily. Karpov engineered
this game beautifully, yet strangely he
did not include it in any books of his
selected games. Maybe he did not want
to alert his rivals to such an effective
pawn-pushing device. Naturally, the
game did not escape my attention.
11
20 c4!
Karpov gets rid of the doubled pawns
and opens the position for his bishops.
20 ... ttJac6 21 iib2 ttJb4
See diagram on page 9.
22 as!?
This is a hard move to come up with.
Perhaps it had been planned earlier.
Had he already anticipated the role of
this pawn or did he just want to prevent 31 a6! l:tfi 32 ttJe4
Black from playing as - a move which The a7-pawn is fixed. Karpov now
fixes White's a-pawn on the colour of brings up his bishop to place it under
the c8-bishop? closer surveillance.
22 ... l:tfB 23 iiaJ 32 ... ttJrS 33 .i.eS! %le8 34 iifl l:tre7
23 hS, playing extravagantly with the 35 l:be7
other edge pawn, was also possible. Now Karpov starts to exchange
pieces around the weak a7-pawn. All
23 ... dxe4 24 ttJxe4 1:r4
part of the plan.
3S ... l:txe7 36 l:lb1 ttJe7 37 %lb8+
~h7 38 <it>h2!
This is a typical Karpovian king
move. It prevents Black from
delivering a check on c 1, which
would be followed by an attack on the
a6-pawn with llal.
38 ... ttJg6?
This only helps White. He moves
away a valuable piece from the area
25 ttJd6 where the battle will take place.
39 ttJeS l:le6?
Karpov sacrifices a pawn to keep his
Returning the knight was better.
opponent's rook out of the game. Here
40 l:td8 l:le7
2S l:te4 holds on to the pawn by
stopping .. .lixc4.
2S ... ttJxd3 26 exd3 l:txh4 27 ttJe4
l:thS 28 l:tee1 iib7?!
After 28 ... l:td5 29 l:tc3 a6 Black can
live with his position.
29 ttJxcS iidS 30 f3 l:trs
Black could improve his knight with
30 ... ttJc6!? Then 31 a6 ttJd4.
Not to be sidetracked, Karpov now
plays according to a well-formulated
plan. Firstly he fixes Black's a7-pawn. 41.ttd7!
12
Anatoly Karpov the J 2'h
Karpov continues to play with great our matches forced him to increase his
purpose. He will exchange the standard of play in the openings.
defending rook as well. 12 ..:ihc5 13 dxc5 lDc8
41. ..lbd7 42 lDxd7 ~c6 43 lDb8 The variation has continued to
~b5 develop ever since our game. The
knight can also be retreated to d7.
14 h3 ~xO 15 ~xf3 ~xe5 16 i..xc6
bxc6 17 ~d4
White achieves domination along the
d-file - which provides compensation
for the pawn deficit.
17... ~f4 180-0
18 ... a5?
At this moment I adopted Karpov's
plan of pushing the a-pawn as far down
44 ~xa7 the file as possible. And I really paid
Finally the ripened fruit drops quietly the price for this misguided decision.
from the tree. White wins the pawn and A few months later Timman
so the rest is simple. improved on this game with 18 ... e5!.
44 ... lDe7 45 ~b6 lDc8 46 ~c5 ~g6 Maybe he never bothered to investigate
47 a7 lDxa7 48 ~xa7 e5 49 d4 exd4 Karpov's earlier games. After 19 ~e3
50 ~xd4 1;n 51 f4 g5 52 fxg5 hxg5· ~xe3 20 fxe3 lDe7 21 ltd7 lDf5
53 <itg3 ~g6 54 ~f3 <ifi>f5 55 g3 1-0 Timman achieved a draw against
Karpov in Tilburg 1986. Black has
Karpov won this game in impressive done well in this position ever since.
style. This plan was implanted in my 19 %:tfel a4?!
brain and I was just waiting for an I stuck to the plan that I had learned
opportune moment to carry it out in from Anatoly Evgenievich.
one of my own games. Quite incredibly 20 ':e4 ~h6 21 ~e5
I had my chance against Karpov
himself.
A.Karpov - G.Kasparov
Game 17, World Championship
LondonlLeningrad 1986
1 d4 lDf6 2 c4 g6 3 lDc3 d5 4 lDo
~g7 5 'ifb3 dxc4 6 "'xc4 0-0 7 e4 ~g4
8 ~e3 lDfd7 9 l:dllDc6 10 ~e2 lDb6
11 'W'c5 "'d6 12 e5
This was my third match against
Karpov and he had prepared most
diligently for it. Here he sacrifices a 21 ... a3?
pawn - something he had rarely done I was still playing in the spirit of
before in this kind of situation. I think Karpov, in the hope that somehow I
13
would be able to get down to the

a2-pawn. However it proves to be an
illusion.
22 b3 ltJa7
This was not my day, I was unable to
push either of my rook's pawns, but
putting the knight on the edge was also
unfortunate. 22 .. .iLg7 was probably
better.
23 .tid7! iLeI 24 l:be7 iLb2 25 ltJa4
ltJbS 26 l:txe6 20 liJh3!
Now White is already a pawn up. A strong move that my team and I
26 .. JHd8 27 l:tb6 l:1dS 28 .i.g3 missed during preparation. I hoped the
g7-bishop would become a powerhouse
on the long diagonal, but it stayed
buried all the way to the end.
20 ... aS?
Again I push the rook's pawn just

like Karpov.
21 f3 a4
The plan must be pursued.
22 :thel!
28 ... ltJe3 Another strong move as it preserves
This is the closest I got to attacking the e5 pawn and makes sure the
that a2-pawn. g7-bishop remains bottled up. I should
have copied this aspect of Karpov's
29 ltJxe3 ~xc3 30 e6
style!
The c-pawn simply kills Black.
22 ... a3
30 ... ~d4 31 l:tb7 1-0 Nothing will divert me from pushing
To end the misery I resigned. the a-pawn.
23 ltJf2
Game 5, World Championship,
1 d4 ltJf6 2 e4 g6 3 ltJe3 dS 4 iLf4

iLg7 5 e3 eS 6 dIeS 'ii'aS 7 lteI ltJe4
8 exdS ltJxe3 9 'ii'd2 'ii'xa2 10 bxe3
'ii'xd2+ 11 ~xd2 ltJd7 12 iLbS 0-0
13 .i.xd7 .i.xd7 14 e4 fS 15 eS e6 16 e4
.tife8 17 e6 bxe6 18 d6 eS 19 h4 h6
14
Anatoly Karpov the J2th
23 ... a2
The pawn has got this far yet it is to
no avail. I was so happy to see the open
road ahead but should have checked
more carefully where that road would
lead.
24 lbd3 l:ta3 25 l:r.al g5
A desperate attempt to stir things
up.
26 hxg5 hxg5 27 iLxg5 ~f7 28 .ltf4
l:lb8 29 l:tecl iLc6 30 l:r.c3 1:a5 31 l:tc2 Here I resigned and decided that in
l:r.ba8 32 lbcl 1-0
the future I would be far more cautious
Finally White wins the a-pawn.
about following Karpov's method of
Black spent four tempi advancing the
pawn to its doom. play.
Karpov contributed to one of my losses in the final of a knockout

tournament I played on the internet. From the first position below he taught
me that in a rook ending, with 3 pawns versus 4 on one side of the board, the
game can be saved.
The second diagram shows the very similar position that I reached.
J.Piket - G.Kasparov
V.Korchnoi - A.Karpov
V.Korchnoi - A.Karpov 8 cxd5 xd5 9 xd5 exd5 10 • xe7

Game 5, World Championship, ~xe7 11 g3 .lta6 12 e3 c5 13 dxc5
Merano 1981 bxc5
Two games later in the match Karpov
1 c4 e6 2 lbc3 d5 3 d4 iLe7 4 lbo played 13 ... .ltb7 and drew.
lbf6 5 iLg5 h6 6 iLh4 0-0 7 :tel b6 14 iLxa6 lLlxa6
15
15 'ilVxd5 28 ... h5!

Petros ian, the specialist of this line This is a very useful refinement in
where White delays e2-e3, did not take the ensuing rook ending.
the pawn and went on to beat Portisch 29 'it>g2 ~g7 30 11a5 tLJc6
this way. He finally agrees to defend the
15 ... tLJb4 16 'Vic4 iVf6 17 tLJh4 'ilVxb2 4 against 3 ending. But Black was not
18 0-0 'ii'xa2 forced to give up the pawn as 30 ... ~f6
was an option. This misled me and gave
me the impression this it is an easy
draw.
31 tLJxc6 l:txc6 32 lha7 lha7
331ha7
33 ...nc2
Karpov often pins pieces. With
White's king on f2 the pawn can't go
much further. I used to think it requires
some effort to draw this type of
position but the ease with which
Taking the last white queens ide pawn Karpov held this one made me think
Black can't lose at all. I was wrong.
is an achievement, yet there is no
34 e4 l:tc3 35 l:ta2 <it>f6 36 0 l:tb3
guarantee of a draw as the knights are 37 ~f2l:tc3 38 <it>e2 l:tb3 39 11a6+ ~e7
still on the board. 40 l:ta5 ~f6 41 nd5
19 ~xa2 tLJxa2 20 lhc5 ~fc8
21 lIa5 tLJc1 22 tLJf5 nc7 23 tLJd4 nb8
24 na 1 tLJd3
Karpov is not yet ready to enter the
4 pawns against 3 rook ending. If
24 ... tLJb3 25 nfb 1 llcb7 26 Itxb3 l:txb3
27 tLJxb3 l:txb3 28 l:txa7.
25 nfdl tLJe5 26 l:ta2 g6 27 ndal
l:tbb7 28 h3
16
41..J:ta3 42 nd6+ ~g7 43 h4 nb3 Would you believe what happened?

44 nd3 %:tb5 45 'ite3 liz-liz Karpov You will see it in the Fischer chapter.
effortlessly held this position. 24 ... 4Jd4 25 ~g2 neR 26 l:tb 1 nfd8
27 ~xd4 %:txd4 28 b4
J.Piket - G.Kasparov In my case this plan did not work, so
KasparovChess Grand Prix I was optimistic that I would do well
60 minute game, Internet 2000 against it.
1 ttJf3 4Jf6 2 e4 c5 3 4Je3 d5 4 cxd5 28 ... axb4 29 axb4 'iVd7 30 bxe5
4Jxd5 5 g3 4Jc6 6 ~g2 4Je7 7 d3 e5 bxe5
8 0-0 ~e7 9 4Jd2 .i.d7 10 4Je4 0-0 Here I realised I couldn't win. But
11 ~xc6 ~xe6
okay, no problem. It was a knockout
final and I thought I would draw the
next game with the white pieces.
31 1:[bb2 h6 32 1:[a2 <ith7 33 1::ta5
:d8 34 'iVxe5 .i.xd3
12 4Jxe5
Piket accepts my pawn sacrifice.
12 ... ~e8 13 'ilfb3 ~f6 14 4Jg4 ~d4
15 e3 ~xe3 16 1i'xc3 b6 17 f3 .ltb5
18 4Jf2 'ifd7 19 e4 4Je6 20 .lte3 a5 21
35lhd3
:ladl l::tad8 22l:td2 ~e6
After 35 l:tb2 l:tc8 36 .f5+ 'iYxf5
According to my opponent's analysis
37 l:txf5 it.c4 it is hard to do anything
the queen should go to b7.
23 %:tel 'ifb7 with White's pieces.
35 ... l::txd3 36 4Jxd3 'tIi'xd3 37 Ua2
'iWb3
24 a3
I also tried a6 and b5 with Black in a
number of English opening games. 38 'ii'e2
17
Anatoly Karpov the J 2'h
With queens on the board White can't Korchnoi did not try anything like
really push the pawns, therefore it this - and Korchnoi was a really strong
should be an easy draw, but I knew how endgame player
easily Karpov drew with Korchnoi, so I 42 ... Ae3
decided to follow him. I was also aware I just keep moving like my
that Karpov wasn't able to squeeze a predecessor before me.
win against Olafsson when he had an 43 ~h4 ~g7 44 ~g5
extra pawn in this kind of endgame. Here I deviated from Karpov and
Even before the Karpov game I knew removed the rook from the third rank.
this position was a draw, however it
was Anatoly who convinced me it was
easy and made me play too casually.
38 ..:.xc2+ 39 lhc2 h5
44 ...:tel?
This was my independent idea - but
it loses. I was short of time. Correct'
was 44 .. .l':ta3! 45 Ac7 :a5.
I play just like Karpov. 45 Ac7 l:le2 46 :e7! Aa2
40 f4 g6
See diagram on page J5.
Karpov also had his pawn on g6.
4i e5 lId3
I'm just following Karpov, who kept
his rook on the third rank, did nothing
and held easily.
47 f5!
This is nasty indeed. I was in time
pressure because it was a 1 hour game
with no increment.
47 ... gxf5 48 e6!
Ohno.
48 ... h4 49 Axfi+ ~g8 50 ~f6 1-0
42 ~h3 And I had to resign.
18
Anatoly Karpov the 12'"
Karpov has played some very well-known games in which he moved his
knight backwards to the first rank. I also know some games where he placed
the knight on the rook file. Here are three of his positions - followed by three
of mine.
Sadovsky - A.Karpov G.Kamsky - G.Kasparov
Z.RihU - A.Karpov Y.Nikolaevsky - G.Kasparov
J.Nunn - A.Karpov L.Zaid - G.Kasparov
19
Sadovsky - A.Karpov
USSR Olympiad 1967
1 c4 ttJf6 2 ttJc3 e5 3 g3 g6 4 ~g2

~g7 5 e4 d6 6 ttJge2 ~e6
Karpov was fairly young when he
played this game.
7 d4! 0-08 d5 ~d7 9 0-0 lDh5

31. .. ~g4 32 f4?! ~Ie2 33 'ii'xe2
exf4 34 ~d4 %:l6f7 35 b5 lDf5
Karpov develops his knight on the
For a long time the knight which had
edge.
been on h5 had no useful move, now it
10 ~e3 f5 11 ~d2 ttJa6
intervenes with decisive force.
The other steed does the same thing. 36 .i.n fxg3 37 .i.xg3 l2Jxg3
12 t1 .l:tf7 13 a3 ttJc5 14 'ii'c2 'it'e8 38 ':xg3
15 b4 ttJa4
38 ...'ii'e5
The rest is simple.
The third knight move to the edge.
39 1:.aa3 axb5 40 cxd6 b4 41 :tad3
16 c5 f4 17 ~f2 fxg3 18 hxg3
cxd6 42 'ii'el ltc7 43 ltgO 1IxO
Now the players enter a long
44 .i.xt1 ~c2 45 ~g2 ltel 46 lIdl
manoeuvring phase.
~e3+ 47 ~hl 1:.xdl 48 1i'xdl b3
18... l2Jxc3 19 l2Jxc3 'ile7 20 'ii'd2
49 'ii'xb3 'ii'g3 50 'ii'xb7+ ~h6 0-1
~af8 21 'Vie3 a6 22 a4 ~f6 23 'it'd3
~g5 24 :ta3lDg7 25l2Je2 ~e8 26 ~c2
Z.Ribli - A.Karpov
h5 27 iLel ~f6 28 Jtn ~h7 29 ~el Tilburg 1980
~h6 30 iLf2 Jtc8 31 l:Ifal?
This is overdoing a good idea. One 1 c4 e5 2 lDc3 lDc6 3 g3 g6 4 ~g2
piece too many goes to the edge. ~g7 5 d3 d6 6 e3 lDge7 7 l2Jge2 0-0
31 :tfb 1 was correct. 8 0-0 ~d7 9 h3 l:.b8 10 'ii'd2 iLe6
20
Anatoly Karpov the J2,h
llllJd5 Karpov achieves symmetry with his

The knights still look to the centre, knights and rooks on the two sides of
but maybe this just diverts the attention the board.
of the opponent. 16 ltJe2 b5 17 llJcl llJb6 18 llJb3
11 .. :ii'd7 12 ~b2 f5 13 f4 b5 14 l:tbl ttJe4 19 l::thgl as 20 ~cl a4 21 llJe5
~xc5 22 dxe5 llJxe5 23 'iYg3 llJef3
bxc4 15 dxc4 e4 16 b4? llJa5!
24 l::thl e5 25 i.e4 'ii'f6 26 ~xf3 ltJxf3
27l::td3
See diagram on page J 9.
Karpov puts his knight on the edge

and wins instantly. What a lucky
fellow!
17 llJd4 llJxc4 18 liJxe7+ 'ifxe7
19 'iVc2 ~xd4 20 exd4 llJb6 21 ~b2
ltJd5 22 a3 ltJe3 23 'it' c3 ltJxfi +
24 ~xfl ~d5 25 i.e4 'ife6 0-1
The next game is unusual, as I was

not selecting lost games from a 27 ... e4
champion, but here Karpov reached Black is winning. If I wanted to be
a position with two knights on the sarcastic I could say he has a winning
edge. edge (advantage) in the middle of the
board.
J.Nunn - A.Karpov 28 l::txf3 exf3 29 l:tel + <it>fS 30 ~g5
3rd Amber-rapid, Monte Carlo 1994 'ii'f5 31 h4 ~g8 32 l:e5 'ii'd7 33 l:te7
'ii'f5 34 l:e5 'iVe8 35 i.f6 tth7 36 h5
1 e4 c6 2 d4 d5 3 e5 i.f5 4 llJe3 e6 'ii'fS 37 a3 'iVxe5 38 'ifgl l:h6 39 'ii'g5
5 g4 i.g6 6 ttJge2 ttJe7 7 i.e3 h5 8 ttJf4 'it'fS 40 hxg6 fxg6? 40 .. .lbg6 wins.
hxg4 9 llJxg6 llJxg6 10 'ii'xg4 ltJb4 41 l:te7 .:te8 42 l:d7 :'hl + 43 ~a2
11 ~d3 g6 12 0-0-0 ~e7 13 ~blltJd7 'it'h6 44 'ifg3 :thel 45 'ii'e7 IHe7
14 llJe2 ltJb6 15 llJr4 llJa4 46 l:txe7 lbe7 47 'ii'xe7 'ii'fS 48 'ii'e6+
21
jVf7 49 'iWxc6 d4+ 50 ~bl d3 51 cxd3 17 g3 liJdf6 18 c5 ~d7 19 lIb3 ~h6

~h7 52 'it'xf3 'iWb3 53 'iWh3+ ~g8 Against Karpov in Tilburg 1991 I
54 'ir'h8+ ~f7 55 'ii'g7+ 'it>e8 0-1 played 19... ttJxg3 and the game
continued 20 bxg3 lLlh5 21 f4 exf4
Karpov's knight on the brink was in 22 c6. I later drew the game, despite
my mind almost all the time. Let me finding myself in an almost hopelessly
show you one of my games from the lost position. See page 34.
time when I reigned as world champion 20 ':'c3
and two when I was a junior player.
G.Kamsky - G.Kasparov
Dortmund 1992
1 d4 ttJf6 2 c4 g6 3 ttJc3 ~g7 4 e4 d6

5 ttJf3 0-0 6 ~e2 e5 7 0-0 ttJc6 8 d5
ttJe79 ttJd2 as 10 a3 ttJd7
There is no chance of this knight
going to h5, but with a couple of moves
I manage to close the diagonal and 20 ... ~f4
make it' possible to place the other This was a novelty in 1992.
knight in an attacking position on that 21 cxd6
square. I was hoping for 21 gxf4?! as I could
11 l':tbl f5 12 b4 ~b8 13 f3 then have demonstrated some
remarkable footwork along the touc~
line. 21.. .ttJx f4 22 ~c4 (22 tiJb3
tiJ6xd5!) 22 ... tiJ6h5. The other knight
goes there as well. Please note that all
Black's moves now will be played on
the flanks. 23 tiJb3 fxe4 24 fxe4 tiJh3+
25 'it>h 1 'ii'h4 26 tiJxc7 ':xfl + 27 ~xfl
:f8 and Black's subtle play has earned
him a winning attack.
21 ... lDxg3
After 21.. .cxd6 22 lDc7 .i.xg3!
The diagonal is now closed and it's 23 hxg3 lDxg3 24 Ae 1 lDfh5 Anand
worth spending a few tempi getting the produced some analysis and concluded
knight where Karpov had put it. that Black has the initiative;
13 ... ttJg8 14 'ii'c2 ttJgf6 15 ttJb5 axb4 Not 21.. ..i.xb5 22 dxc7; or 21.. ..~xg3
16 axb4 ttJh5 22 dxc7 "ile7 (22 ... ~xh2+ 23 'it>xh2
Mission accomplished. lDg4+ 24 fxg4 ~4+ 25 l:.h3 wins.)
23 d6! when Black is in trouble.
See diagram on page 19. 22 bxg3 ttJb5
22
Anatoly Karpov the 12 th
Black gets his knight to the h-file, but

does not have enough fire-power to
back it up. This is rather transparent,
but I also have one particularly nice
memory of a surprising knight check on
h3.
Here it is:
A.Grischuk - G.Kasparov
The other knight follows to h5. I did Linares 2001
not pay due attention to the fact that
Karpov had not played with such
ferocity.
23 gxf4 lLlxf4
After 23 ... ~4 24 :f2 lLlxf4 25 .i.f1
.i.xb5 26 dxc7!! wins as Anand pointed
out.
24 ... lLlh3+!! 25 c;t>g2 l:hb5 26 ~g3

lLlg5 27 .i.f2 .i.b7 28 .i.gl l:tc8 29 h4
.i.xD+ 30 l:hD lLlxD 31 WxD ~xc5
32 lLlxc5 l:tbxc5 33 .i.xc5 llxc5 34 c3
h5 35 gxh5 :xh5 36 b4 axb4 37 cxb4
l:[xh40-1
24.i.c4! Now back to my game with Kamsky:

If 24 :f2 lLlh3+ - a common theme 25 <sPhl
by now - (if 24 ... 'ii'g5+ 25 ~hl! 'irb4+ If 25 ~g2 'ti'g5+ 26 ~xh3? f4+
26 :h2 'iVe 1+ 27 .i.f1 .i.xb5 28 :xc7 27 ~h2 'it'g3+ 28 ~hl 1fh3+ 29 ~gl
lLlh5 - even this doesn't help - 29 'ifb2 'ii'g3+=
wins.) 25 ~f1 (25 ~g2? 'ii'g5+!
26 ~xh3 'ifg I! leads to a checkmate.)
25 ... lLlxf2 26 ~xf2 'irb4+ 27 ~g2
(27 ~f1 'irbl+ [27 ... f4 28 .i.c4!]
28 'iiff2 'irb4+=) 27 ... f4 28 .i.f1 White
survives the attack and wins with the
extra material.
24 ... lLlh3+
If 24 ... 'ifh4 25 lLlb3 lLlh3+ 26 ~hl;
or 24 ... 'it'g5+ 25 ~f2 'irb4+ 26 ~e3.
23
Allato!y Karpov the 12th
2S ... 'VWh4 Y.Nikolaevsky - G.Kasparov

This sets up a dangerous looking Moscow 1976
battery.
26 lLJb3 fxe4 1 lLJf3 lLJf6 2 g3 g6 3 .i.g2 .i.g7
After 26 ... lLJf2+ 27 ~g2 there is not 4 0-0 0-0 S d4 d6 6 b3 cS 7 .i.b2 li)c6
enough juice left in the battery. 27 ... f4 8 dS lLJaS
28 'ikxf2 'ikh3+ 29 <it>g 1 wins as Anand The knight has reached its planned
pointed out. destination on the edge of the board
27 'Yih2 fairly early in the game.
Not 27 fxe4?? lLJf2+ 28 ~g2 .lth3+
29 ~gl 'iig4+ 30 <it;h2 'ikg2 mate. See diagram on page J9.
27 .. JHs 28 f4!
After 28 fxe4 l:ih5 (28 ... 'iixe4+ 9 ne1 .ltrs
29 'Yig2) 29 ncO wins as well. Just like Anatoly Evgenievich I am
28 ...lihS prepared to give up a tempo.
Black's pieces are picturesquely but 10 ltJbd2 'YIkc8 11 e4
precariously placed on the h-file. My opponent takes the free tempo,
just like Karpov.
11 ..• Jtg4 12 c4
Black can build up pressure on the
c4-square.
12 ... ltJd7 13 it.xg7 ~xg7 14 f*'c2
ii.xfJ 15 ltJxfJ ~c7 16 'YIkc3+ ~g8
17 ii.h3 b6
29 ~g3 'ii'xg3
This is tantamount to resignation but
other moves also lose. If 29 ... lLJxf4+
30 'ii'xh4; or 29 ... 1IVf6 30 fxe5 'YIkg7
31 dxc7; or 29 ... 'ikd8 30 dxc7 lLJxf4+
31 ~gl 'ii'f6 32 .ltxf4.
30 l:hg3 exf4 31 .ltb2+ ~g8
32 dxc7! .ltxbS
After 32 ... fxg3 33 d6+ .lte6 34 .ltxe6 18 eS
is mate. White correctly opens the position in
33 .ltxbS fxg3 34 ~g2! lLJgS the centre. Now the knight is missing
And after 34 ... e3 35 .ltd7 wins. from the action.
3S d6 nh2+ 36 <it;xg3l:txb2 37 .ltc4+ 18 ... dxeS 19 it.xd7 'iWxd7 20 ltJxeS
<it;g7 38 d7 1-0 'YIkd6 21 lLJg4 hS 22 ltJeS 'ii'f6
24
34 l:txf8+ <;t;xf8 35 l:Ixe5 'ii'g4+ 36 Wfl

'irb.3+ and White can do nothing with
his extra exchange. White has no time
to make room for his king with 33 h3 as
after 33 ... %1d2 34 tbe5 l:txf2 35 tbxg6
l:t2xD 36 tbxh4 l:xe3 the position
is equal.) 33 .. .'iittg7 34 l:txf8 ~xf8
35 l:tD+ 'i;g7 36 'iVe8 'ii'g5+ 37 :g3
l1dl+ 38 'i;g2 'iVf6 39 'iYg8+ 'i;xh6
40 1:th3+ 'iitg5 41 ltg3+.
23 %:te3 33 l:txf6 'iYxf6 34 tbh6+ ~g7
The e7-pawn is an obvious target. 35 tbg4 'ii'd4
23 ... l:tadS 24 %10 '1Wg7 25 l:Iel 1:d6 After 35 ... 'ii'g5 36 h3 White will
26 b4 tbb7 27 '1Wa3 as 2S bxa5 tbxa5 quickly bring his queen over to the
It is a true delight that the knight can kingside, while Black's knight has to
return to a5. remain on the queenside.
29 'ii'a4 'ir'h6 30 l:tee3?!
30 tbd7 wins the exchange.
30 ... h4 31 gxh4 'ir'xh4 32 tbxf7?
An imaginative trick that wins the
game, but objectively this thrust spoils
White's position. 32 'ii'c2 keeps an
edge.
36 l:txe7+
The pawn is gobbled up and Black's
king remains vulnerable. Black is
simply lost.
36 ... WhS 37 h3 'iVc3 3S 'iYd7 'iYxh3
39 :leS "f3 40 :le7 'iYh3 41 'iYe61i'h5
42 :leS Wg7 43 'ir'd7+ :f7 44 'fics
32 ...1:f6?
1i'h7 45 d6 g5 46 d7 'iVbl+ 47 Wg2
The exploitation of the pin along the
1-0
fifth rank by 32 ... .:xd5!? allows many
tactical possibilities, however Black
L.Zaid - G.Kasparov
almost miraculously survives in every
Leningrad 1977
variation. 33 tbh6+ Going after the
king achieves no more than a perpetual. 1 d4 tbf6 2 c4 g6 3 tbf3 i.g7 4 g3 d6
(33 tbe5 wins the exchange but leaves 5 i.g2 0-0 6 0-0 tbc6 7 tbc3 a6 S d5
his king too exposed, e.g. 33 ... l:xe5 tba5
25
Anatoly Karpov the J 2th
The knight naturally goes to the edge. Malta 1980, or Bouaziz-Karpov,

9 lbd2 c5 10 ~c2 l:tb8 11 b3 b5 Hamburg 1982. Two of these were
12 ~b2 played after this game and so I hadn't
seen them. In a way my retreat is more
effective than Karpov's. It ends the
12... bxc4 13 bxc4 ~h6 14 lbcbl e5 game far more quickly, in just two
15 ~c3 iLd7 16 lba3 moves.
Sad, but my opponent also knows the 36 iLe7 ne8 37 W'd5 1-0
knight to the h- and a-file strategy.
16 ... iLg7 17 !:tabl ~c7 18 e4 h5 Leaving out the analysis, here are the
I should have followed in Karpov's three retreat pearls mentioned:
footsteps and played 18 ... lbh5.
A.Karpov - B.Spassky
19 f4 l:.b4 20 'i/Vd3 lbb7
Game 9, Candidates semi-fmal
I wanted to leave the knight where it
Leningrad 1974
was, but the fact that it was undefended
worried me.
21lbc2 ~xbl 22 ltxbl h4?1
There is no time to free the h5-
square for the knight. After 22 ... exf4
23 gxf4 ne8 Black is in the game.
23 fx·e5 dxe5 24 lbO! hxg3 25 iLxe5
gxh2+ 26 iLxh2
White's centre is rock solid.
26 .. :iVc8 27 lbe3 ttJg4 28 lbxg4
~xg4 29 ttJe5 ttJa5
24 tObl!! 'ii'b7 25 <it>h2 <it>g7 26 c3
Now White no longer attacks the a5-
tOa6 27 ne2 nf8 28 tOd2 iLd8 29 tOD
square. So I waste no time putting the
f6 30 %td2 il.e7 31 'ife6 l%ad8 32 lhd8
knight back on the edge of the board.
iLxd8 33 l:1d 1 tOb8 34 iLc5 l:th8
30 nfl iLh5 31 d6 iLxe5?
35 l:txd8 1-0
If 31 ... iLg4 32 ttJxg4 'i/Vxg4 33 'iVd5
~d4+ 34 ~h I 'Wid7 Black still resists. A.Karpov - M.Quinteros
32 iLxe5 'Wie6 33 ~f6 Malta Olympiad 1980
The exposes the weak black king
and the d6-pa,wn becomes more
threatening.
33 ...ttJc6
He must come back to hold the pawn.
34 'ife3! ~h7 35 'ii'xc5 lbb8
Karpov played some remarkable
knight retreats during his career, for
example: Karpov-Spassky, 9 th game,
Moscow 1974, or Karpov-Quinteros,
26
Anaioiy Karpov the 12iH
32 lObI! lOg7 33 lOd2 lOe6 34 b4 2S ... lOb8!! 2610D lOd7 27 '1t>g31OcS

l:td8 3S lOc4 l:td4 36 lOd6 l:txb4 28 l:tdl as 29 ~f2 ~a6 30 'lte2 lOa4
37 lOxb7 l:tbS 38 h4 hS 39 ~f2 'ltd7 31 d4 l:tb6
40 lOd6 l:txcs 41 l:tb2 1-0
S.Bouaziz - A.Karpov
Hamburg TV 1982
32 dxeS ':xb2+ 33 ~fl ltJxc3

34 exd6 cxd6 3 S %hd6 l:tb 1+
36 lOel ~f6 37 l:td2 bS 38 l:tc2 b4
39 ~f2 l:ta 1 40 eS+ ~xe5 41 lOD+
~e4 42 ltJd4 ~d3 0-1
Karpov sacrificed the e5-pawn in a sharp Sicilian against Ljubojevic.

I decided to use this weapon as well.
A.Karpov - L.Ljubojevie G.Kasparov - A. Yermolinsky
A.Karpov - L.Ljubojevie
Turin 1982
1 e4 e5 2 1Of3 d6 3 d4 lOf6 4 lOc3

exd4 5 lOxd4 a6 6 il.e2 e6 7 f4 'ike7
8 0-0 b5?! 9 il.f3 il.b7 10 e5 dxe5
11 fxe5 lOfd7 12 il.f4 b4
27
13 lDe4 :u l:tadl f6
Karpov sacrifices the pawn. This Moving the knight with 21 ... lDc6
game was so convincing that the 22 "xg5 'fixd6 23 lDf5 "e5 24 lDh6+
position never occurred again. wins, while if 21.. .h6 22 h4 .i.xh4
13 ... lDxeS 23 'it'xh4 "xd6 24 lDf5 decides.
22 .i.xb8 l:taxb8
14 'it'hl! ~e7
Not 14 ... lDbc6?? 15 lDxc6 .i.xc6
16 .i.xe5 "xe5 17 lDf6+ winning nor
14 ... lDbd7 15 lDg5! and White has nice
play for the pawn.
IS lDgS! .i.xgS 16 .i.xb7! "xb7
If 16 ... l:ta 7 17 lDxe6 or 16 ... .i.xf4
17 .i.xa8 lDg6 18 'iW'e 1 0-0 19 .i.e4 and
23 h4! .i.xh4
White is better.
Retreating with 23 ... .i.h6 is met by
17 .i.xeS 0-0
24 lDf5 'ii'c7 25 lDxh6+ 'iPh8
17 ... lDd7 allows 18 .i.xg7 l:tg8
19 lDxe6.
18 "g4 'fie7?
After this White's advantage is
decisive. Also after 18 ... h6? 19 lDxe6
"d7 20 i..xg7 wins. Better is 18 ... lDd7!
19 'fixg5 f6 20 i..xf6 but White is still
somewhat better.
26 l:td8+!! This lovely shot decides

the game instantly.
24 'ih:h4 l:lc4 2S "g3 l:tbc8 26 ltJrs
"a7 27 lDd6 l:t4cS 28 'fib3 1-0
G.Kasparov - A.Yermolinsky
Leningrasl1975
19 "g3! %:tc8
After 19 ... l:td8 20 %:tad 1!. -1 e4 cS 2 lDf3 d6 3 d4 cxd4 4 ltJxd4
20 i..d6 'ii'd7 ltJf6 S ltJc3 lDc6 6 .i.gS a6 7 'ifd2 e6
Or alternatively 20 .....d8 21 lDxe6! 8 0-0-0 .i.d7 9 f4
28
Anatoly Karpov the J2rh
9.•. b5 42 ... ~d6 43 'iVb8+ ~xd5 44 'iVd8+

This is a rarely played line at the top ~e6 45 'iVe8+ ~f5 46 'ifd7+ ~g6
level. It leads to very exciting games. 47 'ili'g4+ 'itf6 48 ttJc3 'iff} + 0-1
10 lDxc6 Karpov-Torre, Manila 1976.
Karpov was nicely beaten by Torre
with 10 'ili'e 1. Here is Torre's Back to the game:
masterpiece. 10 ... lDxd4 11 ':xd4 'irb6
12 :d2 JJ..e7 13 JJ..d3 b4 14 lDdl JJ..b5
15 lDf2 h6 16 JJ..h4 g5 17 fxg5 bxg5
18 JJ..g3 lDh5 19 lDg4 lDxg3 20 hxg3
:xhl 21 -.xhl l:tc8 22 cwttbl JJ..xd3
23 cxd3
10 ... .i.xc6 11 .i.d3 .i.e7 12 e5 dxe5

13 fxe5lDd7 14 .i.xe7 'ibe7
23 ... 'iIi'd4 24 -.dl a5 25 It:Jh2 g4

26 lDxg4 JJ..g5 27 ':c2 ':xc2 28 ~c2
a4 29 a3 b3+ 30 cwttb 1 d5 31 exd5 -.xd5
32 It:Jf2 'ili'xg2 33 lDe4 ~e3 34 It:Jc3
-'c6 35 d4 -'c4 36 d5 e5 37 'iibl -.d3+
38 ~al .i.d4 39 'ii'h8+ ~d7 40 "a8
-.f1+ 41 lDb 1 'ii'c4 42 'irb7+ 15.i.e4
29
I did not mind sacrificing the e5

pawn, hoping I would get enough play
for it in return.
15.....c5 16 lIhe1 lIa7 17 ~Ic6
"xc6 lS"f2 "c5 19 lIe3 0-0
Black has equalised by moving his
king into safety.
20 4Je4 "xe5
See diagram on page 2 7.

32 ...lIe5 33 "d2 lIfeS 34 lIe3 "h6
21 lIg3 35 g3 "h3 36 lIc6
Karpov also allowed short castling 36 lId6 is an alternative.
and caught Ljubojevic's king on the
36 .....f5 37 "c3 "d7 3S b3 :f5
kingside. I hoped my attack would
39 <it'b2??
bring Yermolinsky down too.
21..JlaaS 22 Ibd7 "Ie4 23 lIo White has had to play carefully for
jfg6 quite some time, now a losing mistake
leads to disaster. The king blocks the
queen. After 39 ~a2 :0 40 lIc5 "dl
41 'iib2 White is in the game.
24 a3
White's heavy pieces control many
files and ranks.
24 ...lIacS 25 ~b1 e5!
Keeping White busy and holding on
to the pawn. 39 ...:f1!!
26 lIg3 "e6 27 "d2 g6 2S lIb3 "f6 SUIprisingly catching the king on the
29 "h6 "g7 30 "g5 lIceS"31 lId6 e4! first rank..
321ba6 40 lIf6 :bl 41 .c6 "iid4+ 42 l:tc3
Material equilibrium has been lIf8 43 l:tIn lIxn 44 "xb5 l:tIh2
restored. White is still not worse, but he 45 b4 lbc2+ 46 ~xc2l:tf2+ 47 <itb3
must play with care. 'ii'd1 + 4S ~c4 'ii'e2+ 0-1
30
Anatoly Karpov the 121h
Karpov sacrificed the b6-pawn in S.Garcia Martinez - A.Karpov

a hedgehog position against Garcia
and went on to win.
And below is another example

from my annoying predecessor,
followed by my game against him
where he himself took the b6-pawn.
J .Saren - A.Karpov A.Karpov - G.Kasparov
Garcia Martinez - Karpov

Madrid 1973
1 e4 eS 2 lLlf3 e6 3 d4 exd4 4 lLlxd4

a6
Interestingly Karpov stopped playing
the Paulsen pretty soon after he became
world champion. Avoiding direct
confrontation suits his style better. 13 ... lLle6
S R.d3 lLlf6 6 0-0 d6 7 e4 We7 8 'it'e2 A slightly unusual way to develop.
g6 9 f4 R.g7 10 ~hl 0-0 11 lLle3 b6 Here Black only defends the b6-pawn
12 ~d2 ~b7 13 lLlf3 with his queen. 13 ... lLlbd7 is usual.
31
14 l:.'tac1 l:.1ae8 IS 'it'n 26 lDxe4 dxe4 27 lDxf8 l:hf8

This is a mUlti-purpose move. White Black has two pieces against the
can think about attacking the king with rook. The position is roughly equal and
Vi'h4 or... the stronger player will outplay his
IS ... ttJg4 16 'it'gl opponent.
White keeps an eye on the b6-pawn.
16 ... fS 17 exfS gxfS 18 h3?! By the way, this game misled me in
White diverts the knight to a better several ways. I lost twice by opting for
place and weakens b6 as well. He could two pieces against a rook. Once against
play 18 ttJd5 ~n 19 ttJxb6. Kappe and once against Romanishin.
18 ... tt:)f6 19 tt:)dS ji'd8 20 ~xb6 Here are those examples:
See diagram on page 31. A.Kappe - G.Kasparov

20 ... 'ii'xb6 21 ttJxb6 Cagnes-sur-Mer, 1977
This is the idea implemented later.
Black can give up the b6-pawn in such
a situation.
21. .. ttJe4 22 Jtxe4 fxe4 23 tt:)gS tt:)d8
Karpov sacrifices a pawn and can
also afford to step back with his knight.
Where is it going to? To the edge of the
board of course!
24 lIfe 1 ?!
I t is hetter to defend the f4-pawn by
24 lIce I ! d5 25 cxd5 exd5 26 ttJd7. 11 tt:)dxe4 'ii'xd 1 12 lDxf6+ it.xf6
24 ... dS 2S ttJd7 13 ttaxdl Jtxg5 14 Jtxb7 it.e4
After 25 cxd5!? exd5 26 ttJd7 h6 15 Jtxa8 Jtxa8
27 ttJxf8 llxf8 28 ii.b4 lixf4! 29 Jte7
hxg5 30 ii.xd8 nn 31 b3 d4 Black's
central pawns are menacing. The
position is hard to evaluate over the
board, however White might be better
here.
2S ... h6
16 h4 Jte7 17 Jte5 lDa6 18 c3 it.c6

19 ~f4 Jtf8 20 nd2 f6 21 nfdl cJitfl
22 'iitfl h6 23 Jte3 ttJc5 24 lld8 a6
25 llxe8 'it>xe8 26 f3 ttJe6 27 'iitf2 h5
28 c4 ii.d6 29 llb 1 a5 30 .ltd2 It)d4
31 .i..f4 ~e7 32 ii.xd6+ ~xd6 33 lidl
32
~c5 34 ~e3 lbe6 35 g4 a4 36 gxh5

gxh5 37 bxa4 ~xc4 38 1:c 1+ ~d5
39 a5 .i.b5 40 :gl lbd4 41 .:tg8 c5
42 l1h8 ~c6 43 :xh5 .i.xe2 44 l:th7
.i.xD 45 a6 ~b6 46 a7 .i.b7 47 h5
Q;xa7 48 :f7 ~b8 49 l:txf6 .i.c8
50 ~e4 .i.e6 51 h6 .i.xa2 52 ~e5 1-0
O.Romanishin - G.Kasparov
4·-teams, Moscow 1981 38 ... lbxe5 39 fxe5 .i.g7 40 .:tf7 ~h6
41 h4 ~h5 42 ~h3 .i.e8 43 :a7 .i.g6
44 l:txa6 .i.d3 45 1:[f2 .i.xe4 46 l::ta3
26 l:he7 lbb6 27 :b7 lba4 28 :b8+

.i.f8 29 c4 ~g7 30 ..tg2 .i.d6 31 :la8
46 ... .i.h6?
lbb2 32 a4 lbxc4 33 as ttJe5 34 l:tc8
Karpov rarely misses a trick in
1-0 endgames. However 46 ... .i.xe5 was
winning after 47 l::ta5 .i.d5.
Back to the game. 47 llg3 :1a8 48 1:f7 1:a1 49 l:[h7
l:h1+ 50 ~g2 l:txh4 51 qg,gl
28 b4 .i.e6 29 a4 .i.xa4 30 llxe4 .i.e6 51. .. .i.e2 52 'it>f2 .i.g4 53 b5 .i.f5

31 .:te2 h5 32 ~h2 h4 33 g3 hxg3+ 54 :h8 :Ib4 55 1:g1 ltb2+ 56 ~f3
34 ~xg3 ~h7 35 .i.e3 .i.h6 36 :n l:tb3+ 57 ..tf2 .i.e4 58 1:g3 l::tb2+
:g8+ 37 ~h2 ttJrT 38 .i.e5?! 59 ~gl .i.f5 60 :g2 l:txg2+ 61 ~xg2
33
~e4+ 62 ~g3 <i!?g6 63 b6 ~d5 64 l:tb8 Or should I say has any of my

<it>f5 65 b7 ~xe5 Predecessors had such an endgame?
If 65 ... ~f4+ 66 ~h4 ~xe5 67 ~g4
ii.e3 68 ~h5 holds. A.Karpov - G.Kasparov
Tilburg 1991
66 <i!?g4?
After 66 llh8!! White contrives 111 ltJg6+ ~g8 112 ltJe7+ ~h8
to engineer a miraculous escape. 113 ltJg5 l1a6+ 114 ~f7 :f6+! Yl-Yl
66 ... ii.f4+ 67 <it>g4 ii.xb7 68 nh5+!! Capturing the rook results in stalemate.
Back to the main game:
Recovering a piece, Without this

move White would lose. 68 ... 'it>e4
69 l:tb5! and White wins a bishop. 66 ... ii.e3 67 ~g3 ~g5 68 ~f2 i.e7
Luckily the position is drawish. 0-1
Interestingly, it would take more time
Karpov sacrificed the b6-pawn not
to win if Black had a knight on e6
only in this game but also in an earlier
instead of the pawn.
one where the circumstances were very
On the other hand in the next position
similar.
against Karpov from Tilburg 1991 I did
manage to salvage a draw with rook J .Saren - A.Karpov
versus two knights and a bishop. Skopje Olympiad 1972
People drew the conclusion from
my body language that I considered 1 e4 c5 2 ltJf3 e6 3 d4 cxd4 4 ltJxd4
my position lost at adjournment. My ltJc6 5 ltJb5 d6 6 c4 ltJf6 7 ltJ I c3 a6
reaction to this? No comment! 8 ltJa3 iLe7 9 iLe2 0-0 10 0-0 b6
34
11 ~e3 ~b7 12 f3 l:tb8 13 'iiel lbd7 13 ... lbxe4 14 ltJxe4 1i..xe4 15 'it'xd8
14 'ii'fl lbc5 15 llfd 1 f5 16 exf5 l:txf5 1i..xd8 16 :tad1 d5?!
17 lbc2 ~h4 18 g3 1i..e7 19 b4 lbd7 Maybe defending the pawn was
20 f4 better, but that would be slightly
passive. An interesting psychological
echo is that 16 years later I also gave a
free pawn to my opponent in the World
Championship fInal when neither
player had yet scored a victory.
17 f3 1i..f5 18 cxd5
At the time, commentators thought
18 g4?! ~g6 19 cxd5 exd5 20 l:txd5?
was a losing move, but after 20 ... :te8
21 l:tfd I! White is still a bit better.
20 ... 'iVf8 21 b5 axb5 22 cxb5 lba5
18 ... exd5 19 l:lxd5 1i..e6
23 ~xb6 lbxb6 24 'ii'xb6 1i..d8 25 1i'a7
If 19 ... l:te8 20 Wfl 1i..e6 21 l:td6.
l:tc8 26 'ii'e3 e5 27 1i..g4 lbc4 28 'ili'd3
~b6+ 29 ~f1 l:txf4+ 30 gxf4 'ii'xf4+
0~1
I tucked this idea away safely and

prepared a novelty for my fIrst World
Championship match. I employed it
when the score was 0:0. Let's see how
Karpov's idea worked against him.
Moscow 1984 20 :d6!? 1i..xa2?!
After 20 ... 1i..e7 21 ~xa6 l:txa6
1 e4 c5 2 lbf3 e6 3 d4 cxd4 4 lbxd4
(21 ... 1i..xa3? 22 bxa3 ltJc4 23 l:he6
lbc6 5 lbb5 d6 6 c4 lbf6 7 ltJ1c3 a6
8 ltJa3 1i..e7 9 1i..e2 0-0 10 0-0 b6 wins.) 22 1i..xa6 l:tb8 23 il.d4 lbc6
n1i..e3 1i..b7 12 'iib3 ltJa5 24 ~c31i..c5+ 25 Whllbb4 it would be
This was my prepared novelty - you hard to progress with White.
can guess where it came from. 211:ba6 l:tb8 22 1i..c5 l:te8 23 1i..b5!
13 'ii'xb6 l:le6
Other moves were no better.
See diagram on page 31. If 23 .. J:te5?! 24 1i..d6 l:texb5
25 lbxb5 lhb5 26 lla8 lbb7 27 1Lc7
Black's pawns have not advanced as wins. If 23 ... l:txb5 24 ltJxb5 il.c4
far as in the Garcia-Karpov game, but I 25 ttJd6!? or 25 l:tdl 1i..xb5 26 l:txd8
did not have to sacrifIce a pawn. 1:txd8 27 ':xa5.
35
24 b4 ttJb7 If 28 ....i.xc2 29 ':xe6 fxe6 30 :Xe7

After 24 .. .l:ha6 25 i.xa6 ttJb3 ttJd6 31 i.d7 ':xb4 32 i.c5 :bl+
26 :el!? 33 ~f2 wins as Taimanov pointed out.
25 i.f2 i.e7 26 ttJc2 i.d5 27 :dl 29 ':xe6 :xd7 30 :e1! :c7 31 .i.b6
i.b3 28 :d7! ':d8 1-0
A.Beliavsky - A.Karpov
Karpov won an opposite coloured

bishop endgame where he had a
strong light-squared bishop and a
rook and won despite being a pawn
down. "Furthermore, to make it even
more misleading, he beat a very fine
player - Beliavsky.
Remembering this particular game well, I twice opted for such positions
against Kramnik. In one of them (below, left) I was a pawn up, not down,
and my opponent had no passed a-pawn and I only drew.
But this was not all in the match. In the next example (below. right), I did
not have a strong bishop, but the similarity is still there as my opponent was
a pawn up and possessed a passed a-pawn.
Out of these two games I totalled a miserable half a point whereas
Karpov scored twice as much as that from a single game. To make matters
worse, this happened to me during the World Championship match against
Kramnik.
V.Kramnik - G.Kasparov V.Kramnik - G.Kasparov
36
A.Beliavsky - A.Karpov 28.i.xd7?

USSR Championship, Moscow 1973 Better was 2BlOxf7! ~xf7 (2B .. .lhf7
29 dxe6 lIe7 30 ':xc8 wins) 29 dxe6+
1 d4 lOf6 2 e4 e6 3 lOe3 .i.b4 4 e3 eS 'ii'xe6 30 .i.b3 lOd5 31 'ii'xe6+ ~xe6
S .i.d3 0-0 6 lOn dS 7 0-0 dxe4 32 l:lc6+!; 28 lOxd7 lOxd7 29 'ii'c3+
8 .i.xc4 cxd4 9 exd4 b6 cwt>h7 30 dxe6 fxe6 31 i.g3. In both
The Karpov line brought me my cases White's advantage is sufficient to
second loss in the Kramnik match, sail home with the full point.
from there on it was hard to stage a 28 ... lOxd7 29 dxe6 'ikxe6 30 lOc4
comeback. .i.a6 31 'ifxe6 fxe6
10 .i.gS .i.b7 11 'ii'e2 lObd7 12 :tac1
l:le8 13 lOeS 'ikc7 14 .i.bS 'ikd6
IS l:lfd1 .i.xe3 16 bxe3 'ifdS 17 f4
'ir'd6
32 i.g3?
Placing the bishop on a passive
square. 32 g3 was a better way to
simplify as the bishop is then far more
18 c4 active: 32 ... .:cB 33 :tdl i.xc4
White's centre is huge. Black should 34 Axd7+ ~f8 35 l:a7 i.d5 36 a4.
undennine that zone as the more space 32 ....:c8 33 lId1 i.xc4 34 Axd7+
White has the quicker he will suffocate <ltf6
his opponent. See diagram on page 36.
18... 'ii'e7 19 .i.a4 a6 20 .i.c2 g6 White has an extra pawn but no
21 'ike1 ~g7 22 .i.a4 h6 23 .i.h4 bS longer an advantage. But Beliavsky is a
At a cost of a pawn Karpov gets rid great fighter and still plays for a win.
of the nagging bind. 3S a3
24 cxbS 'ikd6 2S bxa6 .i.xa6 26 dS! After 35 l:d2 ':aB=.
l:lxel 27 l:lxc1 .i.c8 3S ....i.dS
37
36 h3? 45 •.•lIa8 46 ~f2 <iPg6 47 lId7 lIa3

In this case, placing the pawn on the 48 :'d8 :'0+ 49 ~gl
colour of his opponent's bishop means
it's going to fall in the long run.
White can force matters and move
closer to a draw by 36 .i.h4+ ~f5
37 :fl+ ~e4 (37 ... ~g4? 38 i.el and
White can play for a win again.) 38 :'f6
(38 h3 Now he makes room for the
king. 38 ... lIcl + 39 ~h2 <iPe3 40 :'f6
.i.e4 41 f5 gxf5 42 lIxe6 l::tc2 43 <iPhl
f4 44 l:1e8 and White holds.) 38 ... l::tcl+
39 ~f2 l:.c2+ 40 ~g3 (40 <iPgI ~e3) 49 ...:'xh3
40 ... !it>d4 41 ~g4 .i.e4 42 .i.g3. Black's Karpov once held Torre in a single
advantage is no more than symbolic. rook and opposite colour bishop ending
36 .. Jlc1 + 37 <iPf2 lIe2+ 38 ~e3 but there Torre had h- and g-pawns.
lIc3+ Karpov's position must have been lost,
Suddenly Karpov can start to squeeze. but here he wins even though it takes
39 ~f2 lIxa3 40 i.h4+ g5 41 bg5+ time.
hIg5 42 i.g3 lIa2+ 43 <iPe3 50 :'b8 lIe3 51 i.d6 :'e2 52 lI18 lIe6
53 .i.e5 g4 54 :'f6+ ~g5 55 lI18 ~f3
56 i.f4+ ~g6 57 ~f2 lIe2+
In the old days when there were
adjournments the players sealed at
move 56. Here Furman and the other
helpers found a way to win.
o 58 ~g3 lIg2+ 59 ~h4 :'e2 60 ~g3
e5
Karpov wins by pushing the e-pawn
without using his king.
43 ... lIxg2 61 lIb8 e4 62 lIb5 :'e3 63 lIb6+ ~fi
From being a pawn down, Karpov 64 ~g5 lId3 65 ~f5 e3 66 lId6 lIb3
reaches an endgame a pawn up! 67 lId7+ <iPe8 68 ~e6 e2 69 lIe7+ ~f8
44 .i.e7 l:a2 4511h7 70 ~f6 .i.d5 71 -*.h4 lIo+ 72 ~g6
45 i.d6 avoids the immediate loss .i.f7+ 0-1
of more material. 45 ... lIb2 (After
45 ... lIa4! 46 'it>f2 it is hard to tell if V.Kramnik - G.Kasparov
Black can win.) 46 ~d4 lIb3? (This Game 8, World Championship,
direct attempt to win fails. Black London 2000
can maintain his edge by 46 ...:'e2.)
47 .i.e5+ 'iPg6 48 lIg7+ ~h6 49 lIg8 1 d4 t;)f6 2 e4 e6 3 t;)e3 .i.b4 4 'it'c2
and White holds. 0-0 5 a3 .i.xc3+ 6 'iVxc3 b6 7 ~g5
38
.i.b7 8 f3 h6 9 .i.h4 dS 10 e3 lLlbd7 22 .. .lld2 23 ltJe3 .i.b7

11 cxdslLlxdS 12 .i.xd8lLlxc3 13 .i.h4 The g2-pawn is weak and the bishop
lLldS 14 .i.n cS 15 .i.bS lIfd8 16 e4 can target it, just as in the Karpov
game.
24 b4 lIfB 25 lIa2 lIxa2 26 tZ'lxa2
tZ'ldS 27 .i.d4 lIa8
16...lLlc7!!
At the time this was a very important
novelty. We had already played a few 28 ltJc3
games in which I accepted a weakness After 28 lin g5! (28 ... ltJxb4?
or a material deficit for better 29 lIg3) 29 ltJc3 ltJf4 Black has the
development.
upper hand. If 28 .i.b2 lbe3 29 :e I
17 .i.xd7 lIxd7 18 dxeS fS! 19 cxb6
lId8 (29 ... i.d5 30 lbc3 lbc4 31 :te2!
axb6 20 lLle2
Giving back the pawn at once with Probably Black has no win here,
20 lid 1!? would lead to an equal game. because he ean get the a- and the
For example: 20 ...lIxdl + 21 ~xdl fxe4 b-pawn in exchange for the e-pawn,
22 fxe4 .i.xe4 23 ltJD. reaching a rook ending a pawn up but
20 ... fxe4 21 fxe4 .i.xe4 which I evaluated as slightly better for
Black only.) 30 :c7 .i.e4! (30 ... lId2
31 lIxg7+ ~ffl 32 .i.e3 holds.) 31 ..tel
(31 lIxg7+? ~ffl 32 ..te3 ..tb 1 33 lIg3
lbf5 34 lin 1:Idl + 35 ~f2 h5 wins.)
31 ...:dl + 32 ~f2 lbg4+ (32 ... ltJd5?
33 lLlc3=) 33 ~g3 lbf6 34 i.e3 lId8!
Black keeps his winning prospects as I
pointed out in some analysis in
Informant.
220-0? 28 ... ltJxe3 29 ..txc3 lIxa3 30 i.d4 bS
After 22 lLlc3 i.xg2 23 lIg 1 i.D
24 lIg3 lId3 25 i.xb6 ltJd5 26 lLlxel5 See diagram on page 36.
exd5 we could say it's a balanced yet
fighting game. 31l1f4
39
V.Kramnik - G.Kasparov
Game 2, World Chess Championship
London 2000
1 d4 ttJf6 2 e4 g6 3 ttJe3 d5
Against Kramnik, I had quite a few
draws in the Griinfeld during the
second half of the 1990s. Some were
very close, maybe I should have sensed
31..Jld3 that sooner or later I would lose one.
I was a pawn up whereas Karpov was 4 eId5 ttJxd5 5 e4 ttJxe3 6 bxe3 iLg7
a pawn down. He won, so it made me 7 ttJf3 e5 8 iLe3 'it'a5 9 'it'd2 iLg4
too complacent and I missed a chance ... 10 lIbl
After 31...h5! 32 g4 h4 33 g5 lIa2
34 %.txh4 l::tg2+ 35 ~fl l:1xg5 the
situation is almost identical to the
Beliavsky-Karpov game. The extra
b-pawns must increase the stronger
side's chances.
32 l::tg4 g5 33 h4!
Now White holds.
33 ... ~fi 34 h:xg5 h:xg5 35 ~f2 l:td2+
36 c;i(e3
10 ... a6
Later I said I just gave up a pawn
here. I had already visualised the
ensuing opposite coloured bishops
position.
11 :xb7 iL:xf3 12 g:xf3 ttJe6 13 iLe4
0-0140-0 exd4 15 exd4 iLxd4 16 iLd5
.i.e3
Not 16 ... .i.xe3? 17 'it'xe3 lbc8
18 l::tc 1 ttJb4 19 l::txc8 lhc8 20 'it'd2
'it'd8 and Black must suffer in this
36 .. Jhg2
I win the g2-pawn as well, but sadly position a pawn down. If 20 ... e6?
it only leads to a draw since my passed 21 iLxe6 wins.
pawns are too close to one another. 17 'itel! ttJd4 18 iLxd4 .i.xd4
A gap of one rank is usually not I was not particularly unhappy
enough. here.
37 Ib:g2 iLIg2 38 iLe5 In-In 19 ::xe7 :a7 20 lba7 iLIa7
40
Anatoly Karpov the l]flr
35 l:lc4 White retains decent winning

chances.
31 0 l:le7
If 31...l:lb5 32 a4 l:lb2 33 h4!?
3284
21 f4! 'ifd8 22 'ifc3 ~b8

If 22 ...'ifh4 23 'ifg3.
23 "0 'iih4 24 eS gS 2S l:le1 "xf4
If 25 ... gxf4 26 e6 fxe6 27 l:txe6.
Later, in my second loss in this match,
I did not mind a battery when Kramnik
moved a rook to e6. 27 .....g5+ 28 'iig2 32 ..•l:la7
"xg2+ 29 ~xg2 rl;g7 30 l:lxa6. White After 32 ... .i.a7+!? 33 rl;g2 ~e3
is a tempo up compared with that game. Kramnik suggests several plans to
26 "xf4 gxf4 27 e6 fxe6 28 l:lxe6 convert the pawn advantage. Here are
~g7 two of them. 34 ~h3! (or 34 llc6!?
Even 28 ... a5 is possible, but it leads l:ta7 35 ~c2) 34 ... l:te5! (34 ... l:ta7?
to a miserable defence. 35 l:txa7+ ~xa7 36 rl;g4 ~e3 37 a5 h6
29 l:lxa6 38 ~f5 wins.) 35 ~g4 h5+ 36 ~h4
See diagram on page 36. ~f2+ 37 ~h3. It's difficult to tell
29 •..l:lfS which one was the best, but in all lines
Karpov still had to take'the g2-pawn, Black must suffer.
which did not even exist here. The a- 33 l:tb6 ~eS
pawn never assumed a role in that game Maybe I should have tried to defend
- and I was hoping for the same. I knew with 33 ... .i.c7 34 l:tb4 ~d6! 35 l:lc4
in a match Kramnik would not be l:ta5.
careless and lose as Beliavsky did, but 34 l:tb4 l:td7 35 ~g2 l:td2+ 36 ~b3
if Karpov won his game I should have hS 37 l:tb5 ~f6 38 as l:ta2 39 l:lb6+
at least made a draw.
After 29 ... l:lf6 30 l:la8 (30 l:la5!?)
30 ...l:lb6 31 a4 l:lb2 32 a5 ~d6 33 l:lc8
White is likely to win. Indeed 33 ...l:lb5
34 a6 l:lxd5 35 a7 f3 36 l:lg8+ ~f6
37 ~hl is decisive in a nice line shown
by Kramnik.
30 .i.e4 l:leS
Upon 30 ...l:lb5!? 31 ~g2 l:lb2 32 a4
.i.e5 33 l:lc6!? ~d4 34 .i.c2 l:la2
4-1
39 ... ~e7?? 45 1:d5 White loses the opposite

This loses a piece m one move. colour bishop ending as he drops a
However, the position is lost anyway. second pawn or else allows White's
Even after 39 ... ~g7 40 a6 .i.d4 king to invade on the queenside.
41 lIg6+ ~f8 42 .i.b7! is winning. 45 ... 1:xd5 (45 ...1:a2 46 1:xh5 .i.gl
47 ~g4) 46 .i.xd5 <it>g7 (46 ... .i.f2
Here are Kramnik's lines. 42 .. J:ta5
47 ~g2 .i.a7 48 h4) 47 <it>g2 h4
(42 ... .i.e3 43 l%g5 h4 44 1:g4 rj;e7
(47 ... <it>f6 48 h4) 48 ~h3 .i.f2 49 csttg4.
45 ltxh4 .i.gl 46 rj;g4 wins.) 43 1:d6
40 .i.d5 1-0
~gl 44 .J:[dl! .i.e3 (44 ... :g5 45 lhgl)
I should not have lost the title match
against Kramnik. My fIrst loss was
related to Karpov while in the second -
as you will see in the Alekhine chapter
- I followed the fourth world
champion. See page 215. Why do my
countrymen have such an adverse
effect on me - forcing me to lose my
title? Fortunately the crown at least
remained in Mother Russia.
The last kind of position that influenced me from Karpov's games had also
occurred twice in my match with Kramnik. Karpov's contribution to the loss
of my title was considerable.
The 12th world champion won a game where he had a sole extra c-pawn on
the queens ide and both sides had four pawns on the kingside. He also won
another game like this against Van Wely.
A.Karpov - P.Nikolic L. van Wely - A.Karpov
42
Having shown you the positions Quite incredibly the structure

Karpov went on to win, I'll show you occurred once again in the same
mine against Kramnik. I spoiled my match - only that there were
position and only drew - and ended additional a-pawns in each camp and
up not winning a single game in the there were no minor pieces on the
match. board.
V.Kramnik - G.Kasparov V.Kramnik - G.Kasparov
Karpov also had a game against A.Antunes - A.Karpov

Antunes which was very similar to
my second game against Kramnik
and I thought that this time I would
make it as Karpov won the position
despite not even being a pawn up (his
opponent had a doubled pawn), but
again I could not convert the material
advantage.
A.Karpov - P.Nikolic
Tilburg, 1988
1 d4 t[)f6 2 c4 e6 3 t[)f3 b6 4 g3
A quick look in the database shows
that Karpov has had this position with
White 100 times in regular and rapid
games and lost only twice.
4 .•..1a6 5 b3 .1b7 6 .1g2 .1b4+
7 .1d2 a5 8 0-0 0-0
43
Anatoly Karpov the J2111
9 'ii'e2
The start of a long manoeuvring
phase.
9 ... e5 10 ltdl ttJa6 11 ~e3 a4
12 ttJe3 axb3 13 axb3 d5?! 14 ttJa4!
Nicely applying pressure on the
queenside.
14 ... b6 15 'ifb2 l1e8 16 ttJe5 l:tb8
17 ttJd3! dxe4 18 .i.xb7ltxb7 19 bxe4
"ike7? This is a hacking sacrifice. Karpov
This loses a pawn. easily neutralises the ploy and wins.
29 ~xf2 e3+ 30 !ili>g2 lhe5 31 'iibl
':b5 32 h4 'iWb5 33 "ikd3 11f5 34 l:to
lixfi 35 ~xfi "ikg4 36 !ili>g2 1-0
Karpov had another game like this

and won it as well.
L. van Wely - A.Karpov

European Rapid 30 minute
Championship, Cap d' Agde, 1996
20 ttJaxe5! bxe5 21 l1xa6 ttJg4 1 d4 ttJf6 2 e4 e6 3 ttJO b6 4 g3 ~a6

22 Viet exd4 23 ~xd4 5 b3 .i.b4+ 6 .i.d2 ~e7 7 ~g2 e6
8 ~e3 d5 9 ttJe5 ttJfd7 10 ttJxd7 ttJxd7
See diagram on page 42. 11 ttJd2 0-0 12 0-0 ttJf6 13 l:.el e5
14 e4 exd4 15 ~xd4 dxe4 16ltJxe4
White wins a pawn and steadily
converts his advantage.
23 ... e5
Predrag looks for counterplay.
Waiting passively with 23 ... l1c8!? may
have prolonged the game.
24 .i.al e4 25 ttJxb411xb4 26 e5 ':b5
27 l:tad6 'iVg5
Trying to fish in muddy waters.
27 ... e3 28 f3 ttJf2 29 l:tld5 was also 16... ~xe4
hopeless. Karpov creates a weakness.
28 b3! ttJxf2 17 bxe4 lte8 18 lIc1 ~e5 19 ~xe5
44
':xe5 20 'iVa4? 'iVe7 21 e5 tDd7 22 f4 12 ...'ilt'xe3

b5! I was two points behind and had to
Winning a pawn. try to win, even with Black.
23 it' a3 bxe4 13 i.a3 tDd5! 14 1!ibl 'iVf6 15 .td3
See diagram on page 42. h6
If 15 ... 'ifh6?! 16 b5.
24 1:edl tDb6 25 1:e3 g6 26 1:d6
16 b5 l:td8 17 i.b2 'ike7 18 ':a4?1
Waiting passively did not offer much
hope either. This is both imaginative and risky.
26 ...1:e8 27 h4 tDd5 28 l:.e2 1:a5 If 18 i.a3! 'ii'f6. The alternative
29 it'fJ 'ii'e5+ 30 <iPh2 tDe3 31 l:ted2 18 ... tDc5?! is a risky way to play for a
win. 19 i.h7+ <ifi>h8 20 tDe5. Now
19 i.b2=.
18 ... tDe5 19 i.h7+
Not 19 i.a3? tDc3.
19 ... <iPh8
31. .. e3
The c-pawn makes its decisive
advance.
32 l:.d8+ <iPg7 33 1:xe8 'ii'xe8 34 l:e2
tDxe20-1
20 l:.h4
Kramnik moves against the king.
Kramnik could play on the c-file with
London 2000 20 :c4 but 20 ... i.d7 followed by .te8
1 d4 tDf6 2 e4 e6 3 tDe3 i.b4 4 e3 keeps Black in the game. Going after
0-0 5 i.d3 d5 6 tDfJ e5 7 0-0 dxe4 the king on the g-file with 20 l:g4!?
8 i.xe4 tDbd7 9 a3 exd4 10 axb4 dxe3 might have led to some very exciting
11 bxe3 'ike7 12 i.e2!? tactics. 20 ... e5! 21 l:.g3
45
21...e4!! (21...liJf6 22 liJxe5! 27~2

[22 .i.xeS liJh5] 22 ... l::td5 23 f4 liJxh7 If 27 'iib7 'ii'd6; or 27 liJh4 'ii'd6!
24 liJc6 'ii'f8 25 .i.xg7+ and Kramnik 27 ...l:.b5 28 1i'a3 liJb6
prefers White in his Informant Karpov exchanged this at once. I did
analysis.) 22 liJeS! (22 liJd2 liJa4!!) not have the opportunity to follow suit,
22 ... liJf6 (22 .. .r~xh7? 23liJc6) 23 ~g6!
so I had to think about alternatives .
.i.e6 24 f3! with an unclear position.
28 ... liJb4!? was promising. Placing the
20 ... f6
bishop on the long diagonal, just as
Not 20 ... liJf6? 21 ~xf6 'iVxf6 22:f4.
Karpov had done, looks reasonable.
21 l:.e4!
Vladimir wisely returns to the Then 29 .i.bl (29 1i'a7 ji'd6) 29 ... .i.c6
queenside. 30 e4, intending ji'e3.
21. .. ~d7 22 ~a3?!
A trip to the edge of the board with
22 liJh4! would have kept the position
balanced. White has enough play for
the pawn.
22 ... b6 23 .i.e4
If 23 l:Uc 1 .i.xb5!
23 ... a6!
After 23 ... l1ac8 24 liJd4! ~e8
25 l:[fcl it is a matter of personal taste
which colour you prefer.
24 bxa6? 29 l:4e3
This is a serious mistake as it allows Not 29 :xc5? when ~c6!! wins; nor
the rook to come into play. After 2911d4!? 1:c8.
24'liJh4 ~e8; or 241Hcl axb5 25 .i.xc5 29 ...l:.b4! 30 liJd2
bxc5 26 lhc5 and White has just Nor 30 .i.b I?? when ... l:ta8 wins.
enough to hold.
And if 30 "a5 liJa4.
24 ....:.xa6 25 .i.xe5
30 ... f5?
Here 25 ~xd5 ':'xa3! 26 liJh4
This is too hasty. Also 30 ... liJa4?!
(26 ji'xb6? liJa4) 26 ... .i.e8 27 ~c6 and
it is difficult to tell whether or not there allows 31 llxc5. However Kramnik
is enough in the position to win. points out the beautiful 30 ... ~c6!?,
25 ... bxe5 26 %:tfe1 :a5! ( ... l:tc8 is a good move as well)
31 1:xc5 (31 ~xc6 l::txd2) 31.. .lhe4
See diagram on page 43. 32 'iVaS ':xd2 33 'ii'xd2 ~d7 with
excellent winning chances.
Black is a safe pawn up and as
Karpov had won so convincingly with I beat Vladimir from a very similar
the same structure, perhaps I became a position in the 2001 Botvinnik
little too casual. Memorial, my first rapidplay match.
46
V.Kramnik - G.Kasparov 31. .. ltJa4?

Botvinnik Memorial rapidplay But Black blunders the pawn back,
Moscow 2001 after which his dream of winning is
gone. If 31 ... ~c6? 32 :txc5=. Best was
31 ...:c8 which keeps the extra pawn
and retains some chances of winning.
Kramnik mentions this move, but also
31...c4 looks promising.
33 ':a8+ ltJf8 34 'iic2 g6 35 g3 ~g7

36 ~g2 e5 37 :a4 'ifd6 38 'ifc4 ltJe6
39 'ifd5 'iib8 40 :ta8 'ifb2 41 .tla2 'iVc3
42 itd2 'iVb3 43 'ifd5 'iib8 44 l:.a8 'fIc7
45 l:la6 ~d4 46 'ili'c6 'ife7 47 :ta8 'iff6
48 'ili'c2 ltJg5 49 :a3 'ife6 50 h4 'iih3+ 32 l:.xe5! :tb2
51 ~gl ltJe6 52 l:tb3 'iVg4 53 'ifd3ltJc5 If 32 ... l:.bb8 33 :t5c3.
54 'iff3 'ili'xf3 0-1. 33 ltJe4 'iVxe5 112-112
All the queenside pawns are gone
Back to the game and there is nothing left to play for. If
33 ...:xf2? 34 l';tc8.
A.Antunes - A.Karpov
Tilburg 1994
1 ltJf3 ltJf6 2 e4 e6 3 g3 d5 4 ~g2

~e7 5 0-0 0-0 6 d4 dxe4 7 ltJe5 ltJe6
8 ltJxe6 bxe6 9 ltJa3 ~xa3
31 ~f3?
Having little time left White misses
the drawish simplification 31 :xc5! .
Then 31...':xe4 (31...':b5 32 ~f3 ~e8
33 ltJb 1!; 3 1.. .l:.a4 32 'fIb2) 32 ltJxe4
fxe4 33 'ii'b4! and White wins the
e4"'pawn, which is enough to hold on.
33 ... ltJa4 (33 ... 'iVd6 34 'ii'xe4=)
34 l:.5c4 'ifxb4 35 l:txb4.
47
10 bxa3 28 .. .l:tc6?
This somewhat awkward pawn Though this wms this game, he
structure often occurs in the Catalan misses a clearer path to victory by
opening. 28 ... .:bl!! 29 l%xbl (29 'iic2 ':xfl+
10 ....i.a6 11 iLg5?! 30 ~xfl ~1+ 31 ~e2 'iVg2 wins.)
Taking the pawn is the main line and 29 .....xf2+ 30 <it'hl 'iVxg3.
is more natural. 29 J:td4 'ii'e2 30 ':d2 'ii'f3 31 l:ld4
11 ... b6 12 Jtxf6 'ii'xf6 13 .i.xc6 'iie2 32 l:ld2 'iib5
l:Iab8 14 ..wa4 l:Ib6 15 lIfdl J:td8
16 .i.f3 c6 17 ~g2 ..we7 18 e3 l:Ic8
19 b4 iLb5 20 ~4 c5 21 dxc5
33 l::tbl l:td5
Black's advantage has by no means
evaporated.
34 'fic2+ iVg6 35 'iixg6+ ~xg6
21 ...11xc5 36 l:Ic2 l:1d3 37 a4 l:Ia3 38 ':b4 c3
Karpov creates a passed c-pawn. 39 ~n 11a6
22 11d8+ ~h7 23 11adl .i.c6 24 ..wc3
Jtxf3+ 25 ~xf3
25 .. J:tf5+
If 25 ... 'irb7+!? 26 e4 l:lh2.
26 <it>g2 ~7+ 27 <it>gl 'iif3 28 l%n
401:tb3
40 l:tg4+!! would he an interesting
check that gives better chances as it
drives the king further away from the
centre. 40 ... ~h7! (40 ... ~f6 41 l:If4+
~e7 42 ':c4l:16xa4 43 l%2xc3l::txa2 [on
43 .. .l:hc3 44 lIxa4! and White holds]
48
Anatoly Karpov the J 2 rh
44 I:tc7+ ~f6 45 1:.b7 Black has no time 53 ...~c7 54 1:a3 <it'b6 55 l:ta8 l:tc3+
to defend f7 therefore White gets away 56 litd4 l:tc4+ 57 ~e3 I:tc3+ 58 <itd4
with it.) 41 %:tc4 (41 ~e2 1:t3xa4 %:tc4+ 59 <i!fe3 I:tc7
42 ':xa4 l:txa4 43 ~d3 l:ta3)
41...1:6xa4 42 1I2xc3 l:txa2 43 ltc7
~g6 44 h5+ litxh5 45 %:txf7 and White
probably holds.
40 ...%:t3xa4 41 l:tcxc3 I:txa2 42 lIb7
l:tb6
60 l:tg8?!
The subtle intermediate check
60 l:tb8+!! would at least force Karpov
to fight hard for the point. 60 ... <i!fa5
(after 60 .. J:tb7 61 .:tg8 b4 62 <&td2l:tc7
63 g5 b3 64 l:Ib8+ l:tb7 65 ~g8 I think
43 1:xb6 axb6 44 l:tb3 l:ta6 45 e4 White draws - by the way there is no
~f6 46 f4 9;;e7 47 ~e2 ~d6 48 g4 beautiful win by 65 l:hb7+ <iftxb7 66 f5
:a2+ 49 <ifi'e3 ~c6 50 l:tc3+ litb7 exf5 67 g6 as after ... fxg6 Black's king
51 I:td3 l:tc2 52 h5 b5 is close enough to stop the e-pawn.)
61 :g8 b4 62 IIxg7 b3 63 I:tg8 l:tb7
64 ~d2 and White can hold as the
b-pawn can't be pushed because the
rook check on a8 saves White.
60 ... b4 61 ~d3?!
With 61 <i!fd2! White has more
chances of getting behind the b-pawn.
61...<ita7! (Karpov should play
differently from the game. After
61...b3? 62 IIb8+ IIb7 63 ':'c8 ~a6
53 e5? 64 <itc 1 White has decent drawing
53 %:td7+ would have produced a chances.) 62 I:td8 b3 63 %:td3 b2 64 I:tb3
better fight. 53 ... %:tc7 54 l:td8 9;;b6 :b7 65 lhb7+ litxb7 66 ~c2 ~c6 and
55 g5 b4 56 lIg8 ~b7 57 l:td8 and Black wins.
White still has chances of holding the 61. .. b3 62 1:.b8+ %:tb7 63 l:tc8 ~a7
game (57 l:txg7?? b3 !). 64.:tel
49
64 ...:bS! Black has at least equalised by

Karpov's subtle play has provided his isolating White's c-pawn.
king with a path to invade the kingside. 19 ... hS 20 'iWf4 'iWc6 21 .i.xf6! ltJxf6
6S lIbl ~b6 66 ~c3 ~cS 67 ~b2 22ltJxf6+
l:tb4 68 l:to ~dS 69 l:to ~e4 70 :0 After 22 'iWe3 ltJxd5 23 cxd5 'ii'c5.
l:tb7 71 nf2 ~e3 72 no ~e2 73 l:.gl 22 ....i.xf6 23 ltJdS .i.b2 24 :bl .i.g7
~f2 0-1
2S 'it'gS ~f8 26 l:tdc1 e6 27 tiJf6l:1ed8
28 h4!
London 2000
1 ltJo ltJf6 2 c4 b6 3 g3 cS 4 il.g2

il.b7 5 0-0 g6
Once I drew a very important game
against Karpov with this variation.
This was in the 23 rd game of our third
match when I needed a draw to hold on 28 ..:.a8!
to my title - and I achieved my If 28 ... 'iIi'c5 29 :b7!
objective with this particular English 29 cS!
opening. Kramnik prefers to sacrifice the
6 ltJc3 il.g7 7 d4 cxd4 8 'it'xd4 d6 c-pawn rather than let my rook get to
9 ndl ltJbd7 10 .i.e3 lIc8 11 l:tacl 0-0 c5.
12 1Wh4 a6 13 ltJel?! .i.xg2 14 ltJxg2 29 .. JbcS!
l:te8! IS b3 If 29 ... dxc5 30 nxc5 .ixf6
31 l:1xc8=.
If 15 .i.g5 l:tc5! to prepare 16 ... h6.
30 :xcS Jtxf6 31 'iVxf6 dxeS
Then after 16 b4 lIe5! 17 f4 l:e6 Black
is okay. See diagram on page 43.
IS .. :it'c7 16 JtgS ~7 17 ltJe3 bS Almost the same pawn structure has
18 ltJedS bxc4 19 bxc4 occurred in the match again. This time
50
Anatoly Karpov the J21h
I decided not to give the pawn back. Not 34 ... 'ii'c8? 35 l:tc6=
32 ~h2! ~g8 35 exf3 l:tc8 36 l:ha6 c4 37 %:td6
33 ':b6! 37 ... c3?

This is a nasty move, White now Pushing the pawn too far. 37 ... ~f8!
tlueatens simultaneously to take on a6
would have put much more of an
and e6.
obstacle in White's path. White has
33 ...l:te8
chances to survive but he must play
a) If 33 ... c4 34 ':xe6 ':dl 35 ':c6
holds. very carefully. I show Kramnik's line.
b) Alternatively 33 ... l:tb8 34 'ifn! 38 'it>g2 ~e7 39 l:tdl l:ta8 (39 ... e5!?
'ika7 35 :c6 'ikb7 36 a4 l:tc8 37 ':xc5! 40 a4!? %:ta8! 41 l:al 'Wtd6 42 a5
as Kramnik pointed out. [42 ~fl ~c5 43 ~e2 ~b4 44 l:tb 1+
c) If Black tries to go after White's ~c3 wins] 42 ... l:ta6!? [42 ... ~c5 43 a6
king with 33 .. Jldl there is a draw. ~b6 44 l:tb 1+ 'Wtxa6 45 ~f1 ]
Kramnik shows some remarkable lines. 43 'Wtfl ~c5 44 '1!i>e2 ~b4 and White
34 e4 \Wc8 35 'ife5! l:[d8 36 ~g2!? c4
has a really difficult task.) 40 l:td4
37 g4! hxg4
(40 ':cl l:ta4) 40 ... c3 (40 ... l:ta4 41 f4)
el) 37 ... c3 38 gxh5= c2? 39 h6 wins.
~1 ':c4 (41 a4 %:tc8 42 %:tdl c2 43 %:tel
c2) 37 ... ~h7 38 gxh5 (38 \Wf6 'it'd7)
38 ....:g8 39 hxg6+ :xg6+ 40 cwth3; ~d6 44 ~f1 ~d5 45 'Wte2 l:tc3!)
38 h5 gxh5 (38 ... ~h7?! 39 'ii'f6) 41..J:ta3 42 ~fl ~d6 43 'Wte2 (43 %:tc8
39 'fig5+ (39 l:[b7? f6!) 39 ... ~f8 ~d5 44 ~e2 e5! [44 ... ~d4 45 %:td8+]
40 l:txa6! and Black's king is too 45 ~e3 f5) 43 .. Jlxa2+ (43 ... crPd5
exposed to do anything. 44 l:tc7 ~d4 45 ':d7+) 44 ~e3 l::ta3 45
34 'iif3? f4 and maybe this position can be held.
After 34 "'e5? \Wd5. But 34 "'c3!,
38 ':dl %1a8
blocking the c pawn earlier, was
After 38 ... ~g7 39 ~g2 'Wtf6 40 f4!
stronger. 34 .. J:tc8 (34 ...1i'e4 35 'ii'd3)
35 'iic4 'iid5 (35 ... l:tc6 36 'it'e4; 35 ... a5 Iifilf5 41 ~n c2 (41..J:ta8 42 :tel l:ta3
36 ':xe6!) 36 'ifxa6! l:ta8 37 'ifb5 c4 43 l:tc2) 42 ':c 1 %:tc3+ 43 ~e2 ~e4
38 a4 and White gets away with it. 44 <it>d.2 l:tn 45 <i!te2=
34 ......xf3 39 l:tel
51
43 l%b7 <it>e8 44 1.'tb8+ <it>e7 45 :b7+

<it>f6 46 <it>n
Better was 46 f4!
46 ... e5 47 l:tb6+ 'iPf5 48 :b7 'it>e6
49 ltb6+ <it>f5 50 :b7 f6
39 .. Jba2
I had to allow the proud c-pawn to
fall after all. From here on the position
is drawish. After 39 ... l:ta3 40 <it>g2 <it>g7
41 f4 ~f6 42 l:Ic2 e5 43 ~f3=.
40 lbc3
If 40 <;itg2? fla3 41 <it>f1 ~g7 42 'it>e2 51 l%g7
Unnecessarily providing Black with
<;itf6 43 l:tc2 ~f5 44 ~d3 e5 45 lhc3
another chance. 51 l:tf7! was called for.
(45 ~e3 ':b3 46 ~d3 flb2 wins)
51. .. g5
45 ... ':xc3+ 46 <it>xc3 e4 47 fxe4+ <it>xe4
48 <;itd2 ~f3 49 <;itel f5 wins according If 51...11d2 52 flg8.
52 hxg5 fxg5 53 l:tg8 g4
to Illescas.
After 53 ... h4 54 flf8+ <;itg6 55 ttg8+
40 .. Jbf2+ 41 'it>gl lla2
~f6 56 gxh4 gxh4 57 %:tg4 h3 58 ~gl
h2+ 59 ~h 1 flf2 60 IH4+ White holds
with the help of the stalemate motif.
54 l:tf8+ ~e6 55 tte8+
42 l:tc7
42 f4! looks dodgy because of the
isolated king, however Black still can't
win. 42 ... ~g7 43 ':c5 'it>f6 44 ~f1 1.'td2
45 ':a5 l:td5 46 fla7 ~f5 47 flxf7+ 55 ••• <;itf5
~g4 48 flf6 'iitf3 49 'iitgl=. This was just not my World
42 ... 'iitf8 Championship match. With 55 ... ~f6!?
If 42 ... e5 43 l:tc5 f6 44 ':c7 ~f8 I could still have created problems.
45 'iitf1 fld2 46 :la7 ':d8 47 ~e2 ne8 However White can save the game with
48 ~e3 l:te7 49 na4 'iitf7 50 f4. precise play. 56 llg8! gxf3 57 :f8+
52
cwtg5 (57 ... cwte6 58 l:Ixt) cwtd5 [58 ... e4 This was my last game with the
59 1:.b3] 59 1:.f5=) 58 %;txt) cJa>g4 Black pieces as the reigning world
(58 ... e4 59 :f4) 59 l:[e3; or 55 ... cJa>d5!? champion.
56 l:[d8+ (56 f4 e4) 56 ... <ili'c4 (56 ... ~c5
57 fxg4 hxg4 58 %;tg8 l:Ia4 59 ~e2=) Tal said once that Karpov was the
57 fxg4 hxg4 58 l:[g8 ~d3 59 ':xg4 and honoured trainer of Azerbaijan. Yes,
though the position is equal, White still Tal has a point as I improved during my
has to be careful. matches with Karpov. On the other
56 ':f8+ !itg6? 57 l:Ig8+ cwtf5 hand you can see I lost games because
This allows a threefold repetition. of him. Had I won these two games the
After 57 .. .rJ-if7 58 l:[g5 <it>f6 59 f4 exf4 aggregate score in our five World
60 gxf4 :h2 61 ~g 1 l:[h3 62 <it>g2 <ite6 Championship matches would not have
63 <1tf2=. been 21 wins 19 losses for me, but 23
58 lIf8+ liz-liz wins and 17 losses in my favour.
53
Robert James Fischer the 11th
Fischer won the title at the end of the better chance to defeat me in our
1969-1972 cycle. On his way to the matches.
final he beat Taimanov and Larsen 6-0 Fischer's influence included bringing
and Petrosian 6 1h-2Ih. In the world
more money into the game which also
title match he dethroned Spassky
was beneficial for me. However as you
12lh-81J1 which ended a 24 year-long
will see from the following examples I
Soviet dominance of the World
may have had even better results if I
Championship, which began in 1948.
had not followed his games so closely.
It was only natural that I should have
Let's see one of his games which is
investigated Fischer's games deeply,
well known and a focal point for me
and the effect can be seen in my
later when I played against Karpov.
repertoire with the Black pieces,
especially as we both played the Fischer had a strong e-pawn, Black had
Najdorf most of the time. Bobby also a good queenside. Below you see his
had an opening repertoire which was position first and then my own:
ahead of his time.
R.Fischer - L.Stein
Intriguingly, he actually helped me Sousse Interzonal 1967
indirectly by not playing against
Karpov - whose name in English 1 e4 e5 2 ltJf3 ltJc6 3 i..b5 a6 4 i..a4
means 'carp' and this darting fish ltJf6 5 0-0 iJ..e7 6 Ilel b5 7 iJ..b3 d6
eluded the reluctant fishennan! - as he 8 c3 0-0 9 h3 i..b7 10 d4 ltJa5 11 i..c2
would have strengthened Anatoly with ltJc4 12 b3 ltJb6 13 ltJbd2 liJbd7 14 b4
the experience of additional match~ White intends to transfer his knight
play. Karpov would have had an even to a5.
A typically occurring situation in the main Ruy Lopez: White has a strong
e-pawn, kingside majority and central advantage - and Black a good queenside.
R.Fischer - L.Stein G.Kasparov - A.Karpov
54
Robert James Fischer the 11 th
14 ... exd4 15 cxd4 as Then 21 ... .txD! 22 ttJxD c4 23 ttJd4

Black starts to undermine the centre. i.c5 24 i.b2 'ii'b6 and it is not easy to
16bxaS cS 17 eS attack with White.
Fischer pins his hopes· on the idea 21. .. lOb6?
that the e5-pawn provides space in the With his last three moves Stein has
centre and opens the way for White's strengthened the queens ide but
pieces to irritate Black's king. In neglected to take precautions on the
my Predecessors book 17 .i.b2 was other flank. 21...lle8 followed by ltJ:ffi
recommended. ensures a playable position.
17... dxeS 18 dxeS
22 lOfgS!
In one of my later games I also
Fischer starts operations against
adopted the idea of having an e-pawn
Black's king.
like this.
22 ... ~xe4
18 .•. lOdS
If 22 ... h6 23 lOh7 wins.
See diagram on page 54. 23 'ifxe4 g6 24 'ifb4!
19 lOe4 No more preparatory work is needed
White has to act quickly since if and Fischer forces matters with direct
Black gets to an endgame his chances threats.
are encouraging as he possesses a pawn 24 ... bS 25 'iVg3! lbc4
majority on the queenside. Not 25 ...... d4? 26 ltJxfl llxfl
19 ...lOb4 27 ~xg6 when White has a devastating
After 19 .. .l:ha5 20 ttJeg5 h6 21 'ifd3 attack.
g6, I showed that the tactical shot 26lbf3?
22 lOxfl wins. (Fischer's 22 lOe6 Fischer's move is slow. He could
is not so convincing because of land a potentially lethal harpoon by
22 ... lOb4!) 22 .. .l:hfl 23 'ii'xg6+ l:tg7 26 e6!? f5 27lbfl l:lxfl 28 exfl+ ~xfl
24 'ii'e6+ <ith8 25 'ii'xh6+ ~g8 26 e6 29 ~xf5 gxf5 30 'ifD ~g6 31 g4
ltJ:ffi 27 l:te5 and White has a winning generating a strong attack. Even more
attack. And if 19 ... c4 20 .i.g5! deadly however is the sacrifice
20 ~bl l:bas 21 'ife2 26 lbxfl! which leads to a win after
Keres' suggestion was 21 ttJeg5. 26 ... l:xfl 27 i.xg6 l:tg7 28 ~h6 "'f8
55
29 iJ..xg7 "'xg7 30 e6 h4 31 'iib8+ ~fB

32 .i.f7+ ~g7 33 'fif4.
26 ... ~g7
Upon 26 ... ttJd3 comes 27 l:tdl ttJxcl
28 l1xc l.
27 'ii'f4 l:th8
35 a4!! ttJb2 36 l:tbl liJxa4 37 'iVe5

'ii'e8 38 e7 White has a winning attack.
30.i.e4?
Sacrificing the knight with 30 ttJh4!
would have finished off Black.
30 ... iJ..xh4 31 'it'xh4 'it'xf5 (31 ... 'iif6
32 'fig3 l1e8 33 .i.bl l1a7 34 .i.g5
28 e6! 'ii'xa 1 35 iJ..h6+ wins so nicely.)
White has to hurry with his 32 'ii'e7+ 33 'ii'd8+ ~g7 34 'ii'c7+ ~g8
operations against the black king. 35 e7 wins as Fischer pointed out.
28 ... f5 30 .. :li'xf4 31 .i.xf4 ':e8?
After 28 ... iJ..f6 29 exf7 iJ..xal This time another queens ide move
30 fB='fi+ 'iixfB 31 'iic7+ ~g8 was required. Better was 31.. .l:txa2!
32 .i.xg6 ttJd5 33 'iib7 ttJf6 34 l1e6 32l::txa2 ttJxa2 33 ttJe5 g5 34 iJ..g3 ttJb4
l:ta8 35 'ii'xb5 ttJd6 36 'lib I .i.d4 35 ttJxc4 bxc4 36 .i.e5+ iJ..f6 37 .i.d6
37 ttJxd4 cxd4 38 'iWb3l:lb8 39 'ii'g3 h4 ':e8 38 iJ..xc5 ttJd3 39 .i.xd3 cxd3
40 'ii' g5 l::tb5 41 iJ..f7+ wins as I pointed and Black can simplify to a tenable
out in the Predecessors book. endgame.
29.i.xf5! 32 ':adl 1:[a6
White demolishes the pawn chain Regaining the extra pawn with
around the black king. 32 ... .:xa2 does not solve Black's
29 ...1r'f8 problem. After 33 l:td7 l::ta6 34 .i.bl
Stein settles for an endgame a pawn ttJc6 35 iJ..g5 White is better.
down which, however, still requires
special skill to win. If 29 ... 'iib8
30 ttJe5! (Hubner). Alternatively
29 ... gxf5 30 'ii'g3+ (30' 'it'xf5 'it'e8
31 ttJe5 ttJxe5 32 'fixe5+ iJ..f6 33 'ii'g3+
~h7 34 'fi'c7+ 1r'e7 35 1r'xa5 iJ..xal
36 ~xb5 looks better for White.)
30 ... ~h7 31 ttJg5+ .txg5 32 .txg5
l1a3! 33 'fif4 'ii'fB 34 l1adl l:ld3
(34 ... ttJd3 35 'fig3)
56
33 l::td7! lhe6 34 ltJg5 ':f6 35 ~f3! In the next game I constructed very
The most artistic way to win was much the same Spanish centre as
with 35 a3. Then 35 ... ltJxa3 36 ~e5 Fischer did against Stein. But, to put it
ltJc4 37 ~a1! ltJb6 38 ':b7 ltJc8 mildly, I should not have done this.
39 ~b1!
35 .. Jbf436ltJe6+ G.Kasparov - A.Karpov
Moscow 1985
1 e4 e5 2 lLlf3 lDc6 3 ~b5 a6 4 iLa4

ltJf6 5 0-0 i.e7 6 1:tel b5 7 St.b3 d6
8 c3 0-0 9 h3 St.b7 10 d41:te8 lllLlbd2
~f8 12 a4 'ilVd7 13 axb5 axb5 14 ':xa8
~xa8 15 d5 lDa5 16 St.a2 c6 17 b4
4Jb7 18 c4 ':e8 19 dxc6
After 19 'iVe2 lLld8 20 ~b2 bxc4
36 ... ~f6 21 ltJxc4 1i'a7! 22 .:tal cxd5 23 exd5
With an exchange sacrifice Black ~xd5 24 lLlxd6 i.xd6 25 St.xd5 'ilVxa I +
stabilises his position and prolongs the Black has compensation.
game considerably. To convert White's 19 ...'ilhe6
advantage requires the technique that
Fischer possesses.
37 ltJxf4 ltJe5 38 l::tb7 ~d6 39 <it>f1
ltJc2 40 l::te4 ltJd4 41 l:tb6 l::td8
42 ltJd5+ <itf5
20 c5?!
Just as in the Fischer game the
developing move 20 iLb2 was stronger.
20 ... 4Jd8 21 iLb2 dxe5 22 bxe5
'iVxe5 23 iLxe5
43 ltJe3+
Fischer sealed this move.
43 ... <it>e6 44 ~e2 ~d7 45 ~xb5+ When I captured the pawn I was
lLlxb5 46 :'xb5 ..ti>c6 47 a4 ~c7 inspired by the Fischer game. Here, just
48 ~e2 g5 49 g3 ':a8 50 1:tb2 ':f8 like the American champion, [ had an
51 f4 gxf4 52 gxf4 lLln 53 :e6+ ltJd6 extra pawn in the centre while Black
54 f5 :a8 55 l:[d2 lba4 56 f6 1-0 had an extra pawn on the queenside. I
57
was hoping to create an attack on birds with one stone. But the catapult is
Karpov's king and even dared to think pointing backwards. The queen should
that I would conduct the attack without have gone to the diagonal with 25 'ifb 1
any mistakes. Maybe this distracted me as happens later in the game.
from the reality of the game. 2S ... ltJeS!
23 ... ltJd7 24 ~b2 'ifb4!
This intermediate move is
undoubtedly strong as it allows White
no time to build up an attack on the
king. However I was still relaxed
because another famous Fischer game
sprung to mind - one which was very
similar to the present one. Here it is:
R.Fiseher - B.Spassky
Reykjavik 1972 26 ~al
I gave up material in order to ensure
play against the black king. It was
possible to hang on to the pawn by
26 ltJxc5 and then defend passively -
but I did not like to do that as after
26 ... iVxb2 27 ':e2 'ii'c3 28 ltJd3 Black
stands rather better with his two
bishops and distant passed pawn.
26 ... ~xe4 27 ltJfd4 tiJdb7 28 'iVe2
ltJd6
26 oltb3 axb5 27 _f4 l:'ld7 28 ltJe5 Karpov should have pinned my
'fie7 29 l:lbd 1 l:'le7 30 ~xf7+ l:'lxf7 queen to the defence of the a2-bishop
31 'iVxfl+ ~xfl 32 ltJxfl ~xe4 by 28 ...:a8!? His extra pawn would
33 l'lxe4 'it'xfl 34 :d7+ 'it'f6 35 l:'lb7 probably then be decisive.
l:tal + 36 ~h2 oltd6+ 37 g3 b4 38 'it'g2 29 ltJxeS 'iVxeS
h5 39 l:tb6 l:tdl 40 ~f3 'it'fl 41 'it'e2
lId5 42 f4 g6 43 g4 hxg4 44 hxg4 g5
45 f5 olte5 46 l:tb5 ~f6 47 l:texb4 oltd4
48 l:'lb6+ 'it'e5 49 'it'f3 l:td8 50 :b8 l:ld7
51 l:t4b7 ':d6 52l:tb6l:td7 53 l:tg6 <ittd5
54 l1xg5 ~e5 55 f6 ~d4 56 l:tb I }-o
2S tiJb3?!
By defending one piece and attacking
another, you might think this kills two
58
30 'ii'g4 Karpov likes to restrict his opponents

Just like Fischer I create some play and often does so with a pin.
against the king. However, in my case 33 'ii'cl .te4 34 l:el 'ii'a5 35 .ib3
this proves to be insufficient. 'ii'a8!? 36 'itb2 b4
30 ...:te8 31 :dl Karpov takes no risks and declines
If 31 'ili'f4 .id5! the pawn on g2.
31. ...ig6! 37 :e3 .ig6 38 l:he8 'iVxe8 39 'iVc1
White no longer has any realistic ltJe4 40 .id5 ltJe5 41 ltJb3 ltJd3 0-1
attacking chances against the king. White has nothing for the pawn,
32 'ii'f4 1fb4 which is why I resigned.
Let's continue with another even more famous Fischer game - or should I
say endgame. This example of domination by the bishop in endgames is often
taught to young players. Here the opponent's pawns are fixed and the pawns
are positioned on both wings. My game which follows has similarities.
k.Fiseher - M.Taimanov G.Kasparov - N. de Firmian
R.Fiseher - M.Taimanov 33 a4 tOe7 34 h3 ltJc6 35 h4 h5

Game 4, Candidates match, 36 ltd3+ ~e7 37 l::td5 f5 38 l::td2 1:f6
Vancouver 1971 39 l:te2 <it'd7 40 l:te3 g6 41 .tb5 %td6
42 ~e2 ~d8
1 e4 e5 2 tOf3 tOe6 3 d4 cxd4
4 tOxd4 'ike7 5 tOe3 e6 6 g3 a6 7 .ig2
tOf6 8 0-0 ltJxd4 9 'ii'xd4 .ie5 10 .tf4
d6 11 'i"d2 h6 12 :adl e5 13 i.e3
.ig4 14 .ixe5 dxe5 15 f3 ~e6 16 f4
:d8 17itJd5 .ixd5 18 exd5 e4 19 l:tfe1
l:bd5 20 :xe4+ <at>d8 21 'ii'e2 %:txdl +
22 'ifxdl + 'ii'd7 23 'ifxd7+ ~xd7
24 :e5 b6 25 .in a5 26 .ie4 l.tfB
27 ~g2 ~d6 28 ~f3 ltJd7 29 lte3 llJb8
30 :d3+ ~e7 31 e3 tOe6 32 l:le3 ~d6 43 :d3!
59
Fischer exchanges rooks, after which

the bishop's domination over the knight
will be even more potent.
43 ... ~e7 44 l1xd6 ~xd6 45 ~d3
White has a winning advantage as

Black's kingside pawns are fixed on the
same colour square as the bishop and
he has a route for a queens ide invasion.
62 ~xg6!
45 .. .ltJe7 46 iLe8 ~d5 47 ~fi+ ~d6 White's passed pawn will be
48 ~e4 ~e6 49 ~e8+ ~b7 decisive.
62 ... lDxg6 63 ~Ib6 ~d7 64 ~xe5
lDe7 65 b4 axb4 66 cxb4 lDc8 67 a5
lDd6 68 b5 lDe4+ 69 ~b6 ~e8 70 ~e6
~b8 71 b6 1-0
Fischer played the entire endgame
very powerfully.
Now let's have a look at an endgame

with a very similar queenside pawn
structure - just like the famous Fischer-
50~b5 Taimanov encounter.
The way Fischer improves his
position is quite beautiful. G.Kasparov - N. De Firmian
50 ... lDe8 51 ~c6+ ~c7 52 ~d5 lDe7 PCAlIntel-Grand Prix, New York 1995
53 ~fi ~b7 54 ~b3 ~a7 55 ~d1
~b7 56 ~f3+ ~e7
34 ~Ie6!
My pawn structure is very similar to
57~a6 the Fischer example. In that game
The white king invades. Bobby swapped rooks. I knew the
57 ... lDg8 58 ~d5 lDe7 59 ~c4 lDe6 bishop was not the same piece but I
60 .i.f7 lDe7 61 .i.e8 ~d8 followed his exchanging idea to invade.
60
34 ...~xe6 35 ~f3 <std6 44 c4 B lack is in zugzwang and

Closing the queens ide with 35 ... .i.c1 White still has a spare tempo at his
would be clever, but White is not disposal. 44 ... ~c7 45 ~a7 ~c6 46 Wb8
obliged to allow that. 36 ~e2 .i.xb2 and White invades.
37 ~d2 ~a3 38 ~e3 White wins. 37 ...Q;c6 38 ~el ~g5 39 ~c4 ~e3
36 Cifi>e2 ~c1
40 b4!
Aiming to open the position for
37 <it'd3 invasion. As Black is in zugzwang the
I stopped short of exchanging; game continuation is forced.
luckily it did not spoil anything. 40 ... cxb4 41 cxb4 axb4 42 .ltxb4
After 3 7 ~e3 the pawn ending was ~cl 43 ~f8 ~g5 44 .ltg7
simply winning. 37 ... .ltxe3 38 <itxe3 f6 As often happens in same coloured
(on 38 ... c4 39 g5! White soon promotes bishop endings, the weaker side is
the pawn to a lady.) 39 <iitd3 ~d7 caught by a zugzwang.
40 ~c4 ~c6 41 b3 (White has a lot of 44 ... f6
spare moves to lose a tempo.) 41...<iitd6 A sad necessity. Black has to put one
42 <iitb5 Q;c7 43 ~a6 (White has to be more pawn on the same colour as the
careful; he can't do whatever he wants, opponent's bishop. 44 ... tit>d6 was not
e.g. 43 g5? fxg5 44 g4 [44 Cifi>a6 g4] any better. 45 ~b5 ~e3 46 ~f6.
44 ... ~b7 and Black holds.) 43 ... Cifi>c6
(43 ... c4 44 bxc4 ~c6 45 q;a7 1itc7
46 c5 bxc5 47 Q;a6 <it'c6 48 ~xa5 and
White wins.)
45 ~h8!
I was lucky to have an extra square
available on the diagonal.
61
45 ... ~d6 46 ~b5 ~c7 47 .i.g7 I followed Fischer. It was a close call
The bishop's objective is to get to d8 but nevertheless I won!
in order to net a pawn. In general I am not going to compare
47 ... ~b7 48 ~f8 the champion's effectiveness at
Transferring the bishop to d8. damaging my career. Maybe this
48 ... ~c7 49 iLe7 ~d7 50 iLb4 ~e3 'lucky' win makes Fischer's effect on
If 50 ... ~c7 51 .i.e I .i.e3 only me less negative.
temporarily prevents the bishop from Before I show the games in which I
invading. 52 g5! fxg5 53 g4 .i.c5 emulated Fischer's play, I would like to
(53 ... .i.d4 54 iLb4 White wins.) 54 a5 present one game on a topic already
~b7 55 axb6 (55 a6+ ~a7 56 .i.c3 discussed in the Karpov section. This
it.d6 57 iLb2 and Black is in game was also planted in my mind as
zugzwang.) 55 ... .i.f8 setting up a well as Karpov's.
fortress can sometimes save an
identical coloured bishop ending. R.Fischer - M. Taimanov
Though this time it is ineffective, such Palma de Mallorca Interzonal 1970
a device can sometimes rescue the
1 e4 c5 2 ttJO ttJc6 3 d4 cxd4
weaker side. (55 ... it.xb6 56 it.b4 wins.)
4 ttJxd4 e6 5 ttJb5 d6 6 c4 a6 7 ltJ5c3
On 56 .i.c3 iLd6 57 .i.b2 Black is in
lDf6 8 iLe2 ~e7 9 0-0 0-0 10 ltJa3 b6
zugzwang.
11 iLe3
So far the players have followed
main line theory. Now the Russian
grandmaster deviates from the most
conunon 11.. ..i.b7.
11 ... .i.d7 12 l:ct 1t'b8 13 0 :a7
14 ttJc2 l:d8 15 'ilt'el iLe8
Black plays for b5. In the main line
they play for d5 or even ~h8 and :g8
with g5.
16 'Wif2 :b7 17 a4
51 g5!!
This stops b5 once and for all.
This is a very nice and instructive
17... a5 18 lDd4lDxd4 19 .i.xd4 ttJd7
breakthrough. Black's pieces are
overloaded.
51...fxg5 52 g4 ~e6 53 ~c6
The simplest option. Taking on b6
wins more quickly than going after the
h6-pawn. 53 .i.f8 would be winning as
well, since after 53 ... ~f6 54 .i.xh6 ~fl
55 ~c6 il.d2 56 ~d5 iLf4 57 ~d6
Black is in zugzwang.
53 ... iLd4 54 iLd6 1-0
62
Robert James Fischer they It!'
20 'iig3
I usually play on the queens~de
against the hedgehog set-up.
20 ... ~f6 21 ~xf6 lDxf6 22 l:[fdl e5
Black has obtained a fully playable
game.
23 'ii'h4 h6 24 l:[d2 lDd7 25 ~dl
lDc5 26 f4 exf4 27 'ir'xf4 lDe6
According to Vasiukov the position is
equal after 27 ... lle7! 28 ~c2 l:te5.
40 ~f3!
28 1fg3 1fc7 29 lDd5 'ifc5+ 30 'ithl
The bishop is very nicely placed on
~c6 31l:[c3 lDg5 32 ~c2 ~xd5 the diagonal. If White creates a passed
White is just a little better after a-pawn it will get tremendous support
32 ... :te8 33 h4! lDh7 34 lDe3 l:e6 from the long diagonal bishop. This is
35 lDf5 l:[g6 36 'ii'e3. what I wanted to do against Karpov, but
33 l:[xd5 'irc7?! the circumstances there were far less
This blocks the b7-rook. Better is fortunate than in Bobby's game.
33 ... 'irc6! 34 e5 (34 b3 l:[e7) 34 ...11bd7 40 ....t1d7
Black can live with his position after
35 ~f5 lDe6 when Black is safe.
40 ... l:tc7! 41 l:tb5 l:tc5.
41 l:[b5 l:[d4?
This wins a pawn but allows White to
open the queens ide. Better was
41...l:tdd8.
42 c5! l:txh4+ 43 'ltgl l:tb4 44 l:txb4
axb4 45 ':c4 bxc5
Or 4s ... lDd7 46 c6.
34 e5!
The more the position opens up, the
more the bishop has a chance to
dominate the knight.
34 ... dxe5 35 'iixe5 l:tdb8?!
Vasiukov's move 35 ... lDe6 IS more
46 l:txc5
natural.
Now the bishop is a real powerhouse.
36 ~f51lfxe5 37 l:.xe5 g6 38 h4 lDh7 46 ...ct>g7 47 as l:te8 48 l:tct! l:.e5
After 38 ... f6 B lack could exchange 49 ':al l:e7 50 ~f2!
the light pieces. Four-rook endings tend The a-pawn and the bishop are
to give considerable drawing chances. indeed strong, however they still need
39 ~g4lDf6 the help of the king.
63
Robert James Fischer the 11'h
50 ... ttJe8 51 a6 l1a7 52 ~e3 ttJe7 This move was inspired by one of
53.i.b7 Fischer's ideas - Random Chess. At the
White buries the rook. start of the game the pieces are
53 ... ttJe6 54 lIa5! ~f6 55 ~d3 ~e7 positioned on the first rank in irregular
56 ~e4 ~d6 or random positions. Somehow I must
have thought we were playing his brand
of chess, so I started to arrange my
pieces on the first rank in an
unorthodox manner.
17 ttJd2 g6
Of course the knight can't retreat to
b8 but ttJe7-g6-h8 would have given
a most exciting Fischer Random
position. Black would then only have to
transfer the c8-bishop to a8.
57 l:[d5+ ~e7 58 ~b5 1-0 18 ttJd5 f5 19 exfS
The king soon invades on b6 as well
and this decides the outcome of the
game.
In the next game Fischer had a

negative effect on my play like no other
champion.
V.Akopian - G.Kasparov
Russia v The World, Moscow 2002
19 ... gxfS?
1 e4 e5 2 ttJf3 It)e6 3 .i.b5 e6 4 0-0 Preoccupied with thoughts of Fischer
It)ge7 5 b3 a6 6 .i.xe6 ttJxe6 7 .i.b2 b5 Random Chess, I just wanted to keep
8 e4 bxe4 9 bxe4 nbS 10 .i.e3 d6 my pieces on the back rank. But better
11 It)a3! e5 12 ttJe2 .i.e7 13 ttJe3 0-0 was 19 ... .i.xf5.
14 d3 'ii'e8 15 l:[bl l:[xbl 16 'iixbl 20 f4!
Akopian puts pressure on the centre
and the king.
2o ...11n 21 ~el ng7 22 ttJf3 ~g6
23 g3
Here it dawned on me that we were
playing ordinary chess and that I was
now simply lost.
23 .. Jln
23 ... e4 was no better. 24 .i.xg7 ~xg7
25 dxe4 fxe4 26 ttJd2 .i.a5 27 ~a 1+
16....i.d8? ~f7 28 f5 wins.
64
24 fxeS f4 A.Karpov - G.Kasparov

Linares 1993
25 exd6
I had to resign here, because 22 ... c3 23 lOxa2 c2 24 'ii'd4
White has a decisive battery: 25 ... fxg3 cxdl='ii'+ 25 <iitxdl lOdc5 26 'ii'xd8
26 'ii'e8+ tUB 27 'ii'xf8+ 1-0 ':xd8+ 27 <it>c2 lOa 0-1
A most unfortunate encounter.
Incidentally Karpov himself also got After this effort my games against
caught by the Fischer Random virus. Karpov were far less regular than they
Here is his position: had been previously...
Fischer won a game in a I c4 c5 English type position, where his opponent

had a c4-pawn, while he himself had a d6-pawn and undennined White's
pawn structure with ... a6 and ... b5 and went on to win.
I played ... b5 under very similar conditions in three games, losing all three,
against Romanishin (below), Shneider and Anand (next page).
M.Aaron - R.Fischer O.Romanishin - G.Kasparov
65
A.Shneider - G.Kasparov V.Anand - G.Kasparov
M.Aaron - R.Fiseher 14... ttJxd5

Stockholm Interzonal 1962 14 ... .i.xd5!? 15 exd5 .i.h6
16 'iVxh6 ~a5+ 17 ir'd2 ~xb5 leaves
1 d4 ttJf6 2 e4 g6 3 ttJe3 .i.g7 4 e4 d6 Black slightly better.
5 f3 0-0 6 .i.e3 ttJbd7 7 'iVd2 e5 8 ttJge2 15 exd5 .i.Id5 16 a4 e6 17 0-0 ..-h4
a6 9 ttJg3?! exd4 10 .i.xd4 ttJe5 18 ttJe2 lUe8 19 .i.e3
11 .lte2 .i.e6 12 ttJd5 b5! It would be more appropriate not to
give up the two bishops by 19 lHc 1.
Then after 19 ... .i.h6 20 l:txc8+ l:txc8
Fischer undennines Black's pawn 21 ..-e 1 'ifg5 22 'IlVg3 'iff5 the position
structure. would be unclear.
13 exb5 axb5
19 ... ttJe4 20 ~xe4 'ilt'Ic4 21 :fc1

14 .i.xb5 'ifa6 22 lhe8+ ltxcs 23 ttJe3 .i.c4
White could not have gone a pawn up 24 f4
by 14 ttJxf6+ since after 14 ... .ltxf6 After 24 .i.d4 ~xd4+ 25 'ifxd4 e5
15 .ltxb5 comes the lovely tactical shot 26 'IlVd2 (26 'iVe3 ~b8) 26 ... 'fkb6+
15 tOxf3+! 16 gxf3 .i.xd4 17 'ii'xd4 27 'it>hl d5 it is easier to play Black's
'iWa5+. position.
66
S... liJxd4
I beat him in the Kosmos 1998 blitz
match with 8 ... fl)g4.
9 'ifxd4 d6 10 'ifd3 a6 11 ~e3 .i.d7
12 ~d4 .i.e6 13 e4 l:teS 14 lIfel .:teS
15 l:tad1 "a5 16 a3 bS
I also undermine the centre.
17 exb5 axb5
24 ... d5 25 .i.d4 .i.xd4+ 26 1ixd4
'iltb7!
Fischer improves his position with
strong, natural moves.
27 'iVf2?! ~a6 2slId1l:Ie4
IS e5!
It's a smart idea to weaken the
b5-pawn.
lS ... dxe5 19 .i.xe6 exd4!
I planned this exchange sacrifice.
29 l:td2? Unlike the Fischer game, here the
A bad blunder in a tough position. b5-pawn is really weak after 19 ... l:txc6
29 ...Axe30-1 20 ':xe5.
20 ~xeS dxe3
This was a convincing game, sO
decided to give the a61b5 plan a try:
Moscow-4-teams 1981
1 lbn g6 2 d4 .Jtg7 3 g3 lbf6 4 .i.g2

0-0 5 e4 e5 6 0-0 exd4 7 lbxd4 lbe6
Slbe3
I had a few irregular games against
Kramnik in this variation. Once he
drew after playing an early d6 and .i.d7 21 ~d7!
instead of castling - and once when I This subtle intermediate move forces
withdrew my knight to c2. the rook to a less effective square.
67
21. .. l:t.d8 despite being the exchange down.)

Black has nothing for the exchange 27 ... h6 28 h4 c2 (28 .. .'''xb4 29 l:.d4
after 21.. .l:t.b8 22 ~xc3. If 21...l::tc7 [29 g5 'iWg4+=] 29 ... l:.e7 30 l:.xb4 l:.xe5
Black can exchange all of White's 31 l:.xe5 .i.xb4 32 .i.a4 [32 f3?? c2]
queenside pawns, but only just - and it 32 ... .i.d6) 29 l::lal (29 g5 cxdl='iW=)
requires very precise calculation. But 29 .. .'''xal 30 1i'xal (30 l::lxal
there is no point entering into this when
there is a more comfortable line in
22 b4! 'iVxa3 23 ~xb5
30 ... J1.d6!! 31 'iWxf6+ 4it>xf6 32 l:.cl

l:.c7 Black is safe.) 30 ...:xal 31 %1xal
tOd5 (Black can also draw by
23 ... ~f8!! (After 23 .. J%c8 24 ~a6 31.. .J1.xb4 as well as White has no time
White is better.) 24 'ii'd8 (24 l:tbl 'iWa2 to keep all his kingside pawns. 32 l::lc 1
25 l:te2 [25 'iWd8 tOe4 26 ltfl tOxf2! tOxg4 33 Ibc2 h5) 32 ~a4 tOxb4
27 l:tal tOh3+ 28 'it>hl 'iVe6 29 l:.ael 33 ltcl J1.d6 and Black has no problem
'ii'a2 and White is unable to take the at all.
rook, so Black is not worse.] 25 ... c2 22 'ii'xb5
26 l:lc I tOd5! Black wins the b-pawn
and survives. 27 :exc2 tOxb4 and
Black gets away with it.) 24 .. .l:ta7
25 'ii'd4 e5! 26 'iVxe5
22 ...1Vc7?
After 22 ... 'ii'xb5 23 ~xb5 l:tb8
24 bxc3 (upon 24 a4 cxb2 the pawn on
b2 is really strong and compensates for
26 ... ~g7!! 27 g4 (27 -.d4 :b7 the exchange. After 25 %:te2 tOd5
[27 ... J1.xb4] 28 l:t.al ~xb4 29 -.xb4 White's advantage is symbolic.)
~xb4 30 .i.d3 ~c5 and Black holds 24 ... l:Ixb5 25 llxe7 Now one can see
68
why my text move was a big mistake. 34 nc8!

Black still has the rook to hold the Now I can't even sacrifice a single
queenside pawns until his minor pieces one of my pieces.
come over to help. 1-0
23 bxc3 J:[xd7 24 l:txd7 'ii'xd7
25 1i'xd7 lLlxd7 A.Shneider - G.Kasparov
EU Cup, Lyon 1994
1 d4 lLlf6 2 c4 g6 3 lLlf3 ~g7 4 g3 c5

5 ~g2 cxd4 6 lLlxd4 0-0 7 0-0 lLlc6
8 lLlc3 lLlxd4 9 'i!fxd4 d6
26l:he7
Black has no pieces on the queenside
to hold back White's pawns.
26 ... lLlb6
After 26 ... lLlc5 27 ':'e8+ ~f8 28 l:a8
lLle4 29 a4 Black is in trouble as well. 10.d3
27 l:b7 lLla4 28 l:[b8+ ~f8 29 c4 I knew that Fischer had beaten
~g7 30 ~g2 ~d6 31 1:Ia8 lLlb2 Spassky in the 8 th game of their 1972
match but had an idea I wanted to try
out to combat Fischer's plan. I must
admit that those games where two
champions played each other always
gave me such a headache. Which
champion to follow? When I began
to realise that they could also be
inaccurate even when they won, then it
became even more confusing.
10... a6 11 ~d2 l:tb8 12 Aael b5
32 a4 lLlxc4 33 as lLle5?
Taking the pawn was possible, but it
offers no chances of holding the game. 13 b3
I had virtually the same position against Oops, what to do now - they took on
Speelman in Graz 1981 and easily b5 in the previous examples.
converted my advantage. However, if 13 ... ~f5
White's pawn were on h4 and Black I wanted to close the diagonal,
had one on h5 that might be a draw. thinking it was worth a tempo.
69
14 e4 Jtd7 15 h3!? 23 lid 1 ttJe8

This stops ... ttJg4. Here 23 ... ':fd8 24 ~b6 (24 a3 'ii'c3
15 ... bxc4 16 1t'xc4 "a5 2.5 'ii'xb5 :b8 26 :xd6 wins prettily.)
16 ... 'iVb6 is less provocative. 24 ... ':d7 25 a3 'ii'c3 26 'ii'xc3 l:txc3
17 -.d3 ~a3? 27 11a8+ Jtf8 28 Jtd4 and Black must
resign.
24 a3 'ii'c3 25 'ii'xb5 'iib2 26 ~d3
Simpler is 26 h4.
26 ....:c3 27 '6b1 'iixb3 28 'ir'xb3
1:tJ:b3
18 l:tc2! Jtb5?
It may seem weird but perhaps
retreating with the queen was
objectively stronger than this move.
Not 18 ... l:f.fc8? when 19 liJbl! traps the
queen.
19 ttJxb5 axb5 20 l:f.c7 e6 21 ~e3! 29 a4
l:f.bc8 In Fischer's game the queenside
Other moves were miserable as well. pawns played no role. My opponent?
a) 21. .. '1IVxa2? 22 :a7 'iWb2 23 Jtd4 unlike Fischer's, opened the back rank,
wins. so there was no real hope.
b) 21 ... :a8 22 :fcl "xa2 23 e5 29 ... h5 30 a5 ':a3 31 a6 ~e5
hurts. 32 Jth6 Jtg7 33 Jtg5 Jtf6 34 ~xf6
c) 2L.J:Hc8 22 lIa7 (22 ':fcl l:txc7 liJxf6 35 lbd6 :al + 36 ~h2 l::ta2
23 lbc7 ttJe8 (or 23 ... ~xa2 24 :a7 37 e5 ttJb7 38 lIad7 lIxfl 39 l::td2 1-0
W'b2 25 "xd6) 24 :a7 'iib2) 22 ... ~4
23 .%:dl liJe8 and Black is passive.
V.Anand - G.Kasparov
22 l:f.a7 'irb4
PCAllntel-Grand Prix, Moscow 1995
1 e4 c5 2 ttJf3 d6 3 d4 cxd4 4 ~xd4

In 1999 Peter Svidler tried this line
against me. That time I won.
4 ... ~d7 5 c4 ttJc6 6 "d2 g6 7 Jte2!?
Anand now stops 7 ... Jth6, as 8 'ii'c3
is then possible.
7 ... Jtg7 8 0-0 ttJf6 9 ttJc3 0-0 10 l::tb 1
a6 11 b3 "a5 12 Jtb2 :fc8 13 l:f.fdl
70
In the preceding game (against

Shneider) White did not take and that
led to trouble. Maybe it was primarily
because of that game that I
subconsciously did not expect my
opponent to capture here. It was better
to stop the h-pawn with 21...h5.
22 cxbS
13 ... ~g4 Taking the b5-pawn simply wins.

Perhaps it is would be more efficient Black doesn't lose just one pawn, but
to prepare b5, e.g. by 13 ... l:.ab8 14 h3 two.
b5. 22 ... axbS 23 ttJxb5 ttJbc6 24 a3 d5
14 'i:Ye3 ttJd7 15 ttJd5 ~xb2 16 %bb2 25 exdS lIxd5 26 l:txd5 exd5 27 b4
~xf3 17 ~xf3 e6 18 ttJc3 'it'a4
18 ... lId8? 28l:bdS 1-0

This is too passive. Usually playing a
dynamic move like 18 ... b5!? would There are many elements to Fischer's
come naturally to me. Fischer also did contributions to chess culture, some
not move his centre much, maybe that of which are quite complex. However,
is why I postponed it. 19 lIxd6 ttJce5 in pure chess terms, perhaps his
20 ~e2 (20 lIbd2 ttJf6 21 c5 1i'a3) contribution to his own pet opening -
20 ... bxc4 21 b4 \!Va3 22 \!Vd2 a5 23 b5 the Najdorf - is the biggest. He adopted
ttJd3 and Black is in the game. it in almost all his games when faced
19 lIbd2!? with 1 e4. I employed some other
When the rook took on b2, it was variations at times but the Najdorf was
simultaneously building up White's my most common response.
position. Suddenly the d6-pawn is
Fischer was so good at taking pawns
vulnerable.
and calculating precisely. I also took
19 ... ttJde5 20 ~e2 ttJb4 21 h4 bS??
the e5-pawn once ... but let's start with
See diagram on page 66. Fischer's game.
71
B.Larsen - R.Fischer
Fischer took the centre pawn

on e5 with his knight and
calculated precisely to win.
So I decided I would try the

same thing.
H.Lehmann - R.Fischer Velibekov - G.Kasparov
H.Lehmann - R.Fischer 11 ~hl <lth8 12 a4 nb8 13 g4 b6 14 g5

Capablanca Memorial, Havana 1965 tbe8 15 Jtg2 ~b7 16 b3 fi'd8 17 h4 g6
18 .i.b2
1 e4 cS 2 tbf3 d6 3 d4 cxd4 4 tbxd4
tbf6 5 tbc3 a6 6 ~e2 tbbd7 7 0-0
I had a nice and in my opinion very
instructive win against 7 a4, as played
by Short in the Moscow Olympiad
1994. White should try to follow my
plan despite the different move order.
G.Kasparov - N.Short
Moscow Olympiad 1994
I e4 c5 2 tbO d6 3 d4 cxd4 4 tbxd4 18 ... tbg7 1911dl l1e8 20 f5 e5 21 f6

tbf6 5 tbe3 a6 6 Jte2 e6 7 f4 ~e7 exd4 22 fxe7 'fIxe7 23 l1xd4 f6 24 'ilfd2
8 0-0 "ike7 9 "ikel tbbd7 10 JtO 0-0 fxg5 25 l::txi8+ l1xi8 26 l1xd6 tbe5
72
27 l:hb6 gxh4 28 ltJd5 ..i.xd5 29 'ii'xd5 17... 'iYhS!

':e8 30 .th3 ~c7 31 :e6 llxe6 Gaining an important tempo.
32 .txe6 ltJc6 33 1fg5 ~d6 34 .td5 18 'ifel .te7 19 b3 0-0
ltJb4 35 'ii'f6 'ir'xf6 36 .txf6 ltJxc2 Fischer has survived and the win was
37 .tc3 h6 38 b4 ~h7 39 b5 axb5 just a matter of time.
40 axb5 ltJh5 41 b6 ltJg3+ 42 ~h2 1-0. 20 a4 'ir'cs 21 'ii'e2 l:tac8 22 e4 bxe3
23 ..i.a3 'ike7 24 i..xe7 ihe7 25 'ilfxeS
Back to the Fischer game.
':cS
7... e68 f4 bS 9 ..i.o i..b7 Fischer calculated so well.
26 'ife2l:hgS 27 'ii'xa6"'4 28 :tbl
.:td8 29 as h6 30 'ilfe4 ihc4 31 bxc4
c2 32:c1 0-1
B.Larsen - R.Fischer
Game 6, Candidates match,
Denver 1971
1 f4 cS 2 ltJo g6 3 e4 i..g7 4 i..e2

ltJc6 5 0-0 d6 6 d3 e6 7 ltJa3 ltJge7 8 c3
10 eS 0-0 9 i..e3 a6 10 d4 exd4 11 ltJxd4 bS
White feels like forcing the issue 12 ltJxe6
without delay.
10....txO 11 ltJxo dxeS 12 exeS
ltJg4 13 'ii'e2 b4 14 ltJe4
White scores well without sacrificing
the e5-pawn. He could opt for 14ltJa4.
14 ... ltJgxeS 15 ltJxeS ltJxeS
16 ltJgS
White has to react quickly since if
Black castles then that's it. 12 ... ltJxc6
16...'ifb6+ 171it>hl Black has equalised effortlessly.
13 'ifd2 'ilt'c7 14 .:tadl ];Id8 15 ltJc2
l:tb8 16 83 ltJaS 17 eS
At this point the match stood at 5-0-
such a shock for Larsen. Nevertheless,
he presses on as he has nothing really to
lose.
17.....i.fB 18 b4 ltJc6 19 ltJd4 dxeS
20 exeS ltJxeS
73
Fischer again takes the e5-pawn in a 30 ':xf7 is one way to draw. Then
Sicilian. 30 ... ~xf7 31 lin + ltf5 32 1i'xh7+.
21 ~g5 l:td5 22 ~f4 ~g7 23 b4 30 ttJxe6 is another. Then after
l:tb7?! 30 ... .i.xe6 31 ':xd5 .i.xd5 32 -'g5+.
Alternatively, 23 ... iYxc3! and White 30 ... f5 31 -.f6??
has almost nothing for the sacrificed Larsen fmally cracks. How else to
material. explain why a world-class player
24 ~f6? ~xf6 25 ~xf6 makes a losing move like this.
31 ~h2 allows the knight to move.
3l ... -.c7 (3l ... ~c8?? 32 ttJxe6 wins.)
32 -'g5+ ~f7 33 'ir'h5+ and there is
another perpetual.
After 31 ~g5+ rj;f7 321rb5+ Black's
king should not try to run away from
the checks by 32 ... ~e7 33 -'g5+ ~d6?
25 ...'iVxc3 26 h5 gxb5
After 26 ... l:txd4 27 ~xe5 l:td3
28 'iif4 ~xdl 29 lhdl ~d7 30 ':d3
White has compensation.
27 ~bl
27 ttJxe6 simplifies to an equal
position after 27 ... ~xe6 (27 .....e3+
28 Wh2 ~xe6 29 l:txd5 ~xd5 ... as 34 ':xf5!! and suddenly Black's
30 ~d8+) 28 lhd5 ~xd5 29 ii'd8+. king is under fire.
27 ... ttJg4 28 ~xg4 hxg4 29 1i'h6 31. .. .i.c8 32 l:[ffl
~d7
32 .. .l:1n
30 l:H4 It's all over now.
Larsen refuses to force a perpetual, 33 'iWh6 ~b7 34 ttJxe6 ii'f6 35 1i'e3
still hoping to break his duck. l:[e7 36 l:[del l:[d6 37 'it'g5+ iYxgS
74
38 lOxg5 l:hel 39 ~hel .i.d5 40 ':e8+ such a good player. 26 ... h5 27 <atgl
cwfi>g70-1 lDe7 28 ~d3 lDd5 29 ~d2 i.e7
30 ~f2 i.c5+ 31 ~e2 i.d6 32 c4 bxc4
Karpov also took an e5-pawn and 33 .i.xc4 ~xg3 34 .i.xa6 lOf4+
went on to win. 35 .i.xf4 .i.xf4 36 lDe3 g5 37 lDd2 f5
38 a4 ~f7 39 lDdc4 %:tb8 40 lDc2 g4
V.lvanchuk - A.Karpov 41 lDd4 h4 42 lDc6 %:ta8 43 i.b5
Sicilian tournament,
Buenos Aires 1994
I e4 e5 2 lOO e6 3 d4 exd4 4 lOxd4

lOe6 5 lOe3 fIIc7 6 ~e2 a6 7 0-0 lOf6
8 cwfi>hl ~e7 9 f4 d6 10 ~e3 0-0 II 'iVel
~d7 12 'fig3 ~h8 13 ':adl ':ac8
14 lOo lDb4 15 lDel b5 16 a3 lDc6
17 e5 dxe5 18 fxe5 'iVxe5
43 ... h3 44 gxh3 gxh3 45 lD6e5+ ~f6

46 lDd7+ ~e7 47 lDdb6 J::th8 48 i..c6
h2 49 a5 i..b8 50 a6 ~d8 51 ii.hl ':g8
52 lDe5 1;c7 0-1
Of course I too developed an appetite

for swallowing the e5-pawn in the
Sicilian.
19 .i.f4 'iVe5 20 .i.e3 fIIe5 21 l:bf6 Velibekov - G.Kasparov

'iVxg3 22 hxg3 .i.xf6 23 ':xd7 ~g8 USSR 1976
24 lDd I .J:tfd8 25 lhd8+ ':xd8
1 e4 c5 2 lDn e6 3 d4 cxd4 4 lDxd4
lDf6 5 lDc3 d6 6 ~e2 a6 7 0-0 lObd7
8 f4 b5 9 ~n ii.b7 10 a3
My opponent may have known the
Fischer game.
10.....c7 11 ~hl
It would also be nice to know
whether he was aware of the following
game between two former champions.
26 lDo It is hard to imagine how After 11 'iVel ~e7 12 ~hl %:tb8 13 b3
Karpov can win this endgame against 0-0 14 ~b2 ':fe8 15 'ii'g3 ii.f8
75
Yugoslavia 1959. What a hacking

game.
16 l:tae 1 e5 17 ttJf5 ct>h8 18 ~4

exf4 19 1i'xf4 ttJe5 20 l:te3 g6 21 ttJh6
iLg7 22 ttJd5 ttJxd5 23 exd5 f6 24 iLe4 1l ... bS?!
g5 25 'iff5 i..xh6 26 "xf6+ i..g7 This is a risky move!
27 "f5 ttJg6 28 lih3 12 -'.e3 ttJcS 13 eS dxeS 14 fxeS ttJg4
15 -'.xb7 "xb7 16 i..gS ttJxeS
I took the pawn - just as Fischer liked
to do.
17 "e2 ttJcd7 18 :lad1 -'.e7 19 ttJe4
Suddenly Black has difficulty in
finding a continuation.
19 ...1i'c7 20 iLxe7 ct>xe7
Steinitz didn't mind placing his king
28 ... i..xb2 Black IS winning. in the centre.
29 "xg6l:te7 30 l:th6lig8 31 "f5 iLc8 21 ttJgS l%afS 22 'iVe1
32 1i'0 g4 33 'ifd3 iLe5 34 c4 bxc4
35 bxc4 lieg7 36 c5 dxc5 37 d6 1i'a7
38 i..d5 l:td8 39 'iWe4 i..d4 40 'iH4
22 ... ttJcS??
This is a dreadful mistake. I came to
the conclusion that it is not as simple to
40 ... l:tgd7?? (40 .....d7 was winning.) take the e5-pawn as one may think.
41 1:I.f6! 1-0 Tal-Smyslov, Candidates, 23 'iVxeS! 1-0
76
Robert James Fischer the JJilt
When the King's Indian becomes a Benko Gambit...
V.lvanchuk - G.Kasparov
M.Cuellar - R.Fischer
The Najdorf was my pet opening, well. On the other hand, I already
whereas the Benko Gambit was a rare hinted that a Benko type pawn sacrifice
choice. I played it only a few times. might occur.
I should clarify that the Benko type 13 liJxa5 'ft'xa5 14 liJxb5 ~xb5
of position in the King's Indian 15 cxb5 l:[fb8 16 ~fJ liJf6 17 a4 a6
sometimes transposes to the Benko and
then a similar pawn sacrifice occurs. See diagram above!
18 bxa6
M.Cuellar - R.Fischer After 18 ~d2!? 'fVb6 19 e3.
Stockholm Interzonal 1962
1 d4 lLlf6 2 c4 g6 3 g3 SLg7 4 .i.g2
0-0 5 lLlfJ d6 6 0-0 lLlc6 7 lLlc3 .i.f5
8 d5 lLla5 9 lLld4 ~d7 10 'ifd3 cS
11 lLlb3 lLlg4 12 f4
18...'ii'Xa6
For players who do not know the
Benko, it comes as a small surprise that
Black exchanges pieces when he is a
pawn down. But the idea has its logic.
Black exchanges in order to clear
12 ... b5 squares for an invasion by his well-
This pawn sacrifice is rather a positioned pieces. Black has no need to
surprise as the knight can capture as fear the endgame.
77
19l'la3 ~xd3 20 exd3 :b4 21 as 32 ... lbxdS

21 b3!? keeps the pawn, but Fischer Black moves ahead in material and
would have compensation anyway. White's position deteriorates very
21 .. J~bS quickly.
33 llb3 lbb4 34 fS gxfS 35 ~gS e6
36 ..i.d8 lIa8 37 .i.b6 lIxc8 0-1
I beat Bareev with the Benko at

Linares 1994 and in the last round of
the Dubai Olympiad I defeated
Schmidt when the Soviet Union needed
to win 4-0. Evgeny took the pawn
whereas Schmidt kept the position
closed. Despite these pleasant
22 ~d2 nxb2 memories, when I think of the Benko
Black recovers the pawn which Gambit it is the Fischer effect that is the
most pronounced and it really hurts.
makes him feel more comfortable,
however he is not yet better as the The next defeat helped Karpov to win
white a-pawn stymies him and his another Linares tournament.
opponent's position is not loose enough V.lvanchuk - G.Kasparov
to invade. Linares 1997
23 ~c3 l:tb7 24 l:[el lbe8 25 ~d2
This defends a5 but allows Black's 1 d4 lbf6 2 c4 g6 3 lbc3 Jtg7 4 e4 d6
dark-squared bishop to assume a 5 f3 0-0 6 .i.gS a6 7 'it'd2 cS 8 dS bS
dominating role. 9 cxbS lbbd710 a4 ~aS lllbge2lbb6
2S ... Ci!ff8 26 ~d1l:tb2 27 il.c1 ~d4+
28 ~h1l::[f2 See diagram on page 77.
White has the usual problem in the
Benko of keeping his position together. 12 lbc1 axbS 13 .i.xbS .i.a6
29 ~g4 lbf6 30 il.h3 l::[c2 31 a6 :a7 14 lb1a2
32 ~c8 If 14 .i.xa6 -'xa6!
If 32 il.g2 lbg4. 14 ... .i.xbS
78
15 axbS 22 c;t>xf2 c4 23 b4 iJla7 is unpleasant for

The pawn fonnation is slightly White) 20 ... 'ii'xb5 21 lDac3 'iWa6 is
different from that in the Fischer game. lovely for Black.
Is ... lbhS 18 b3 e6 19 dxe6 l:txe6!?
Black could try swapping pieces on After 19 ... fxe6!? comes 20 i.e3.
the queens ide with 15 ... lba4!? 20 .i.e3 .i.xe3 21 'ii'xe3 dS
16 ':'bl
After 16 0-0 .li.d4+ 17 Wh 1 .li.xc3
18 bxc3 f6 19 .li.h6 l:tfb8 Black has
compensation in an unusual form.
16....li.d4
I also manage to put my bishop on
the dominating d4-square.
17.li.h6
22 b4
Not 22 'ili'xc5? lDf4 23 0-0 d4!
24 'ili'xd4 'ir'xa2 25 :f2 (25 lDxa2??
lDe2+) 25 ... 'ii'a5 and this time the three
pawns are not enough for the piece.
22 ...iJla3?!
After 22 ... cxb4 23 lDxb4 dxe4 24 0-0
exD 25 'iVxD Black should be able to
17...l:tfe8 live with his small disadvantage.
Bobby put his rook here too. Of 23 bxeS lDe4
course he was defending the e7-pawn If 23 ... dxe4 24 0-0 exD 25 iJld4 fxg2
whereas I was looking for dynamic 26 :f2 lDa4 27 c6 White's passed
play, but his rook move affected me and pawns are menacing.
played a negative role in my decision 24 'ii'd4 lDf4
making.
Maybe the simplest way was to exert
pressure on the queenside with
17 ... .:.fb8!? and try to gobble up the
b5-pawn. 18 b3 lDd7 19 .li.e3 (19 b4
'ili'a3; 19 <ifi'e2 lDe5) 19 ... .li.xe3 20 'iWxe3
':'xb5 and Black is better. I also
suggested the dynamic attempt 17 ... f5?
sacrificing the exchange. Then 18 .li.xf8
':'xf8 19 b3 (19 b4!? 'iWa7 and Black has
nice play for the material.) 19 ... fxe4
20 lDxe4 (20 fxe4 :f2 21 'iixf2 .li.xf2+ 250-0!
79
Ivanchuk simply sacrifices a knight 271O:.:dS

to neutralise Black's play and get his White's connected passed pawns
passed pawns rolling.
2S ...1ha2 26 .:tn! can't be contained and he has a
Not 261Oxa2? because oflOe2+. completely winning position.
26 ... 'iIr'a3 27 .•.1Wd3 28 'ihd.3 lO:.:d.3 29 l1c2
lOa3 30 :a2 lO:.:cS
After 30 ... lOxbl 31 lba8+ ~g7
32 c6 wins.
31 lIbal fS 32 lOc7 lieS 33 lOxa8
lOxbS 34 exfS g:.:fS 35 lLlb6 lLlc3
36 lIc2 1-0
And here I lost on time.
80
Boris Spassky the 10 th
The tenth world champion reigned that mantle, and in 1972 won the title.
from 1969 to 1972. He defeated Tigran Spassky, like Capablanca, Euwe,
Petros ian at the second time of asking. Smyslov, Tal, Fischer and Topalov
Beating the Armenian world champion never successfully defended the title.
was in itself a great achievement, but Of these, Fischer was the only one who
winning the Candidates matches twice did not actually try to do so. I wrote in
was also great. In the second half of the the Predecessors book that Spassky's
60s he was probably the strongest style was more attacking than
player. From 1970 Fischer took over universal.
I copied a positional idea from Spassky, but it did not payoff. He used to
create many problems for Black in queenless variations of the Queen's Gambit
Accepted. The pawn structure is symmetrical and yet Spassky managed to
inject power into the proceedings. I also tried the idea of pressing in a
symmetrical queenless opening. Here are the positions:
B.Spassky - R.Fischer G.Kasparov - P.Leko
Let's start with a game by two world This is one of Spassky's pet lines. He
champions. managed to breathe life into this
seemingly dead boring variation and
B.Spassky - R.Fischer after his return match with Fischer the
Game 4, St StefanlBelgrade match line caught on. Kramnik tried it against
1992 me as well. Now we understand this
1 d4 d5 2 c4 dxc4 3 liJo liJf6 4 e3 e6 position much better than before
5 ~xc4 c5 6 0-0 a6 7 dxc5!? Fischer-Spassky 1992.
81
Boris Spassky the 10lh
7 ... 'ih:d1 9 ... ltJbd7 10 il.b2 b6

In a match, players tend to be happy This was Fischer's first reaction to
with a draw. For a long time people did the problem. He holds back b5 as he
not understand how many pitfalls there doesn't want the queenside to be
were in this variation. attacked by a4.
8 lb~1 il.xc5 On the next two occasions the
American no longer rejected the
tempo-gaining 10 ... b5 and after 11 il.e2
il.b7 12 ltJbd2 Fischer castled when he
faced the line for the third time.
12 ... 0-0. Interestingly, castling has an
unusual advantage compared to what
happened in his second attempt at this
variation. Here the king defends the
g7-pawn. The more natural king move
12 ... ~e7 was Fischer's choice upon
facing the line for the second time,
Despite its symmetrical nature, this when play continued 13 a4 bxa4
endgame provides White with more 14 l1xa4 l:[hb8 15 nc1 iLd5 16 ltJe5
opportunities to fight for an advantage il.d6 17 ltJxd7 ltJxd7 18 l:1xa6 lha6
than was at first thought. 19 il.xa6 f6 20 il.c4 il.xc4 21 l'hc4
9 b3 ltJc5 22 :c3 f5 23 il.a3 ltJe4 24 l:lc7+
Spassky tried this move three times ~d8 25 il.xd6 ltJxd2
against Fischer in the match.
In the fourth and [mal occurrence of
the line Boris changed over to another
plan with 9 ltJbd2. By this time Fischer
had learned how to neutralise the
endgame advantage. 9 ... 0-0 10 a3 b5
11 il.e2 il.b7 12 b4 iLe7 13 il.b2 ltJbd7
14 ~ac 1 ~fc8 This way of developing
looks very convincing for Black and he
has equalised the position. 15 ltJb3
lhc 1 16 ~xc 1 ~c8 17 ~xc8+ il.xc8 And after 26 lhg7 Bobby survived
18 ltJfd4 ltJb8 19 il.f3 ~f8 20 ltJa5 this lost position.
iLd6 21 ltJdb3 e5 22 ltJc5 ~e7 23 h3 It may have affected me in some
ltJfd7 24 ltJd3 f6 25 il.e4 g6 26 f4 exf4 games, but I have not worked out yet in
27 exf4 ltJb6 28 ltJb7 il.c7 29 ltJbc5 which one. 26 ... ~xb3 27 h4 h5 28 il.f4
ltJc4 30 iLc 1 ltJd7 31 ~fl ltJxc5 ~e8 29 ~h2 :b2 30 ~h3 ltJe4 31 f3
32 ltJxc5 il.b6 33 il.d3 il.xc5 34 bxc5 ltJf2+ 32 ~g3 ltJd3 33 il.g5 e5 34 <it>h3
.i.e6 35 ~f2 ~d7 36 il.xc4 il.xc4 Ih-l,h ltJf2+ 35 'i!fh2 ltJd3 36 il.h6 ltJel
Spassky-Fischer, Belgrade 1992. 37 ~gl ltJd3 38 il.g5 ~bl + 39 c;t>h2
82
Boris Spassky the 10,h
l:tb2 40 l:le7+ <;tJf8 41 l:le6 q;g7 42 ~h3

%:te2 43 l:td6 4:Je 1 44 .i.f6+ c;t>g8
45 .i.xe5 .the3 46 Sl.f4 %:te2 47 l:lg6+
~f7 48 l:g5 litte6 49 .i.c7 l:a2 50 .i.b6
4:Jd3 51 ~h24:Je 1 52 ~h34Jd3 53 .i.c7
%:tc2 54 .i.b6l:ta2 55 ~g34Jel 56 l:lxh5
l:txg2+ 57 ~f4 4Jd3+ 58 ~e3 4:Je5
59 l:Ih6+ <ttd5 60 .i.c7 %:tg7 61 .i.xe5
~xe5 Ih-th Spassky-Fischer, Belgrade
1992} 13 %:tac1 l:tfc8 14 h3 ~f8
(Fischer still returns to the centre.) 14 f3!?
15 c;t>n q;e7 16 4:Je 1 .i.d6 17 a4 I also won a few games by freezing a
b7 bishop, for example against Karpov
in the second game of our 1990 World
Championship match. However, as you
will probably guess, this plan is not as
sparkling as it looks.
14 ... b5 15 i.e2 Sl.e5! 16 ~fl! ~e7
17 ... .i.c6 (Black defends the

queenside in a different way.) 18 axb5
axb5 19 l:c2 l:lc7 20 lidc1 l:tac8
21 .i.f1 i.xf1 22 4Jdxf1 e5 23 :xc7
l:txc7 24 l:Ixc7 .i.xc7 25 4Jc2 4:Je4 17 e4
26 liJa3 b4 27 4:Jc4 f6 28 4:Je 1 liJdc5 This is the position that confused me.
29 4Jc2 liJxb3 30 4:Jxb4 4:Jbd2+ The outcome of the game and the
31 4Jxd2 liJxd2+ 32 <ite2 liJc4 th-Ih pressure Boris managed to exert in this
Spassky-Fischer, Belgrade 1992. line prompted the idea of going for a
llliJc3 position with a similar pawn structure.
This was Boris' deviation from one See diagram on page 81.
of his earlier games. 17.•. g5
1l ... .i.b7 12 l:Iacl i.e7 Bringing the h8-rook to the
After 12 ... h6 comes 13 liJa4! which queenside was playable for Black.
is the point of White's 11 th move. Then IS 4:Jbl g4 19 i.a3 b4?
13 ... i.e7 14 i.xf6 .i.xf6 15 l:Ixd7! and Taking on a3 was better.
Black loses a pawn. 20 l:1xe5!
13liJd4 lieS The exchange sacrifice brings
Not 13 ... 0-0? 14liJxe6! Fischer into an eternal pin.
83
Boris Spassky the 10th
20 ... 4JxcS 21 -ltxb4 l:thd8?! If 33 ~d3 l1bcS 34 e5+ rt;e7 35 b5+

This turns out to be a loss of tempo. 4Jbc5 36 4Jb4 ~b7 (36 ... l:th8 37 ltJa6
Better was 21...4Jfd7 wins.) 37 b6 wins. 33 f4 wins as well.
22 4Ja3! 33 ... 4JbcS 34 4Jd4 eS
A subtle way of developing the
knight.
22 ... gxO 23 gxO 4Jfd7
After 23 ... 4Je8 24 4Jc4 (or 24 :cl
lhd4 25 -ltxc5+ t;t>d7 26 l:tc2 and
Black has nothing for the pawn.)
24 ... 4Jd6 25 4Jb6 l:tc7 26 l:tc 1 White
wins.
3S 4JxeS!
After 35 4Jf5! flxb5 36 :xh5 wins,
as Ftacnik pointed out.
3S ... 4Jxe5 36 1:[f5+ ~g7 37 lhe5
White has obtained a winning
position.
37 ... 4Jxe4 38 ~d3!
If 38 fxe4 Ilc3+.
244Jc4 38 .. J~c3
It is remarkable that the knight on c4 If 38 ... 4Jc3 39 ~d6 wins.
actually blocks the c-file, thus helping 39 -ltb4 flxd3+
Black to defend the c5-knight. Yet it Giving back the exchange in order to
paralyses Black's position. Fischer may prolong the game but really it changes
have underestimated the move. nothing else in the position.
24 ... -lta8 2S t;t>f2 l:tg8 26 h4 l:tc7 40 t;t>xd3 4Jf6 41 ~d6 flc8 42 l1gS+
27 4Jc2 l::tb8?! ~h7 43 ~e5 4Je8 44 lbh5+ t;t>g6
After 27 ... ~f6 28 -lta5 l:tb7 29 ~d2 4S %:tgS+ t;t>h7 46 -ltf4 f6 47 :f5 rt;g6
~e7 30 4Jd4 White has nice play for 48 b6 fld8 49 l:taS ~xO
the exchange.
28 ~a3! hS?
Giving up the g-file turns out to be a
huge mistake. After 2S ... l:tgS 29 b4
4Ja4 30 b5+ t;t>d8 31 b6 l:tc8 32 4J2e3
Black is in trouble.
29 flgl ~f6 30 t;t>e3
30 l:tg5 is met by 30 ... ~xe4!
30 ... aS 31 %:tgS a4?! 32 b4 4Jb7
33 bS
84
Boris Spassky the JOth
50 h5+! 1-0 This is not the best despite the fact

Black resigned as White can soon that piece play is fully in the spirit of
play b7 which wins. the Griinfeld Defence.
14 ... b5 looks more natural but my
It's time to look at the games opponent may have been worried that I
in which Spassky's 'instructive' play had analysed it all previously at home.
affected me. We start with his strategy Then 15 0, Spassky's idea to cut off
in the symmetrical Queens Gambit the b7-bishop in this pawn structure,
Accepted pawn structure. (other moves can be met satisfactorily,
e.g. 15 a4 b4; 15 ~f3 tUe5; 15 tUc6
G.Kasparov - P.Leko ~b7) 15 ... ~b7 16 tUb3 and White has
Fujitsu-Siemens Giants, a small edge as the knight soon reaches
Frankfurt 2000 a5. But then it will be difficult to decide
which small advantage to go for as any
1 d4 tUf6 2 tUf3 g6 3 c4 ~g7 4 tUc3 of them might prove to be more
d5 5 ifb3 dxc4 6 'ikxc4 0-0 7 e4 a6 significant than expected. Alternatively
The Hungarian variation, played by a 14 ... b6!? deserved some consideration.
Hungarian grandmaster. Pushing the b-pawn only one square
8 'ifb3 allows him to control more squares.
White makes a third move with the
queen. This subtle move order is the
result of experience.
8... e5 9 dxe5 'iVa5 10 'iVb6 'it'xb6
11 exb6 tUbd7 12 ~e2 tUxb6 13 ~e3
tUbd7
15 f3
In the queenless middle game I am
building up my pawn structure in the
same way that Spassky did. In my case
it did not bring the same result.
14 tUd4! 15... e5
An endgame has arisen which Or 15 ... tUe6 to get rid 0 f the
greatly resembles that mastered by dominating d4-knight and develop his
Boris Spassky in the Queens Gambit bishop on e6. Then 16 tUb3! and Black
Accepted. I follow Spassky's strategy. is still under pressure. His queens ide is
Let's see where it led me! vulnerable and he must also reckon
14•.. tUc5?! with e5. 16 ... ~d7 (16 ... b5 17 ~f2 ~d7
85
Boris Spassky the J Oth
[or 17 ... ~b7 18 ~a5] 18 l:Ihdl and

White has an edge.) 17 f4 ~c6 18 .i.f3
i..h6 19 g3 l:Ifd8 20 ~e2. White's space
advantage in the centre makes Black's
game difficult.
18 ~f2?!
At least an inaccuracy as it allows the
c8-bishop to move to e6. 18 .i.c4!
is a better move as it hampers the
opponent's bishop. Having the king on
e2 would be better and even queenside
16 ~c6!
castling might be possible. Another try
This is the only jump that causes
is 18 ~b6!? which would create some
headaches for Black - otherwise Black
confusion in Black's camp.
will just complete his development.
After 16 tiJf5 gxf5 17 ~xc5 l:td8 18 ... i..e6 19 %:thdl
18 .i.b6 lte8 19 l:Id 1 ~e6 and Black is 19 lthc 1 !? looks better. Then
doing all right. 19 .. J:td2 (19 ... ~d7 20 .i.e3 a5 [20 ... f5
If 16 tiJc2 tiJe6 17 0-0-0 b5 18 tiJd5 21 tiJa4] 21 tiJa4 ndc8 22 i..c4 and
~b7 the position is safe for Black. defending c6 is not going to be fun.)
16 ... bxc6 20 b3 .i.h6 21 ~e3 ~xe3+ 22 ~xe3
Leko takes on a pawn weakness. On and Black's rook is active on d2 but
the other hand the move considerably may soon come under pressure.
loosens White's grip. 19 ... ~d7! 20 .i.e3 .i.f8! 21 l:[d2 f5!
Black could think of sacrificing a 22 %:tadl ~e7! 23 g3
pawn instead with 16 ... tiJe6. Then
17 tiJxe5! (17 tiJe7+ ~h8 18 tiJa4!?)
17 ... ~xe4 18 tiJxg6 hxg6 (18 ... tiJxc3
19 tiJxf8 ~xf8 20 bxc3 ~xc3+ 21 ~f2
i..xa 1 22 l:txa 1 and the endgame is not
attractive for Black.) 19 fxe4 tiJd4 and
Black has some counterplay, though he
is a pawn down.
17 ~xc5 l:td8
86
23 ... <i!rf7!
Black has improved his pieces. Now
he has play of his own.
24 b3 as 25 ':c2 lbf6!
Black has certainly played very
purposefully over the last six moves
and managed to equalise.
26 %lxd8 l::txd8 27 exfS
Defending the e4-pawn would not
leave White very much scope for 39 l:te2??
action. After 39 if.g6? Ith4+ 40 <i!rg7 l:tg4
27... gxfS 28 lba4 if.dS 29 if.b6 ':a8 the pin is lethal.
Or 39 :£2 llh4+ 40 r;J:;g7 l:lxh2
30 if.cs lbd7 31 if.xe7 <ifa>xe7 32 <ifa>e3
41 l:txh2 lbxh2 42 <ifa>f6 and White
litd6
achieves a draw.
39 ... l:lh4+ 40 <i!rg7 lbxh2
The intennediate 40 ... lbd4! was
winning at once, e.g. 41 lIe8 lbe6+.
41 lbc3 lbo 42 lbe4+ ~c7 43 itJf6
lbd4 44 lbxdS+ cxdS 45 lId2 ~d6
46 if.d3?
I was already short of time in this
rapid game. After the text my bishop
lands in a losing pin. Maybe I was
angry that the strategy did not work as
33 if.d3?! well for me as it had done for Spassky
I had to win to have a chance of and that affected my concentration.
catching up with Anand who was Black could try playing on but White
can proba~ly hold with 46 if.b 1.
leading the event. Of course White
does not stand at all worse after
33 lbc3.
33 ... f4+!
Black seizes his chance to take the
initiative.
34 gxf4 exf4+ 35 ~xf4 l:f8+ 36 <ifa>gS
llJeS 37 if.xh7
After 37 if.e4 lbxf3+ 38 if.xf3 lhf3
39 llJc3 <ifa>e5 Black's king is somewhat
troublesome, yet White should be able 46 ... ttJe6+ 47 'litf6 l:H4+ 0-1
to live with it. White resigned as after 48 ~g6 :d4
37 ... lbxo+ 38 Wh6l:tf4! Black wins.
87
Boris Spassky the 10,h
B.Larsen - B.Spassky
We now look at how Spassky

used the h-file as a stunning
avenue for attack.
I knew these games and wanted

to hammer my opponents in the
same way along the h-file.
B.Spassky - J. van Oosterom G.Kasparov - V.Anand
It is worth seeing Spassky's games. flank. The alternative 11...f6 12 h4 ~f7

is depressing for Black but he can at
B.Spassky - J. van Oosterom
least last longer than in the game.
Junior World Championship,
12 ltJg3 bxc4
Antwerp 1955
1 d4 ltJf6 2 c4 g6 3 ltJc3 ~g7 4 e4 d6
5 f3 0-0 6 ~e3 eS 7 ltJge2 ltJc6 8 'ifd2
ltJd7
Black opens the diagonal for his
bishop. In fact this move is still played
competitively.
9 0-0-0 a6 10 dS ltJa7
10 ... ltJe7 looks more natural than
putting the knight on the edge.
11 g4 bS 13 h4
Black doesn't get enough play on the Spassky commences operations on
queenside and wastes time on the other the h-file.
88
Boris Spassky the JOth
13 ... f6? I also made a similar knight sacrifice

This buries the g7-bishop. After in my game against Chiburdanidze.
13 ... lOb5!? 14 ~xc4 liJd4 15 'iVa 17... gxfS 18 'ilfhS+ <t>g8 19 gxfS 1:[f7
White should do well - nevertheless 20 ~e2 lbcS?!
Black is still alive and kicking. This allows a forced checkmate, but
14 hS 'ike7? Black is completely lost anyway. If
Black played 13 ... f6 to make an 20 ... <ittf8 21 lldg 1 'ife8 (21...<ite8
escape route for the king, but now he 22l:txg7; 21 ... lbb5 22 'iVh8+ and White
delivers a nice checkmate.) 22 'ifh7!
blocks the path with this natural-
lbb5 23 l:hg7 (Capturing with the
looking yet losing move. Better
queen would be more elegant but this is
resistance was offered by 14 ... lIf7.
far more decisive.) 23 .. .l:txg7 24 ~h6
15 hxg6 hxg6
wins.
21 l:tdgl 'iWd7?!
Other moves would have lasted
longer but it's all the same now.
16 'ii'h2!
So natural. And what makes it even
nicer is that it wins directly.
22 'ilfh8 mate
16 ...~n
After 16 ... lbb6 17 'ii'h7+ <ifi>f7 18l:th6 The next game is probably Spassky's
g5 19 'ilfg6+ wins. most famous masterpiece and of course
I knew it well.
B.Larsen - B.Spassky
USSR v Rest of the World,
Belgrade 1970
1 b3 eS 2 ~b2 lbc6 3 c4 lbf6

4 lOf3?! e4 5 lOd4 ~cS 6 lUxc6 dxc6
7 e3 ~fS 8 'iVc2 'iVe7 9 ~e2 0-0-0
10 f4?
B lack is ahead in development, so
17 lOfS!! White has no time for this.
89
Boris Spassky the J Oth
14 ...:hl!!
Black sacrifices a whole rook for a

decisive tempo.
15 lbhl g2 16 :O?!
In a hopelessly lost position White
walks into a checkmate. But if 16l::tgl
'int4+ 17 ~dl 'intI wins.
10 ... lbg4!
Spassky acts at once, before White 16 ...'ii'h4+
can bring his pieces into the game. Spassky's queen invades on the h-file
11 g3 hS! and checkmates White.
To open the kingside. 17 <iifdl gxO ='ii'+ 1-0
12 h3?
White could make only three more
White allows the opening of the
kingside. After 12 h4 f6 Black stands moves.
better anyway.
We saw how effectively Spassky
used the h-file for attack. I was not
careful enough when I tried to emulate
his play. This game was unfortunate
indeed!
G.Kasparov - V.Anand
Reggio Emilia 1992
1 e4 e6 2 d4 dS 3 It)d2 eS 4 exdS
12 ... h4! "xdS 5 dxeS
Black opens the h-file at the cost of a The first sign that White may castle
piece - which turns out to be a low
queens ide.
priced but highly fruitful investment.
S...1.xeS 6 lbgf3 It)f6 7 1.d3 0-0
13 hxg4 hxg3 14 ~gl
8~e2
White persists with his idea of
castling long.
8 ... lt)bd7 9 It)e4 b6!
Black would rather give up the
two bishops than fall behind in
development.
10 It)xeS 'it'xeS! 11 ~e3 iie7
12 ~d4 ~b7
90
Boris Spasslo/thittIO~A"
130-0-0 17 ...•xa2!
It has taken some effort but at last Black is confident that he has got his
White can consider launching a snap bearings rig.ht in this complicated
attack on the kingside. position.
13 ... 4Jc5! 14 i..e5 18 i..xf6
After 14 i..xf6 White can't double If 18 l:he4? 4Jxe4 19 'ii'xe4 'ii'al +
Black's pawns because of 14 ... 4Jxd3+ 20 !if;d2 'it'xhi 21 'iVg4 f6 wins.
15 l:txd3 'ii'f4+. 18 ... i..g6!
14 ... 4Jxd3+ 15 l:txd3?! 'ii'c4! Black can't win the rook with
Grabbing a counetrattacking chance, 18 ... 1i'al +? as White's attack has
offered by the unprotected state of the grown too strong. 19 rj;d2 1i'xh 1
a2-pawn. 20 l:txe4 gxf6 21 'ii'g4+ rj;h8 22 'ii'h4
164Jd4?! l:[g8 23 'ii'xf6+ l::tg7 24 l::tg4 l:tag8
Going for a slightly worse but 25 4Jf3 and quite incredibly Black is
tenable ending with 16 i..xf6 was defenceless.
White's best option. After 16 ...•f4+ 19 l::ta3 'ii'd5 20 h4
17 ~b 1 "'xf6 White can live with this After 20 i..e5 f6 21 Jig3! 'ii'xd4
position. But not 16 ~b 1?! i..e4 17 lle3 22 'ii'xe6+ the position is equal
'iVxe2 18 l:txe2 i..xf3 and the doubled according to Ernst.
pawns make it a really tough endgame 20 ... gxf6 21 h5 'ii'xd4 22 hxg6 hxg6
for White.
16... ~e4
After 16 ...... xa2 17 i..xf6 gxf6
18 "'g4+ <it>h8 19 'iWh4 'iVal+ 20 ~d2
'iVa5+ 21 l:.c3 'ii'g5+ 22 'iVxg5 fxg5
23 lic7 White has some compensation
for the pawn.
17 lie3
91
23 ~ah3 A sad moment for White, who gives

See diagram on page 88. up his hopes of creating an effective
This is the position I was hoping for. attack on the h-file.
Spassky's raids on the h-file were very 28 ...'fie5 29 g3 'iie1+ 30 ~c2 licd8
much the models I had in my mind. He 31l:td4 'iie5 32 l:thf4 'iic7 33 'ii'e3 e5!
succeeded with this theme several 34 lhd8 l1xd8 35 l:Ie4
times. 35 lih4 would allow no more than a
23 ... f5 24 ~h4 check on the h-file. Black replies
35 ... 'iid6.
35 ... l:td5
Black gradually makes progress with
his extra pawn.
36 g4 b5 37 g5 ~d6 38 f3
24 ... f4!
Black correctly keeps his queen in
the centre where it can easily deny its
counterpart access via the h-file. If
instead 24 .. .'ii'f6 25 'ii'e3!? l1fd8
26 'ii'h3 c;!(f8 27 ~h8+ c;!(e7 28 ~a3+
'lifd7 29 ltd 1+ and White has an attack. 38 ... a5!
25 'iWf3 Ibc8 26 ~xf4 Now he starts opening up White's
The queen can reach the h-file with king.
26 'iib3, but it would all be in vain. 39 ~e2 'ife6 40 'ifh2
26 ... 'ii'xf2! 27 l:r.h8+ ~.g7 28 l:th7+ A useless demonstration on the h-file.
(28 'iWh6+ 'liff6) 28 ... ~f6 29 c3 'ii'e3+ 40 .•."f5 41 'ifg3 'ii'd7 42 'ife1 b4
and Black wins. 43 cxb4
26 ...'ii'c5 27 c3 ~g7
Upon 43 l:r.xe5 ~a4+ 44 <it>c 1 bxc3!
wins.
28l:thh4
92
43 ...'iVa4+ 44 b3
This is particularly annoying as it is After 44 ~c3 Ernst gives 44 ... 'ifc6+
Black who creates a winning attack 45 :'c4 axb4+ 46 ~xb4 ltb5+ 47 <itc3
from an edge file. This was exactly 'ifxf3+ wins.
what I wanted to do to my opponent 44 ..•'ifa2+ 45 ~c3 a4 46 bxa4 'iVa3+
and now I get checkmated in the same 47 ~c2 'ifxa4+ 48 ~c3 'iVa3+ 49 <ittc2
way. So not only Spassky but also lld30-1
Anand gets it right - just not me!
93
Tigran Petrosian the 9 th
Petrosian's wish came true in 1963 challenged. Tigran is in many ways the
when he defeated Botvinnik. The flrst closest to me, as he is from a Caucasus
Soviet world champion was not republic. He spent his childhood in
afforded the privilege of a rematch, so Tbilisi, Georgia, just a few hundred
Petrosian enjoyed a complete three kilometres away from Baku where I
year cycle as Champion before being grew up.
In which particular way did Tigran Vartanovich affect me? He is known for
his exchange sacrifices. I also tried them a few times.
Here are a couple of his exchange sacrifices which I had in mind when I
made my own.
S.Resbevsky - T.Petrosian A.Yusupov - G.Kasparov
25 ... 'Ae6!! 14... gxf5 15 ~h5 VJic8 16 iLe7 lIe8
M. Tal - T.Petrosian J.Timmao - G.Kasparov
31 ... ~4!! 16... 'Ah4
94
Tigran Petros ian the 9rh
J. Van der Wiel- G.Kasparov
I lost all these games after

sacrificing the exchange.
Let's look deeper into them

and see what went wrong!
24 ... g4
S.Reshevsky - T.Petrosian 14 'ire2 c4 15 .lic2 b5 16 e4 ~e6

Candidates tournament, 17 'irel 4Jd7 18 'irg3 f6 19 .lif4 %:[f7
Zurich 1953 20 l:tfe 1 4JiE 21 ~d6 ttd8 22 ~c5
lh-th Rabar-Petrosian, Belgrade 1954.
1 d4 4Jf6 2 e4 e6 3 4Jc3 ~b4 4 e3 White obtained some advantage and
0-0 5 ~d3 d5 6 4JO e5 7 0-0 4Je6 8 a3 probably convinced Petrosian that he
~xe3 9 bxe3 b6 should switch to another Nirnzo-Indian
This is a relatively rarely played line. line.) 1411Vel 4Jd7 15 e4 c4 16 J.c2 f5
Petrosian adopted it four times in 1953 17 e5 ltf7 18 a4 a5 19 f4 b5 20 axb5
and 1954, then he stopped using it. 'ii'xb5 21 ~a3 4Jb6 22 'ii'h4 'ii'e8
Before him, Keres was the only great 23 %:[f3 4Jc8 24 ~a4
player to employ this line. For example,
he drew against Alekhine with it in
1938. Petrosian used this move for the
first time in this game.
10 exd5 exd5 11 ~b2
At the 1953 Candidates tournament
in Switzerland, Petrosian played the
line twice more. In the first two games
the bishop move was played. In a third
game at Zurich, Taimanov beat him
after 11 4Je5 when, interestingly, 24 ... l:td7 (24 ... ~d7 25 e6 wins.)
Petros ian tried a similar exchange 25 l:lb 1 'ifd8 26 ~xd7 'ifxd7 27 Itg3
sacrifice but this time it didn't work 4Ja7 28 ~e7 il.f7 29 'tIVg5 J.g6 30 h4
just as in my games! 11...'iVc7 4Jc6 31 .i.a3 4Jd8 32 h5 ltJe6 33 'iVh4
12 4Jxc6 1i'xc6 13 f3 ~e6 (after 13 ... a5 iJ..f7 34 h6 g6 35 'iff6 lid8 36 iLe7
95
TIgran Petros ian the 9th
'ii'c7 37 l:txg6+ hxg6 38 h7+ 'it>xh7 22 ~g4

39 'i¥xn+ !Dg7 40 ~f2 1-0 Taimanov- By inserting this move he announces
Petrosian, Zurich 1953 his intention of pushing the e-pawn.
11. .. c4 12 ~c2 ~g4 Alternatives were 22 h4!? and 22 lle3!?
Reshevsky played 12 ... !De7 In a 22 .....e8 23 e5
match against Najdorf the same year. This is menacing since White can
He drew one and lost one out of these open up the position with e6. However
games. it gives up the d5-square and the bishop
13 'ii'e1 !De4 14 !Dd2 tDxd2 on b2 is out of play.
A Soviet player Ababkarov played 23 ... a5 24 lIe3 l%d8 25 l%fel
14 ... ~f5 twice in 1957, interestingly he Crouch recommends 25 l:tef3 as it
won both games. prevents 25 ... f6.
IS 'iVxd2 .Ji.h516 f3 .Ji.g617 e4 'ji'd7 See diagram on page 94.
18 l:tae 1 dxe4 25 .. J~e6!!
A couple of rounds later Petrosian Black blocks the e6-thrust, and
diverged and played what was perhaps at the same time Petros ian clears the
his most famous game. Against e7-square.
Smyslov he went 18 ... f5 and the game 26 a4?!
continued 19 exd5 tr'xd5 20 a4 lHe8 In the Predecessor series I indicated a
21 '6'g5 'VJkn 22 .Ji.a3 h6 23 'VJkg3 lhel preference for the immediate capture
24 l:txe 1 l:te8 25 nxe8+ tr'xe8 26 'it>f2 26 iLxe6. Then 26 ... 'ihe6! (26 ... fxe6
tDa5 27 tr'f4 tDb3 28 .i.xf5 ~xf5 27 l::tg3 !De7 28 lIfl !Dd5 29 tr'g5 'ike7
29 'ii'xf5 'iWxa4 30 'iWc8+ 'it>h7 31 'ii'f5+ [29 .. Jld7 30 h4] 30 .i.cl 'iVxg5
~g8 32 'ii'e6+ ~h7 33 'ii'e4+ 'it>g8 31 .i.xg5 lIb8 32 .i.d2 and White has
34 ~a8+ 'it>h7 35 'ii'e4+ 'it>g8 36 'i¥d5+ an advantage in the endgame.) 27 J:.g3
'it>h7 37 iLe7 !Dc 1 38 'ii'f5+ 'it>g8 !De7 28 h4 !Dd5 29 'iVg5 lId7 30 h5 h6
39 1i'f8+ 'it>h7 40 'ii'f5+ <i!7g8 41 d5 31 'ifh4 .i.d3 and Black has a
'ii'a2+ 42 ~g3 'ii'd2 43 d6 'iVel + reasonable fortress. If 32 iLc 1 tDxc3.
44 ~g4 !Dd3 45 'iWd5+ ~h7 46 d7 'iWe5
47 'ii'xd3+ cxd3 48 d8='ii' '12-112
Smyslov-Petrosian, Zurich 1953.
19 fxe4l:tfe8 20 'ii'f4 b5 21 iLdll1e7
26 ...!De7!
Tigran Vartanovich radically
improves the position of his knight.
27 .i.xe6 fxe6 28 ..n
96
After 28 ':f3 b4! 29 1:.en ~d5 8 e3 d6 9 h3 If::)aS 1 0 ~e2 eS 11 d4

30 'ii'g5 ':b8 Black stands well as I 'ike7 12 ~bd2 ~d7
pointed out in the above-mentioned Petrosian played this line seven
analysis. times, holding three world champions.
28 ... ~dS! Both Karpov and I were unable to hurt
There are no open files for the rooks him.
and both Black's minor pieces have 13 lLlfl lLlc4 14 lLle3
wonderful play. 14 b3 is the main line.
29 1:.f3 ~d3 30 ':xd3 14 ... lLlxe3 15 .Jtxe3 ~e6?!
White gives back the exchange, but Players no longer put the bishop on
Black's knight remains very strong and e6 nowadays it's c6. Tigran drew
easily compensates for the pawn against Karpov in Milan 1975 with
deficit. 15 .. .lHc8.
30 ... exd3 31 'it'xd3 b4 32 exb4 16lf::)d2 %:tfe8 17 f4 Aad8 18 fxeS
If 32 c4 lLlb6. If 18 f5 exd4! or 18 d5 exf4.
32 ... axb4 33 as 18... dxeS 19 dS i.d7 20 c4 l:tb8
21 a4 b4
33 .. .1138
Even though he is a pawn down, 22 as!
Black does not even stand worse. Tal wants to exchange the light-
341:al 'ilic6 35 ~el 'ikc7 36 a6 'ifb6 squared bishop in order to remove an
37 ~d2 important defensive piece.
If 37 h3 lLlc7 picks up the pawn, as 22 ...1:.f8 23 ~a4 ~xa4 24 %ha4
pointed out by Crouch.
37 ... b3 38 'iVc4 h6 39 h3 b2 40 J:[bl
<ith8 41 ~el th-th
M. Tal - T.Petrosian
USSR 1958
1 e4 eS 2 lLlf3 ~c6 3 ~bS a6 4 ~a4

lLlf6 5 0-0 ~e7 6 ltel bS 7 ~b3 0-0
97
24 .. Jtbd8! If 36 h6 f5.
Black improves the rook, but his 36 ... h6 37 lbel?
position remains troublesome. After 37 b3!? c4 38 4Jxd6 (38 bxc4
25 'iff3 %:td6 26 4Jb3 4Jd7 27 :aal .i(.b8) 38 ... 'it'xd6 39 l:tael f6 40 bxc4
l:tg6 28 l:to ~d6 29 h4 'ii'd8 30 h5 :f6 l:tc8 41 l%e4 'it'c5+ 42 'it>h2 1Vxa5
31 'ifg4 43 'it'f5 the position is unclear.
See diagram on page 94. 37 ... ~b8 38 l%dl
31..J:tf4!!
A great saving concept, Petrosian
sacrifices the exchange for a blockade.
38 ... c4 39 d6 4Jd3 40 'it'g4?

Better was 40 1i'e3! f5 (40 ... 4Jxb2
41 lId5! 'ifd7 42 4Jc5 1Vc6 43 lHf5!)
32 ~xf4?! 41 'ir'd4 with a still unclear position.
Even such a broad-minded player as 40 ... ~a7+ 41 'it>hl f5 42 4Jf6+! 'it>h8
Tal could not resist taking the 43 'ifxc4 4Jxb2 44 'ifxa6 tiJxdl
exchange. Tal could have retained an 45 'ii'xa7
advantage by playing 32 l:txf4. Then Tal was not yet prepared to defend
32 ... exf4 33 ~xf4 "iJie7 (After other passively with 45 l%xdl I?, but maybe it
moves White will stand clearly better, was the better option as after 45 ... ~b8
e.g. 33 ... ~xf4 34 'ir'xf4 'ir'e7 35 h6 g6 46 4Jd5 .i(.xd6 47 4Je7 'ir'xe7 48 1i'xd6
36 l:tfl; 33 ... 4Jf6 34 'iff3 ~xf4 'ilr'xd6 49 l:txd6 %:ta8 50 l:tb6 l:txa5
35 'ilr'xf4 ti)xh5 36 'fie3; 33 ... 4Je5 51 %:txb4 White probably holds.
34 'iVg3 %:te8 35 h6 g6 36 %:tel 'ilr'c7) 45 ... 'it'xd6?
and Black can either try to blockade or Petrosian misses a win. Black can
break out. There might follow 34 h6 g6 force the issue after 45 ... 4Jc3! 46 'ife7
35 'ilr'g3 .i(.xf4 36 'ilr'xf4 f5 37 lIel. gxf6 47 %:txf5 'ii'xe7 48 dxe7 l:.e8
32 ... exf4 33 4Jd2 4Je5 34 'ir'xf4? 49 lhf6 'it>g7 50 l:tb6 (50 %:tg6+ ~f7
White had two moves to obtain the 51 %:txh6 4Ja4!! wins.) 50 .. Jhe7
better prospects. Either 34 'iih3 'ir'g5, 51 l:txb4 ne5 52 g4 nxa5 53 %:tb7+ 'it>f6
or 34 'ir'f5 g6 35 'ilr'h3. 54 l:tb6+ 'it>g5 55 'i!7g2 ti)e4!
34 ... 4Jxc4 35 e5 4Jxe5 36 4Je4 46 'ifd7 'ifxf6
98
'iVf6 69 'i!Ve4+ ~g8 70 'iVe8+ 'ii'fB

71 'iVxfB+ ~xfB 72 ~g4 ~n 73 ~f5
1/z-Yz
Here is my game against Timman.
J.Timman - G.Kasparov
Tilburg 1981
1 d4 lLlf6 2 c4 g6 3 g3 iLg7 4 .tg2

0-0 5 ltJf3 d6 6 0-0 c5 7 lLlc3 lLlc6 8 d5
47 'ifxd1 :b8 48 IU'3? ltJa5 9 lLld2 a6 10 'iVc2 l:tb8
He should play 48 'iVd3 b3 49 a6.
A deviation from this line
48 .. J:ta8?
with 10 ... e5 resulted in Tigran's
After 48 ... .c.b5 49 'iVe 1 '1fi>h7 50 l:tb3
Vartanovich's most beautiful and most
l:txa5 51 'iVxb4:a I + 52 Wh2 :fl wins.
famous combination from the 1966
49 'ii'e1 :xa5 50 'iVxb4 l:te5 51 'ii'f4
World Championship final match
'1fi>h7 52 <1th2 ':d5 53 :n 'iVg5 54 'iVf3
against Spassky. Here is how that game
:e5 55 ~g1 ':c5 56 'iVfl l:te5 57 'iVf3
went: 11 b3 ltJg4 12 e4 f5 13 exf5 gxf5
:a5 58 <1th2 ~h8 59 IiPg1 l:ta2
14 ltJdl b5 15 t] e4 16 iLb2 exD
60 'ifd5?
17 .txt] .txb2 18 'ii'xb2 ltJe5 19 .te2
f4 20 gxf4 iLh3 21 ltJe3 iLxfl 22 llxfl
ltJg6 23 iLg4 ltJxf4 24 lhf4 l:txf4
25 .te6+ l:If7 26 ltJe4 'ii'h4 27 ltJxd6
'iVg5+ 28 ~hl :aa7 29 iLxf7+ l:txf7
60 ....c.c2?
60 ... 'ii'e3+!! would have been
decisive. Then 61 '1fi>h2 :a4 62 'ii'd8+
~h7 63 ltxf5 l:td4 64 :d5 :g4 65 IId3
'it'e5+ 66 c;!tgl 'iVel + (66 ... 'ii'e4 67 'ii'd5
30 'ifh8+!! 1-0 Petrosian-Spassky,
':xg2+ 68 c;!th 1 'it'xd5 69 :xd5 :g5 Moscow 1966.
Black wins.) 67 ~h2 l:th4+ 68 l:lh3
'it'e5+ 69 ~gl I:td4 and Black catches 11 b3 b5 12 .tb2 bxc4 13 bxc4 i.h6
White's king. In the databases Donner is credited
61 'ii'a8+ ~h7 62 'iif3 1Ic1 63 l%xc1 with playing this for the first time
'it'xc1 + 64 c;!th2 'i'c7+ 65 ~h3 'it'e5 against Botvinni!<, way back in 1958.
66 g4 fxg4+ 67 ~xg4 'ii'g5+ 68 ~h3 14 ltJcb1
99
Tigran Petros ian the 9 th
According to the database this move did not offer this exchange sacrifice at
was introduced by Udovcic. It was then once. I lost a game against Zaid after
seen regularly, although it has never playing 16 ... iLg7. I have not yet
been played as often as 14 f4. Petrosian decided which world champion I
tried 14 f4 unsuccessfully with White should blame for that. 17 lIabl "ike7
against his problem opponent in the 18 e4 h5 19 f4 lIb4 20 "ild3 ttJb7
sixties. 14 ... e5 15 l:lae 1 (Petrosian 21 ttJc2 lixbl 22 ':xbl h4?!
deviated from Portisch's play when he (22 ... exf4!?) 23 fxe5 dxe5 24 ttJf3! and
faced the young Chiburdanidze by White's centre pawns will be strong.
playing 15 dxe6 fxe6 16 l:lab 1 .i.g7 24 ... hxg3 25 iLxe5 gx.h2+ 26 iLxhl
17 ttJce4 lIxb2 18 ltxb2 ttJg4 19 l:lbb 1 "ilc8 27 ttJe3 ttJg4 28 ttJxg4 .i.xg4
~d4+ 20 ~hl ttJe3 21 'fie 1 .i.b7 29 ttJe5 ttJa5 30 lin .i.h5 31 d6 iLxe5
22 ~f3 ttJxfl 23 ~xf1 .i.xe4 24 ttJxe4 32 iLxe5 "ike6 33 iLf6 ttJc6 34 "ike3
"ile7 25 ttJg5 ~h8 26 "it'h3 e5 27 ttJe6 ~h7 35 'ii'xc5 ttJb8 36 iLe7 lte8
lie8 28 it.d5 exf4 29 ttJxf4 "ilf6 30 "ild7 37 "ikd5 1-0 Zaid-Kasparov, Leningrad
'ii'd8 31 "ilxd8 ltxd8 32 ltb6 .i.e3 1977.
33 ttJe6 1-0 Petrosian-Chiburdanidze, 17 iLxb4 cxb4 18 ttJabl
Vilnius 1978) 15 ... exf4 16 gx.f4 ttJh5 White's rooks have no open files,
17 e3 lie8 18 ttJce4 .i.f5 19 .i.e3 ttJb7 therefore the sacrifice comes into
20 "ila4 a5 21 ltbl 'ii'e7 22 l:lfel .i.d7 consideration. However, Black doesn't
23 'ilVe2 .i.f5 24 "ila4 ~f8 25 ltb6 ltbd8 even have a pawn for it.
26 "ilb3 .i.c8 27 ttJfl ltd7 28 ttJfg3
ttJxg3 29 hxg3 iLg7 30 'iib2 f5
31 iLxg7+ "ilxg7 32 ttJf6 1-0 Portisch-
Petrosian, Santa Monica 1966.
14 ... e5
In the debut game, Suetin preferred
14 ... iLd7 which then became routine.
15 iLc3 iLd7 16 ttJa3
16... ltb4
I got excited when I read about 18 .• .'iWc7
exchange sacrifices in a chapter in Petrosian's opponent followed up the
Petrosian's book. What advantages exchange sacrifice with 18 ... 'iib6. Then
does Black accrue with this exchange? came 19 ttJb3 ttJb7 20 ttJld2l1c8 21 a3
The position is closed so the rooks do a5 22 axb4 a4 (22 .. .'~i'xb4!?) 23 ttJa5
not work well. In addition the c5- ttJxa5 24 bxa5 'ii'xa5 25 e3 iLg7 26 lIa2
square is firmly under Black's control ttJe8 27 ltfal lIa8 28 e4 ~c5 29 'iVc3
and he has an outside passed pawn. I lic8 30 iLf3 iLh6 31 ltb 1 iLxd2
did not pay attention to the interesting 32 'ilVxd2 "1Wxc4 33 lIb4 'ilVcl + 34 ~xe 1
fact that Petrosian himself had opted nxcl + 35 ~g2 ~g7 36 iLe2 ttJf6 37 f3
for this position. In my younger days I g5 38 g4 h5 39 h3 hxg4 40 hxg4 ttJg8
100
TIgran Petrosian the 9th
41 l:r.b6 1-0 Petrosian-Toran, Bamberg 22 ... b3 23 4Jxb3 ~a4 It looks like

1968. White can live with this pin. 24 ttJ 1d2
19 e3 :b8 25 l:lab 1 4Jxb3 26 4Jxb3 'ifb6
Two years later Kurajica tried an 27 c5 .i.xb3 (27 ... dxc5 28 'ii'xc5 ~xb3
interesting idea, he gave up the c-pawn 29 'ii'xb6 lhb6 30 l:r.ecl and White
with 19 c5 in order to open the file. soon invades on the queenside.)
19 ... 'ii'xc5 20 'ifb2 4Jg4 (20 ...1fb5 !?) 28 'ifxf5 gxf5 29 cxb6 ':xb6 30 ':ec 1
21 4Je4 'ifb6 22 ~f3 ~g7 23 4Jbd2 and White is better.
4Jh6 24 l:tab 1 f5 25 'S'xb4 White is 20 4Je4 ~xe4 21 .i.xe4 tUb7
ready to give up a piece to open files In a closed position one usually has
for his rooks. 25 ... 'ii'xb4 26 l:lxb4 fxe4 time to manoeuvre, however Black
27 4Jxe4 ttJf5 28 l:c 1 ttJd4 29 ~g2 now lacks just one move to obtain a
~b5 30 l:.c7 ~f6 31 ttJxf6+ l:lxf6 32 a4 favourable setup. After 21 ... ttJxe4
~xe2 33 ~xe2 4Jxe2 34 :b6 4Jd4 22 'iVxe4 f5 23 'iVc2 'iVxc4 24 'ii'a4 'ii'c5
35 l:xa6 ttJab3 36 a5 l:ln 37 l:la8+ rj;g7 25 tUd2 f4 26 exf4 exf4 27 l:r.ad 1 White
38 l:taa7 1-0 Kurajica-Filipovic, Banja has the better prospects. And if
Luka 1983. 21 ... ihc4 22 'iixc4 4Jxc4 23 ~d3
19 ... i.f5 White retains an edge.
I wanted to retain the knight for use 22 ttJd2 ttJc5
against the bishop. The alternative
19 .. .l:tc8 has not disappeared from
grandmaster practice and here is a
recent example of it. 20 a3 b3 21 4Jxb3
~a4 22 ttJ 1d2 4Jxb3 23 ttJxb3 :b8
24 lhb 1 'iib6 25 :02 ttJg4 26 'iVe2
4Jxe3 27 fxe3 ~xb3 28 l:tfb 1 ~xe3+
29 <ithl ~d4 30 c5 ~xc5 31 'iVdJ .i.c4
32 .c.xb6 l:lxb6 33 'iidl l:lxbl 34 'iVxbl
~xa3 35 'iib8+ rJ;g7 36 'iib7 ~c5
37 g4 ~d3 38 g5 a5 39 h4 a4 40 ~h3 23 ~g2!
~f8 41 'ilfa8+ rJ;g7 42 'ilfxa4 h6 43 ~e6 Interestingly this move fights for the
hxg5 44 hxg5 1-0 Psakhis-Avrukh, c5-square. Chess can be stunning,
Israel 2001. indeed. The bishop will cut Black's
Maybe gaining space in the centre by kingside knight off from the queenside.
19 ... ttJg4!? is after all a reasonable If 23 a3? b3.
option. 20 l:r.el f5 21 h3 (21 4Jb3?! is 23 .. J:tb8 241:trb1 a5 25 a3!
Timman's recommendation but Black Weakening b4 and opening the a-file
has an aggressive and good reply in for the rook.
21...f4!) 21...4Jf6 22 a3 (22 c5 e4) 25 ... e4 26 axb4 axb4 27 .th3!
101
A subtle move which cuts off the I sacrificed the exchange, just like
f6-knight. Petrosian. But after some mutual
27 ... ~g7 28 l:ta2 h5 29 ttJb3 ttJd3 mistakes I went down to Artur.
30 %:tdl ttJe5 17 ~xe8 'iWxe8 18 ~b4 e4 19 'iWc2
'iWh5 20 ~g3 '::US 21 ~f4 'iWg4 22 g3
ttJg5 23 'it>hl ttJf3 24 l::tac1 ~c5
25 ttJxf3 'ii'xO+ 26 ~gl ttJd3 27 'iWd2
~d4
31 c5!
Black loses an important component
of his compensation for the exchange.
He relinquishes the c5 post for his
knight. 28 l%c2 ~b7 29 b3 lIg8 30 ~b2 'iWh5
31. .. ttJd3 32 cxd6 "ii'xd6 33 ~f1 ttJe5 31 tLldl ttJe5 32 0 tLld3 33 llJe3 ~xf4
34 fla6 'ii'd7 35 nxf6 34 gxf4 ~b6 35 ~f2 1ig6 36 l::te2 .ic5
35 d6!? was also attractive. 37 fxe4 fxe4 38 f5 'iWh5 39 %1d2l%g5
35 ... ~xf6 36 '1Wxe4 nc8?
Black gives up the pawn for free.
36 .. .'~a4 would still enable him to
continue resistance but in the end
White's extra pawn should prevail.
37 "ii'xb4
White is just winning with his two
extra pawns.
37 ... h4 38 'ii'f4 'it>g7 39 gxb4 1id6
40 ttJd2 1-0
40 ~f4 'iie8 41 tLlg4 1-0
A.Yusupov - G.Kasparov
World Cup, Barcelona 1989
J. Van der Wiel - G.Kasparov
1 ttJf3 ttJf6 2 c4 g6 3 ttJc3 ~g7 4 e4
World Under 16 Championship,
d6 5 d4 0-0 6 ~e2 e5 7 d5 as 8 ~g5
Wattignies 1976
h6 9 ~h4 ttJa6 10 ttJd2 1ie8 11 0-0
ttJh7 12 a3 ~d7 13 b3 f5 14 exf5
1 e4 c5 2 llJo d6 3 d4 cxd4 4 ~xd4
See diagram on page 94. tLlf6 5 tLlc3 a6 6 f4 "ii'c7 7 a4 g6 8 .id3
14 ... gxf5 15 ~h5 ~c8 16 ~e7 lIe8 ~g7 9 llJo ~g4
102
TIgran Petrosian the 9,h
It's more common to develop the 20 i.xg4 l%ad8

bishop on b7. After 20 ... g5 21 Wh 1 gxf4 22 i.xf4
10 iLe3 lbc6 11 0-0 0-0 12 'ii'el lbdxe5 23 l:tg 1 the position is unclear.
iLxfJ 13 :txfJ e6 14 'ir'h4 'ii'd8 21 l:tdl f5
The queen retreats to defend the king. Black wants to carry out g5 under
15 l:th3 h5 16 .lte2 better conditions than in the previous
line.
22 iLfJ!
Van der Wiel is alert. He stops g5.
22 ...l:tf7
If 22 ... g5 23 ltJxd5.
23 Whl!
He wastes no time and goes after the
weakness on g6.
23 ... g5
If 23 ... lbf8? 24 i.c5.
16 ... d5 24 ~h5!?
This is thematic. As White is See diagram on page 95.
attacking on the flank, Black switches 24 ... g4?
to the centre. Black could ease the Black gets some play for the
attack by getting rid of the strong queen exchange, and White's rooks have no
on h4 with 16 ... lbg4. Then 17 iLxg4 open files - however that factor can be
(17 'ii'xd8 l':tfxd8 18 iLxg4 hxg4 rectified. After 24 ... 1:.ffS 25 fxg5
19 l:tg3 f5 and Black has a nice game.) lbdxe5 (After 25 ... iLxe5 26 i.g6 'ifb4
17 ... 'ifxh4 18 1:[xh4 iLf6! 19 l:th3 bxg4 27 iLh7+ Wg7 28 g6 Black's king is in
20 l:tg3 ltJb4 and Black has a good danger.) 26 .tc5 (26 iLb6 :tc8 27 g6
endgame. And after 16 ... lbb4 17 l:lel ltJd7) 26 ... 'tixg5 27 i.xfS l:xf8 28 l:gl
ltJd7 18 'ife 1 'ii'c7 Black is doing well. ltJg4 29 'ii'e2 ltJce5 30 lbd 1 and Black
17 e5 ltJd7 has to work hard to keep his position
The queen can still be swapped by together.
means of 17 ... d4. Then after 18 l:td 1 25 iLxf7+ 'ifxf7 26 l:g3 'iYh5
ltJg4 19 i.xg4 'ii'xh4 20 1:[xh4 dxe3 27 qg,g2 ltJf8 28 ltJe2 ltJg6
21 i.f3 f6 Black has nothing to worry
about.
18 'tin 'ike7 19 g4 hxg4
Black can also stir up things with
19 ... 'iib4!? Then 20 gxh5 d4 21 i.el
dxc3 22 bxc3 'itb6 23 h6 'ii'xf2+
24 ~xf2 i.h8 25 h7+ rj;g7 26 f5
ltJdxe5 27 i.h6+ ~f6 28 iLxfS lhfS
29 fxg6 and Black has compensation
for the exchange.
103
29 h3! This mistake completely relaxes the

Interestingly, a similar formation pressure. White's king is no longer in
occurred in my game against another danger and the rooks start to work.
Dutchman, Timman, where I sacrificed Black should protect g4 with 32 ...l:tg8.
the exchange. In that game I had a Then 33 l::tdd3 (33 ~e2 it.f8) 33 ... Jif8
b-pawn (which has the same qualities 34 c3 it.e7 35 cxd4 liJb4 36l::tb3 b6 and
as the g-pawn). He undermined it with though Black's position is troublesome
it has not fallen apart.
a single move of an edge-pawn. With
33 "e2! ~xe2+ 34 liJxe2
the same result - he beat me too. Had
Black is just the exchange down.
Black stopped all that he would have
34 ... liJg6 35 l:bh3 ~f8 36 l::tb3 l:td7
had a decent position. 37 l':tbd3 it.c5 38 c3
29 ... liJh4+ 30 ~n rbn 31 liJgl Now John wins a pawn in addition to
White is not in a hurry to take on g4, the extra exchange. It's all over now.
which would free the f5-square. 38 ... it.a7 39 .i.e3 lId5 40 cxd4 liJb4
31. .. d4 32 it.c1 gxh3? 41 l:tb3 a5 42 ~d2 1-0
Petrosian also had one V.Cbeskovsky - G.Kasparov

remarkable game where he
allowed himself to be saddled
with a doubled f-pawn and it
has remained in my memory.
But when I took on the
responsibility of doubled f-
pawns, I lost, as you will see in
the following two examples. ,.
B.Gurgenidze - T.Petrosian A.Beliavsky - G.Kasparov
104
Tigran Petros ian the 9th
B.Gurgenidze - T.Petrosian 20 ... ~d8! 21 ':xe6

USSR Championship 1967 After 21 <iitfl 'iib5+; or 21 'ii'dl 'itb5
22 b3 ~a5 and Black wins.
1 e4 eS 2 100 d6 3 ~bS+ ~d7
4 ~xd7+ 'ii'xd7 S 0-0 lOf6 6 eS dxeS 21. .. fxe6 22 'iVxe8 ':xe8 23 l:te1
7 ltJxeS 'ife8 8 d4 e6 9 lOe3 lOe6 Black dissolves the doubled pawns,
10 lOxe6 'ii'xe6 11 ~gS 0-0-0 but the knight has a chance to force
matters and improve his situation.
23 ... rJ;;c7!
Just in time, the king lends a hand
thanks to the fact that the e I-rook is
unprotected.
24 lOg7 l:te7 2S lLlhS fS 26 h3 <iitd6
27 g4 fxg4 28 hxg4 eS 29 lOg3
Petrosian puts pressure on d4 at once.

12 ~xf6
White sacrifices a pawn in order to
create a doubled pawn in Black's camp.
12 •.. gxf6 13 dS exdS 14 'iio ':d6
It looks like Petrosian is defending
the f6-pawn with the rook.
ISl:lfe1 d4
The rook can recapture on c6, which 29 ... <iitdS
was the point of 14 ... l:td6. Black's active king prevents an
16 'ii'fS+ ~b8 17 ltJe4 effective blockade by the knight.
The knight's aim is to get to f5.
30 f3
17 ...l:te6
Petros ian organises his pieces while If 30 lOe4 ltg7 31 f3 h5 wins.
White's knight heads for f5. 30 ...:17 311:[0
18 ltJg3 ~e7 19 'ifhs 'iie8 20 ltJfS The natural king move 31 <iitg2
allows an exchange of bishop for
knight by 31 ... ~h4.
31. ..~gS 32 <itg2 bS
White has blocked the kings ide but
Black has a winning pawn majority on
the queenside.
33 ltJe4 ~e3 34 l:lhl h6 3S b3 l1rs
36 l:lbll1f4 37 a4 b4 38 l:thl
105
Tigran Petros ian the 9'h
18 d5!
18.•. exd5
With a pawn sacrifice, White blocks
the d5 square. I was not worried that
Petrosian had won with the doubled
pawns and without having any
knight.
19 ttJd4 fia6 20 <it>bl ~d6 21 fif3
~xf4 22 fixf4 ttJe5 23 'ii'f5+ <it>b8
38 ...1he4
The exchange sacrifice wins because
Black has too many pawns for White to
cope with.
39 fxe4+ <it>xe4 40 ltdl a5 41 <ii7g3
ltd5 42 l:[fl ~f4+ 43 <it>f3 c4 44 <it>e2
After 44 bxc4+ <it>xc4 45 <it>e4 <it>c3
46 ~f2 d3 47 cxd3 b3 48 d4 b2 wins.
44 ... cxb3 45 cxb3 <it>e4! 46 ltf3 ~g5
47 l:r.f7
If 47 ~d3 <it>d5.
47 ... d3+ 48 <it>dl 'it>d4 0-1
24 f4
Unlike Gurgenidze, Cheskovsky
V.Cheskovsky - G.Kasparov
manages to win back the pawn.
USSR Championship, Thilisi 1978
24 ... ttJd7?!
1 e4 c6 2 d4 d5 3 tiJd2 dxe4 4 tiJxe4
After 24 ... ttJc6 25 'ii'xf6 ttJxd4
~f5 5 tiJg3 ~g6 6 h4 h6 7 h5 ~h7
8 ttJf3 ttJd7 9 ~d3 ~xd3 10 'tWxd3 26 'ilxd4 White has a small edge.
'ilc7 11 ~d2 ttJgf6 12 0-0-0 e6 13 ttJe4 25 'tWxd5 ttJe5 26 'ii'e4 ttJg4?
0-0-0 14 g3 c5 15 ~f4 c4 He should try to enter a slightly
This was my novelty. It is an inferior rook ending by 26 ... ttJc6
ambitious move which aims to place a 27 ttJxc6+ 'ii'xc6 28 'tWxc6 bxc6.
knight on d5. 27'ii'e2
16 'iV e2 'if c6 17 ttJxf6 gxf6
White has obtained a better position.
27 .. :iWb6
If 27 .. Jlhg8 28 :1hel.
28 c3 f5 29 l:r.hel 'ii'c5 30 'ii'e7 'ii'xe7
31 ~xe7 ~he8
Black could try to hang on the
material by playing 31 ... l:r.hf8. After
32 l:.de I lld5 33 <it>c2 Black is rather
passive.
32 l1del l:be7 33 ltxe7
106
Tigran Petros ian the 9 'h
33 ..•tOf6 13 ... 'ii'xc3?!

Reducing the number of pawns on This position with 13 ... ltJbd7 14 c4
the board does not lead to salvation. has been played many times. 13 ... lOd5
After 33 ... f6 34 ~cl (34 a4!? l:td5 has not been tested. Black is treading a
35 ~a2) 34 .. J:td5 35 Iitd2 a6 36 ~e2 very narrow path but objectively the
Black is struggling. move might be playable. 14 .i.d2 lOxc3
34 :txn tOxh5 35 tOxf5 ltd3 15 ~xc3 (15 ~xh7+ cj;xh7 16 lOg5+
36 l:tf8+ ~c7 37 ltg8 ~d7 38 llg6 b5 ~g8! 17 'ifh5 tOe2+! 18 ct>h 1 [18 'ii'xe2
39 a3 ltdl + 40 ~a2 l%gl 41l:td6+ Q;c7 iDd7] 18 ... 'ii'c2 and Black is in the
42 :'xh6 tOxg3 43 tOd4 game.) 15 ...'it'xc3 16l:tdcl 'ii'a5 17 l:tc7
Black loses a second pawn after 'ii'd5 (17 ... ~d5 18 lOg5!) 18 ltac1 and
which his position is hopeless. White has compensation for the pawn.
43 ... tOe4 44 tOxb5+ <it>d7 45 ':h7+ 14 ~b2
~e6 46 ':xa7 If 14 ~g5 ltJbd7
It's all over now. 14 ...'it'e7
46 ... tOe5 47 tOd4+ cj;d6 48 tOf5+ After 14 ... 'ifb4 15 d5 ~xd5
~d5 49 tOe3+ lite4 50 ':e7 iDd3 (15 ... iDxd5? 16 lOg5 h6 17 'iVe5 lOf6
51 :e7+ 1-0 18 .i.h7+ <ith8 19 'iVxf6 wins.) 16 i..xf6
gxf6 17 tOd4 f5 (17 ... lOc6? 18 1i'g4+
A.Beliavsky - G.Kasparov ~h8 19 'ifh4 f5 20 .i.xf5 and White
Game 4, Candidates match, mates.) 18 .i.xf5 (18 'ii'e3 ttJd7
quarterfinal, Moscow 1983 [18 ... tOc6 19 .i.xf5±] 19 ~xf5 '1ti>h8)
18 ... exf5 19 tOxf5 'ir'e4 20 lOe7+ ~h8
1 d4 iDf6 2 c4 e6 3 iDe3 ~b4 4 e3 21 'ii'b2+ f6 22 iDxd5 and White stands
0-0 5 ~d3 c5 6 tOo d5 7 0-0 dxe4 much better. If 14 ...'iVc6 15 d5 exd5
8 ~xe4 exd4 9 exd4 b6 16 liJd4 'iVe8 17 'ii'd2 iDe4 18 i..xe4
This system of development is "iixe4 19 l:tel 'ifg4 20 l:te7 i..a6
Karpov's favourite variation. 21 'ii'c3 and White has tremendous
10 'ir'e2 .i.b7 11 l:tdl ~xe3 12 bxe3 compensation for the two pawns.
"iic713 ~d3 15 d5 .i.xd5 16 ~xf6 gxf6
107
Tzgran Petros ian the 9 th
Just like Petrosian, I did not mind I was happy with the doubled f-pawn.
having the doubled f-pawns. 21 ':c4 'ifd7?
I gave back one of the pawns but this
was an wmecessary concession. After
21...llJe7 22 l:th4 llJfS 23 .ttg4+ ~h8
24 'ifd3 llJe7 25 l:th4 llJg6 26 Ah5
Black is safe.
22 .J:[h4 'it'f5
17 W'e3! ~g7
Not 17 ... jtb7 18 tL'lgS!
18 l:tac1 tL'lc6
After 18 ... ~7 19 llJe5 llJd7 20 'ii'h3
f5 21 .lla6 'ifxa6 22llJxd7 Black's king
remains vulnerable.
19 .lle4? 23 lbd5 llJe5
After 19 tL'ld4 'iid6! (19 .. JHd8 After 23 ... 'iib 1+ 24 llJe 1 ~h8
surprisingly loses. 20 .llb5 'iib7 25 l:tdh5 l:tfe8 26 .J:[xh7+ 'it'xh7
21 'if g3+ ~f8 22 .llxc6 ~xc6 23 ~f4! 27 l:txh7+ ~xh7 28 'ii'h3+ ~g8
l:td7 [Other moves also lose. 23 ... -'.xg2 29 'ii'g4+ ~f8 30 llJc2 is dangerous
24 llJxe6+; 23 .. J::tac8 24 tL'lxe6+; according to Beliavsky. The knight
23 ... f5 24l::!el; 23 ... -'.d5 24l:tc7 jtb8 aims to get to f5 after creating a flight
25llJb5] 24 W'xf6 ~g8 25 h4!! [25 ltd3 square for his king with h4.
~e4] 25 ... .ltxg2 26 f3 and White wins.) 24 h3l:Ue8
20 llJxc6 (20 ~3 l:th8 21 llJxc6 .ltxc6) Not 24 ... llJxf3+? 25 gx:f3 and White's
20 ... .ltxc6 and Black probably heavy pieces catch the king.
survives. 25 llJd4 'tig6
After 19 .llbS! l:Ifd8 20 l:td3!? Not 25 ...1i'g5? 26 l:Ig4 winning.
(20 llJd4 also wins as it transposes 26 'ft'f4
to the line with 19 tL'ld4) 20 .. Jbc8 Not 26 f4? 'ft'b 1+ 2 7 ~h2 ttJg6
21 llJd4 'iie5 22 llJxc6 ~xe3 23 fxe3 28 'ii'g3 l:Ie 1.
ltd6 24 .lta4 a6 25 ltc2 and White 26 .. Jbd8 27 llJf5+ ~h8 28 lhd8
wins. l:bd8 29 'ft'e4 l:Ic8?
19...'i!fd6! 20 .ltxd5 If 29 .. J:tg8 30 g4. Alternatively
If 20 l:txc6? ,*xc6 21 ltxd5 exd5 29 ... 'ifg8! to free the g6-square for his
22 llJd4 iVa4! 23 .ltc2 'aVe8 and Black knight. 30 liJe7 (30 l:Ih6 tL'lg6; 30 f4
wins as Beliavsky pointed out. llJg6 31 ,*c6 l:td 1+ 32 <itth2 'iid8
20 ... exd5 33 liJh6 <ittg7 34 liJf5+ is a repetition.)
See diagram on page J 04. 30 ... 'iig7 31 :h5! and White has
108
compensation for the pawn. (31 llJf5 b 1) 33 f4 ltJxg4+ 34 hxg4 'ii'xg4

'iff8 32 llJe7 'ii'g7=; 31 ~h2 l:e8 32 35ltJe3! 'ikg7 (35 ... l:d2+ 36 'ithl 'fig7
llJd5 l:e6 and Black may be able to 37 ltJg4) 36 lLlg4 wins.
hold on.) 31...a6 (31...llJg6 32 ~h2; b2) 33 Ihf6 'ikc5 34 ltJh6 'itg7
31 ... l:e8 32 llJd5 1:te6 33 llJxf6. This is 35 l:tf5 'iVd6 36 ~g2 ~xh6 37 l:xe5
the point of putting the rook on h5; ..t>g7 and though Black's king is
White retains a small edge.) 32 ~h2 exposed, it is not easy to exploit the
and Black must be. careful. situation.
31 'ika8+ 'iVg8 32 'ii'xa7 Ihh4
33 ltJxh4 1i'g5 34 1!:Va8+ <t>g7 35 'ilVe4
30~h2
Not 30 f4? llJf3+. 35 ... h5??
30 .. J:tc4?! This is a bad blunder as Black drops
This looks active but it just drops a the knight. Better was 35 ... lLlg6
pawn. After 30 ... l:td8?! 31 g4 (31 f4 36 lLlf5+ ~g8 37 g3 'iWd2 when Black
lLlf3+!) 31...'ii'g8 32 l:h6 'ii'f8! doubtless has problems but he is still in
a) 32 ... l:d2? 33 lLle7 'fig7 34 'ii'a8+ the game.
wins; 36 lLlf5+ ~g6 37 llJe7+ ~h6 38 f4
b) 32 ... 'fig5? 1-0
Similarities in a rook versus bishop endgame - with rook's pawns too.
T.Petrosian - L.Aronin G.Kasparov - A. Yusupov
109
T.Petrosian - L.Aronin Black can resist.

USSR Team Championship, Riga 1954 32 ....i.d6 33 l:te2 .i.e5 34 ~bl l:tc4
35 l:td3 h5 36 b3 l:th4 37 h3 :f4
1 e4 c5 2 lDo d6 3 d4 cxd4 4 lDxd4 38 ~c2 ~c7 39 .:tg2 ~c6 40 ~dl
lDf6 5 lDc3 a6 6 i.g5 e6 7 'ito lDbd7 :fl+ 41 ~e2
80-0-0 'iJlc7 9 ~g3 h6 10 ~xf6 gxf6
This is a risky option.
11 ~bl lDb6 12 f4 i.d7
41. ..:f4
The rook is actively placed on the
fourth and causes much inconvenience.
42 %:tg8 as 43 nh8 %:th4 44 lta8 ~b6
13 ~h4! ~e7 14 ~h5 45 a4 1:[e4+ 46 <i!ff3 :f4+ 47 ~e2l1.e4+
Black has serious problems with his 48 %:te3 1:[h4 49 1:[c8 l:ld4 50 l:[d3 :e4+
special Rauzer pawn formation. 51 ~O 1:[f4+ 52 ~e2 l:te4+ 53 l:le3
14 .. J~h7 15 f5 e5? :d4 54 l:tc2 l%h4 55 lI.d3 lI.e4+ 56 ~f3
Alternatively 15 ... 0-0-0 16 fxe6 fxe6 lH4+ 57 'it>g2 %:te4 58 ~O l:tf4+
17 ~3 ~b8 18 .i.e2 and White's 59 ~e2 l:te4+ 60 lI.e3 1::tf4
advantage is smaller than in the game.
16 lDe6 'it'c6 17 ~g4 .i.f8 18 'i\fg8
fIe6 19 ii'xh7
White has won the exchange, and in
return Black has very little.
19 ... 0-0-0 20 fIe6 .i.xe6 21 .i.e2 d5
22 exd5 lDxd5 23 lDxd5 .i.xd5
24 1:[hfl?
Best was 24 .i.g4+! ~b8 25 .i.f3
winning. 61 lIc4
24 ... ~b8 25 ii'f5 i.e7 26 .i.o e4 Finally he removes Black's rook.
27 i.e2 1:[c8 28 'ii'xd5 'iJlxc2+ 29 ~al 61. ..1:[xc4 62 bxc4 'it>c5 63 lI.b3
'itxe2 30 l:lfel 1Vxg2 31 'ihe4 'iixe4 'it>xc4 64 lIxb7 h4 65 1:[b5 .i.g3
32 l:lxe4 66lha5
110
Tigran Petros ian the 9tia
Petros ian won even though his

opponent had an additional f-pawn. I
had every reason to be optimistic.
81. ...i.f2
The forthright 81...f3 82 l::te8 f2
83 11£8 sets up a zugzwang.
82 IIg2 .i.g3
66 ...cwtb4
Black is likely to take the a-pawn and
with the h4-pawn he has chances to
draw even with the exchange deficit.
67 1:ta7 f5 68 CiPfJ f4 69 :a8 ~a3
70 ~g2 ~b4 71 %.ta6 cwta3 72 :a7 ~b4
73 'itfJ <ita3 74 ~e4
83 l:b2!
I recalled that Tigran Vartanovich
had won this endgame.
83 ... ~a8 84 ':b7 i.f2
After 84 ... f3 85 l::td7 ~b8 86 lth7 f2
87 It£7 wins.
Alternatively, 84 ... i.h2 85 ~c7 f3+
86 ~c8 f2 87 ltb2 1;a7 88 l::txf2 i.g3
89 l:b2! White must keep Black's king
in the comer. 89 lH6 was winning as
74 ... ~b4 well.
After 74 ...'itb3 75 a5 ~b4 76 a6 ~b5
77 %.ta8 <ittb6 78 ~d5 .Jtf2 (78 ... f3 79 a7
f2 80 lU8 'itxa7 81 ~c6 wins.
81...fl ='ii' 82 l:xfl ltb8 [82 ... .Jth2
83 l:tb 1] 83 ~d7 White cuts off the king
from b 1 and his king goes to g5 and
takes the h4-pawn.) 79 a7 ~b7
(79 ... .Jtg3 80 IIg8 'itxa7 81 <iPc6
transposes to the game.) 80 11£8 'itxa7
81 ltxf4 i.g3 82 l1b4 wins.
75 l1a8 <itb3 76 a5 ~b4 77 a6 <iPb5 85 <it?c7!
78 ~d5 ~b6 79 a7! ~b7 80 ltg8! White not only stalemates the black
~xa7 81 ~c6 king but also creates threats of
See diagram on page 109. checkmate.
III
8S ... .i.g3 86 ~c8 f3 87 :e7! 93 ... rba4 94 rbe6 rbaS 9S rbfS· ~a4
Black is in zugzwang. After 87 ltb3 96 c;!?g4 c;!?aS 97 ltb7 1-0
~a7 88 :xf3 ~b6 Black holds as the
king can leave the comer.
87 ... f2 G.Kasparov - A.Yusupov
Linares 1993
1 d4 dS 2 c4 e6 3 12Jc3 fiLe7 4 cxd5

exdS S !iLf4 12Jf6 6 e3 fiLfS 7 12Jge2 0-0
8 12Jg3 .i.e6 9 .i.d3 cS 10 dxeS .i.xe5
11 0-0 12Je6 12 ltel
881:1f7
Black has no choice but to drop the
pawn because of the zugzwang.
88 ... .i.eS 89 1:Ixf2 rba7 90 l1fS! .i.g3
12 ... .i.d6
Artur is ready to defend the isolated
pawn middlegame. It looks like Black
gets away with exchanging all the
central pawns if he were to play
12 ... d4!? 13 12Jce4 (after 13 12Jb 5 .i. b6
14 e4 12Jg4 Black has a nice position,
as in Lautier-Marciano, France 1999)
911:1bS! 13 ... .i.e7 14 12Jf5 .i.xf5 15 12Jxf6+
White must make sure Black's king .i.xf6 16 fiLxf5 1i'a5 (16 ... dxe3
stays in the comer. It is remarkable that 17 .i.xe3 'i'i'xd 1 18 l1fxd 1 1:Ifd8 19 iLd7
Black loses this type of endgame if his 12Je5 20 .i.b5 1:Ixd 1+ 21 l1xd 1 a6
king is in any comer, whereas he can 22 iLe2 l1d8 23 l1c 1 12Jc6 and Black is
draw if his bishop gets on the e I-h4 able to live with White's two bishops as
diagonal and his king reaches the in Karolyi-Zahilas, Hungary-Greece,
centre. It can be very useful to know E-mail Olympiad 2000) 17.i.bll1ad8
these secrets. Incidentally 91 1:If6 wins 18 'iVb3 Itd7 19 exd4 l1fd8! and Black
as well. is not worse, Gulko-Shabalov, Seattle
91...~a6 921:1bl ~aS 93 ~d7 2000.
White just collects the h-pawn and 13 12Jh5 .i.e7 14 12Jb5 12JxhS
wins. IS 'iWxh5 g6 16 'ii'f3 1:Ic8 17 1:Ifdl 'ii'd7
112
TIgran Petros ian the 9 th
18 b3 ':fd8 19 ~g3 lOb4 20 lOe3 37 iH4

lOxd3 21 ':xd3 ius 22 ':d2 'ii'e6 Reducing the amount of material. On
23 ':cdl h5 24 h4?! the other hand White will now have
Taking the pawn with 24 lOxd5 was more freedom of movement for his
better thanks to a little tactic. 24 ...':xd5 king. Another idea is 37 :a4 a5
25 lhd5 h4 26 "0 i.e4 27 l:.e5!
24 ...':e5 25 f3 'iVe6 26 e4
38 ':c4!?
37 ... i.xf4 38 .t1xf4 !itfd6 39 ~e3 ~e5
40 :d4 i.e6 41 a3 as 42 ~e4 b5
After 42 ... f6 43 f4 White can breach
the fortress by defending the queenside
with the king on c 1 and invading with
the rook on d8.
43 ~e5 34 44 f4 ~c6 45 ~f6 ~e5
46 :b4 i.e4 47 ~e7 .i.e6 48 lte4 ~d5
49 :1d4+ ~e6 50 <itd8
White cannot penetrate yet.
50 ...i.f5 51 ~e8 .i.e6 52 ~fB ~e5
26 .. Jhe3! 53 ~g7 !itfe6 54 ~g8 ~e5 55 ~f8 ~e6
The exchange sacrifice offers real 56 ..tg7 <it>e5 57 ~h8 ~e6 58 Wb7 <;t>e5
chances of survival. 59 !itfh6 i.f5 60 ~g5 i.e6 61 ~f6 ~e6
27 bxe3 'ii'b6+ 28 <ith2 dxe4
29 lhd8+ i.xd8 30 i.e3
30 fxe4!? offers White a more
promising way to enter an endgame.
30 ... i.xe4 31 'ii'e3 i.c6 32 'iVe5!?
(32 'ii'xb6 i.xb6 33 i.e5) 32 ... ~h7
33 l:.d2 'ii'a5 34 'iixa5 i.xa5 35 i.e5.
30 ... 'ii'a5 31 'ii'b8 'ii'e7+ 32 'ii'xe7
i.xe7+ 33 litgl exf3 34 gxf3 b6
35 litfl litfB 36 l:[d4
62 f5!
Reducing the number of pawns while
simultaneously freeing the f4-square
for the rook.
62 ... i.xf5 63 ~xf7 <ite5 64 ~f6 ..te2
65 rJ;e7 i.f5 66 ~d8 ~c6 67 l%(4!
It was 62 f5 that enabled him to play
this move.
67 ...~d6 68 :b4 ~c5 69 ~c7 .td3
70 .t1d4 i.e2 71 ~b7 i.f1 72 <;t>a7 .te2
36 ... ~e7?! 73 .t1e4 i.d3
This allows the exchange of bishops. But not 73 ... i.c4? 74 <;t>a6 b4+
113
75 l:txc4+! and having the king on c4 88 ltc3! ~f7 89 ':c2 ct>h3 90 l:lc5 ~c4!
enables the g-pawn to promote. If (90 ... h4? 91 l:lxb5 Wg2 92 l:tb2+ ct>gl
75 ... ct>xc4 76 cxb4 g5 77 hxg5 wins. [92 ... ct>f1 93 l:.h2 'iPel 94 ct>e3 wins]
74 l:tb4 ~c4 75 Wa6 ct>dS 76 <it>aS 93 <it>g4 <it>fl 94 <it>xh4
ct>eS
Black can't stay on the queens ide
with 76 ... ~c5, because of zugzwang.
77 l:tbl ~d3 78 l:tb2. Then 78 ... ~c4
79 l:td2 ~fl 80 l:td4! ~e2 (80 ... ~g2
81 l:td8 ~fl 82 l:tc8+ ct>d5 83 <it>b4)
81 l:te4 ~fl 82 lIe5+ wins.
77 l:tbl
Black's king is in a poor position and

now he loses. If his king were instead
on f6, that would be enough for him to
obtain a draw. If Black's bishop were
dark-squared and situated on f8, the
position would also be a draw even
though White could defend the a-pawn
with his rook along the third rank.
77 ... gS? 94 ... <it>el 95 ct>g4 <it>dl 96 ct>f4 ~b3
Black reduces the nwnber of pawns 97 <it>e3 <it>c 1 98 l:lh2 .i.c2 99 ct>d4 <it>b2
but the final endgame is unfortunate for 100 <it>c5 wins. [but not 100 <it>c4??
him. Keeping the pawns on with ct>xa3].) 91 l:lxh5+ CiPg2 92 ct>e3 CiPfl
77 ... <it>f4 was just enough to survive: 93 l:tg5 ~e2 94 ~d2 ~c4 95 1:f5+
78 lIgl ~d3 79 'it'b4 ct>f3 80 c4!? ~gl! 96 ct>e3 ~g2 and White can make
(80 ct>c5 ~f2 81 l:lg5 ct>f3) 80 ... ~xc4! no progress as I pointed out in my
(80 ... bxc4? 81 ct>xa4 ct>e3 82 <it>b4 <it>d2 Informant analysis.
83 lIg2+ ~e2 84 a4! c3 85 ':xe2+ 78 bIgS ct>fS 79 1:g1 ct>g6 80 ct>b4
'it>xe2 86 'it'xc3 g5 87 hxg5 h4 88 g6 h3 h4 81 <it>cs h3 82 <it>d4
89 g7 h2 90 g8='ii' hl='ii' 91 'ii'g4+! Not 82 l:lh I? ~e6 83 ~xb5 ~xg5
[or 91 "'e6+!] 91...~f2 92 'ili'f4+ White 84 <it>xa4 <it>g4 85 CiPb5 ~d5! and White
can swap queens and win.) 81 l:txg6 has to give up the rook for the pawn,
<it>f4 82 l:tg5 ~e2 83 ~c5 ~g4 84 ~d4 while the bishop contains the a- and
'it>g3 85 ~e3 <it>xh4 86 ct>f4 Black's c- pawns.
king is choked, and it's scary, but Black 82 ... ~e6 83 ct>eS ~d7 84 'it'f4 ~c6
can hold. 86 ... ~d7! 87 l:tc5 ~e8! 85 <it>g3 ct>IgS
114
TIgran Petrosian the 9 th
otherwise White's king invades, e.g.

89 ... ~e8 90 ~g4) 90 ~g4 ~d3.
88 ... bxc4 89 l:d4+ ~e5 90 l:txc4
86 l:ld1!!
White exchanges the weak c3 pawn.
After 86 <ifi>xh3+ ~f4 87 .J:g3 <ite4
88 ~g4 ~d5 89 ~f4 <itc4 Black's king Since the computer programs have
is active. reached new levels of analysis we can
86 ... h2 87 ~xh2 ~f4 understand this type of endgame much
If 87 ... ~f5 88 l:td4! and White cuts better. Now the bishop can't reach b3.
off the king. (On the other hand 88 c4 90 ...Wd5 91 :b4 ~c5 92 ~g3
allows Black to escape into a Not 92 l:tb2? i..d5!
favourable version of the game. On 92 ....i.b5 93 ~f4 ~b6
88 ... bxc4 89 l:td4 c3 90 l:tc4 ~d5 If 93 ... ~c6 94 ~e3 .i.b5 95 ~e4
~c6+ 96 ~d3 ~b5+ 97 ~c3 .i.e8
98 :f4 wins.
94 ~e3 ~a5 95 ~d4 .i.e2 96 l:tb 1
~h5 97 l:te 1 .i.f7
the bishop would reach b3, and that

would save him as his king is in the
centre and can't be forced into any
comer.) 88 ... ~e5 89 ~g3 and White's
king has time to return to the centre and Now the bishop reaches b3 and yet
push Black back. Black is lost because his king is in the
88 c4 comer. If Black's bishop were dark-
Now cutting otfthe king by 88 lld4+ squared and stood on f8 it would be a
is ineffective, as it is only temporary. draw.
88 ... ~e3 89 ~g3 (if 89 l:td6 ~e4) 98 ~c5 ~b3 99 l:te8 ~a6 100 <ittc6
89 ... ~e4! (Black has to approach c3 <it>a7 101 ~b5 <itb7
1'15
112 l:tc7+ ~b8 [1l2 ... ~d8 113 :a7

wins.] 113 ~c6 ~c2 114 ~b6 i.b3
115 ltd7 ~c8 116 l:[d4 is the same
zugzwang.) 112 <i!i>c6 'it>d8 113 l:ld7+
'it>c8 (113 ... ~e8 114 :d4 wins.)
114 l:lh7 ~f3+ 115 ~b6 .i.dl 116 :h4
~b3 117 ltd4! wins. 102 :e4! rl;c7
103 :d4 ~b7 104 l:td7+ ~c8 105 ~c6
.i.c2 106 1iId6 ~d 1 107 :n ~b3
108 nh7 .i.dl 109 ~c6 .i.f3+ 110 'it>b6
102 1:1e7+?? .i.d 1 111 nh4 ~b3 112 nd4 also wins.
Petros ian won because his 102 •..1iIc8 103 ~c6 ~d8
opponent's king was too far away, I
hoped that I could chase the king to
where Tigran's opponent's king was.
Better was 102 1:1d8! as the king has to
be kept in the comer, White wins even
if the bishop is on b3. This is the case if
Black's pawn were on as, his bishop on
b4 and White's pawn on a4. 102 ... ~c7
103 ltd3 ~c8 Postponing the
inevitable. If 103 ... ~b7 104 1:1c3 ~dl
lOS 1:1c1 .i.b3 106 ~a5 1iIa7 107 l:tc7+ 104 l:td7+ ~e8!
~b8 108 ~b6 .i.a2 [if 108 ... ~g8 Black escapes from the comer and
109 lte7] 109 l:ta7 .i.b3 110 ltd7 ~c8 that saves him.
III ltd4 105 ~c7 ~c2 106 l:td2 ~b3!
107 lte2+ ~f7 1 08 ~d6 ~c4
109 ne7+ ~f8 110 l:te4 ~b3
The key position. Black is in a lethal

zugzwang.) 104 <i!i>c6 .i.a2 105 ltd4
~n 106 llf4 ~e8+ The bishop has to 111 ~d7
leave the diagonal. 107 ~c5 1iIc7 If White had one more move, then a
108 l:te4 ~d7 109 lte7 1iIc8 11 0 ~d6 rook check would help - just as in the
.i.b5 III Iln .i.e2 (111...~d3 Petrosian game and he could then
116
Trgran Petros ian the 9th
rush his king to the b-file and win. He 113 ~d6 ~g5! 114 ~e5 ~g6
can't allow the black king to reach g5 115 ':f3 ~g7 116 :f6 .ltc4 117 ~f5
earlier. ~b3 118 ~g5 ~c2 th-th
111.. .ritn 112 l:H4+ ~g6! When I went home I quickly checked
Not 112 ... ~g7? which loses after the Petrosian game and noticed that he
113 ~c6 ~g6 114 ~b5 and the king was in fact Black and actually lost this
can't get to c8. game.
I had the most annoying experience of all with Petrosian, as he beat me with
a particular central pawn structure seen in the diagram below. I learned from
that and wanted to use my new found knowledge against Karpov when exactly
the same structure arose.
G.Kasparov - T.Petrosian A.Karpov - G.Kasparov
First here is the game against How could I try the Petrosian
Petrosian. variation against its inventor? Could
that in itself have been a mental
G.Kasparov - T.Petrosian blackout?
Moscow, 1981 4 ....i.b7 5 lDe3 d5 6 exd5 lDxd5 7 e3
~e7 8 .ltb5+ e6 9 .ltd3 lDxe3 10 bxe3
1 d4 e5 11 0-0 0-0
This is perhaps my most unfortunate
game. I picked up the idea from
Petros ian when he beat me. Then I used
it at a most critical moment against
Karpov in a World Championship
match. I lost that game as well when I
needed a draw to retain the title. While
analysing it for my Predecessors book I
realised some of my earlier comments
were not quite correct.
1... lDf6 2 e4 e6 3 lDf3 b6 4 a3 12 ife2 g6 13 e4 lDc6 14 iLh6 :e8
H7
Tigran Petrosian the 9th
This time there is no exchange 2S ~d2!

sacrifice. Now I can save it.
IS l:tfdl JJc7 16 1r'e2 Aed8 17 "'e3 2S ...St.d6 26 ':0 "'g7 27 a4?!
eS 18 dS ttJaS 19 c4 Dvoretsky and Yusupov found the
After 19 a4 c4 20 ~c2 .i.c8 the right move which was 27 i..b4,
position is equal. stopping .i.d7. 27 ......e7 28 f4 leaves
White a bit better.
27 ... aS 2811b2 St.cS 29 f4 i.d7
Petrosian suggested 29 ... b6!.
30 hS .i.xa4?
If 30 ... gxh5 31 'iVh4 .i.g4. This move
was not mentioned in my 1981
analysis. The position is unclear.
19... ttJb3
Petrosian centralises the knight. I lost
quite a number of games by leaving a
knight on a5 or h5. Nikolaevsky, Zaid,
Magerramov, Beliavsky and Gulko
beat me in those games.
20 1:[a2 f6 21 h4 .i.c8
Dvoretsky and Yusupov pointed out 31 h6
that Black should have exchanged the As we have already seen, Karpov
bishop with 21.. .id8!. pushed his pawns as far as this.
22 Abl ttJd4 23 ttJxd4 cxd4
31..:iic7 32 fS?!
I thought the d4-pawn was not bad,
In 1981 I did not spot that 32 fxe5!
but hoped to get a good game in the
would give me an advantage. Then
end.
32 ... fxe5 33 .i.g5 11f8 34 St.f6. I shall
24 ~g3
not repeat the analysis from My Great
See diagram on page 117. Predecessors. Suffice it to say White
24 ... .i.f8 retains the advantage against all Black
defences.
32 ... gS 33 .i.xgS! fIgS?
Better was 33 ... <it;f7. Petrosian often
used his king with great style. For
example when he beat Fischer in the
1959 Candidates tournament. After
34 i.c 1 ':g8 35 'iWh3 White has a slight
edge.
34 'iixg5+ <it;f8
118
37 ...1:d6! 38 'ifg8+ 'iVf8 39 li'g3

'ii'xh6
35 'iVf6+?
In time trouble I lose my way. Best
was 35 f6! 'ikfl 36 'iixe5 :e8 (if
36 ... 'iig6 37 :xb6! 'ikxh6 38 'ike7+!
wins) 37 'iig5 'ii'g6 38 lH5 and the 401ba4?
white pawns are too dangerous and he When I wrote the chapter on
will win. I lost to Krarnnik in a 1994 Petrosian I discovered that White could
Intel rapid game in a similar fashion. I still stay in the game or at least
consolidated a piece on d4, he took it, resist with 40 "'g8+. Then 40 ... 'ii'fS
then he sacrificed a piece, and later the
41 'ii'xfS+ ~xfS 42 l:txa4 and Black is
exchange. He went on to defeat me.
somewhat better. However it is hard to
You can see that game on page 166 in
tell whether he can win.
the Smyslov chapter. So even my own
analysis helped my rival, making this 40 ...'iVc1+ 41 ~f2 "'xb2+ 42 <&t>f3
really the most unfortunate game of all! q;f7 0-1
35 ... c1te8! 36 1:a1
My fourth World Championship
match with Karpov stood at 11-11 and
I needed to score one more point to
retain the title. This was the dramatic
end of the match.
Seville 1987
36......e7!!
This is a great defensive move. I 1 e4 e5 2 ttJf3 ttJf6 3 ttJe3 d5 4 exd5
understand now why Botvinnik did not ttJxd5 5 d4 ttJxe3 6 bxe3 g6 7 e3 .ltg7
always anticipate Petros ian 's moves. 8 .ltd3 0-0 9 0-0 'iVe7 10 %1b1 b6
37 'ii'e6? The pawn structure is becoming very
I should have swapped queens and similar to my game against Petros ian.
defended in the endgame. 11 1i'e2 lld8 12 ii.e4
119
19 lUc1 %:tac8 20 iLd2 :n 21 a4

he4 22 lbe4 :'cfS
I even control the f-file.
23 f3
12 ... .i..a6!
To weaken the d4-square.
13 c4lLlc6 14 dS fS IS ~d3 eS! 16 e4
lLld4
I got very excited and realised I did 23 ..• iLc8
not even have to expend tempi like I had yet to become acquainted with
Tigran Vartanovich. the analysis of Yusupov and Dvoretsky
17 lLlxd4 cxd4 who pointed out that ... .i..c8 was not the
best move for Petrosian. I just copied
See diagram on page 1J 7 his play.
24 as
I felt things were really going my Karpov did not let me fix the a5-
way. pawn. Okay, you can't have everything.
It is the same pawn structure and 24 ... .i..fS 2S 1r'e2 l:te8 26 ~e4 ~fS
I would be able to use the knowledge 271r'd3
I had gained from my loss against
Petrosian.
18 iLgS
27 ... ~cS
Now my bishop has arrived at the
same square as Petrosian's. Life can be
18 ... %:tfS sweet, I thought.
In addition I have more space on the 28 %:tal 'ii'd7 29 %:tel 'ir'c8 30 ~hl
queenside. l:tc7 31 %:tabl ~g7 32 l:tec1 ~xe4
120
Tigran Petrosian the 9th
If 48 ':xe5 ~d6 49 .:th5+. I was

hoping that sacrificing material would
not bring any luck, as had happened
16 years earlier. After 49 ... gxh5 50 e5+
~g8 51 exd6 'ii'xd6 52 'ii'xd4 'ii'f6
Karpov and Zaitsev evaluated the
position as equal.
:0
48... 49 ~d2 lLc5 50 l:1c6
33 fxe4
The central pawns are configured in
the same way.
33 ...:n 34 'il'g3 bxa5 35 ~xa5 lH4
36 l:te1 "a6 37 ~d2:n 38 'ii'd3 11efS
39 h3
Karpov doesn't push the h-pawn two
squares like I did.
39 ...:f2 40 11al 'ii'f6 41 :tgl h5!?
50 ... l:t7fJ??
42 l:ta5
A horrible blunder. thought
Or 42 ~el!? l:n 43 ~g3 l:txgl + 44
everything was going my way and it
~xgl.
was time to reap the fruit of my lexical
42 .. :iVe7 43 lIbl
knowledge. I felt it was time for a
knockout punch. In a way I was
successful as the battle does indeed end
fairly quickly. But after the game I was
not satisfied. A much better alternative
was 50 ... ~b4!?
51 gxfJ ]:txfJ 52 l:1c7+ <it>h8
43 ... h4
Showing my optimism, I don't need
to worry about putting pawns onto the
colour of Karpov's bishop as there will
be no bishop ending here.
44 :a6 J:l8n 45 ':c6
If 45 :e6 "f8.
45..••fS 46 J:lgl ~e7 47 l:te6 <ith7!
53 ~h6!
What a shock it was to be on the
48 ~el receiving end of this intennediate
121
Tigran Petrosian the 9'h
move. Suddenly Black is completely In the end I won the last game and
lost. thus kept my title but, my word, it was
53 .. Jbd3 54 .i.d'8 l:h:h3+ 55 ~g2 a close run thing. Had I lost the match,
:g3+ 56 'iii'h2 :xg1 57 .i.xc5 d3 Petrosian would have had something to
1-0 do with it.
122
Mikhail Tal the 8th
Tal decisively defeated Botvinnik in and certainly picked up many ideas
1960 to become world champion, but from these experiences. His calculating
he lost a return match in a similarly ability was one of his strong points. I
convincing fashion one year later. The believe this features in my chess as
magician from Riga was the last well.
champion to gain the title before I was I do have very nice memories
born and is famous for his very associated with Tal, but also some
aggressive attacking style. I played painful ones. Let's have a look at a few
some training and blitz games with him examples.
With symmetrical pawn islands of four kings ide pawns and a- and c-pawns,
whoever exerts greater pressure on the opponent's pawn structure should gain
the upper hand.
N .Rashkovsky - M. Tal A.Karpov - G.Kasparov
N.Rashkovsky - M.Tal He previously beat Gulko with the

Chi gorin Memorial, Sochi 1973 main move 8... c5. Tal was extremely
good in positions where one player had
1 d4 g6 2 c4 ll'lf6 3 ll'lc3 d5 a rook versus bishop and knight and
Interestingly, Tal rarely employed the I lost a slightly similar game to
Griinfeld - according to the databases Romanishin in Moscow 1981. Tal
only four times. It is a bit of surprise as played extremely powerfully when
this opening often provides very there was unbalanced material on the
dynamic positions, ones which Misha board and my most bitter memory in
would have handled so well. such circumstances was my loss to
4 cxd5ll'lxd5 5 e4ll'lxc3 6 bxe3 ~g7 Anand in Tilburg 1991.
7 ~c4 0-0 8 ll'le2 ll'lc6 9 0-0 b6 10 ~e3 ~b7 11 'iVd2 ltJa5
123
Mikhail Tal the 8 th
12 ~d3 e6 14 ... 0-0! 15 cxb6 axb6 However the

Let me stress the effect this game had pawn structure doesn't always give
on me. In my game against Korchnoi in Black a good game. For example,
the World Cup in Reykjavik 1988 I Smyslov beat Ribli with White the
same year in a Candidates match.
also played a similar e6 in a main line
16 0-0 'iWc7 17 ~b5 ~xe5 18 ~h6 ~g7
Griinfeld.
19 ~xg7 'it>xg7 20 "d4+ ~g8 21 ~g5
13 ~h6 cS 14 ~xg7 'it>xg7 15 "f4 h6 22 ~e4 ~xe4 23 "xe4 ttJa6 24 'iWe3
Korchnoi put hls queen on the
"c5 25 "xc5 ~xc5 26 11fb 1 ':fd8
queens ide - that game ended in a draw.
27 lIb4 ':d6 28 .i.fl ~f8 29 a4 lIa5
lS ... liJc6 16 ~bS .f6 17 'iWIf6+
30 g3 ~e7 31 'it>g2 f5 32 ~b5 ':d2
<it>If6 18 dICS
33 11d4 lIxd4 34 cxd4 liJxa4 35 11xa4
l%xb5 36 ':a7+ <it?d6 37 lIh7 h5 38 l:tg7
ltd5 39 l:txg6 b5 40 'it>£1 b4 41 <it?e3 b3
42 <it?d2 l%xd4+ 43 <it?c3 b2 44 'it>xb2
l:ld2+ 45 'it>c3 l%xf2 46 h4 f4 47 l%g5
It£1+ 48 <;!i>d4 lIxg3 49 ltxh5 lte3
50 lIh6 'it>e7 51 h5 e5+ 52 'it>d5 f3 0-1
Kasparov-Korchnoi, London 1983.
19 ttJc1
Whoever starts exerting pressure on
the opponent's pawn structure will be
the one who gets the upper hand.
18 ... bICS
19 ... llJaS! 20 :tel l%ab8 21 eS+ 'it>g7
We have reached a key position. 22 ttJb3
Alternatively, 18 ... lIfc8 would also
22 ~fl !? keeps the pawn structure as
have been interesting. Korchnoi beat
it is.
me only once in our many encounters,
the only time I suffered was during our
Candidates match in London 1983. He
used the pawn sacrifice motif under
rather similar conditions after 1 d4 liJf6
2 e4 e6 3 liJ£1 b6 4 liJc3 ~b7 5 a3 d5
6 exd5 liJxd5 7 e3 g6 8 ~b5+c6 9 ~d3
~g7 10 e4 ttJxc3 11 bxc3 c5 12 ~g5
~d6 13 e5 'iWd7 14 dxe5
22 ...~xg2!
Tal weakens Rashkovsky's pawns on
the kingside.
23 c4 llJxb3 24 axb3 ~b7 25 lba7
ltfd8! 26 ~f1 ~f3 27 ~d7 lIxb3
27 ... ltb7 also gives reasonable
winning chances as after 28 l:txb7
124
i.xb7 29 i.a4 g5 White is rather This is the very same queens ide
passive and his pawns are loose. pawn fonnation that Tal had against
28 i.xe6 l:b7 29 1:a3 fxe6 30 lbo Rashkovsky.
l:d4 31 lIf6 lIe7 32ltal See diagram on page 123.
32 l:tcl l:e4 33 f4 ltf7 34 ltxe6
Black's king is less active than in
l:tfxf4+ 35 <1tgl h5 is tough for Black,
Misha's game, therefore I thought
but not hopeless.
White had better drawing chances.
32 ... l:te4 33 f4 l:be4 34 <t>e2 l:te2+
Nevertheless I had no inkling of what
35 <ifi'd3l:bh2 36 lIct l:h3+ 37 ~e4??
was about to happen.
The king is frequently well-placed
Another famous game by a champion
in the centre, but not always. These
suggested to me that I have a good
exceptions make chess such a
position because of the pawn structure.
wonderful game. White finds himself
Here it is:
in a difficult endgame after 37 ~c4!,
although he has chances of holding on. L.Portiseh - R.Fischer
Piatigorsky Cup, Santa Monica 1966
37 ... l:d7!
Tal grabs the chance to net 21 i.f4 h6 22 lIe2 g5 23 ~e5 'iid8
Rashkovsky's king. 24 l:1fe 1 tJi;f7 25 h3 f4 26 ~h2 a6
38 l:e4 l:dd3! 0-1 27 l:te4 'ii'd5 28 h4ltJe3 29 l::tlxe3 fxe3
Out of the blue White's king is 30 lbe3 'iixa2 31 l:tf3+ ~e8 32 i.g7
getting checkmated. 'ifc4 33 hxg5 hxg5 34 %H8+ ~d7
35 %:ta8 ~c6 0-1
Now we look at the game which was 15 l:tabltOb6 16 i.e2 e5
inspired by Tal.
Moscow 1984
1 tOo d5 2 d4 tOf6 3 e4 e6 4 tOe3
i.e7 5 i.g5 h6 6 i.xf6 i.xf6 7 e3 0-0
8 'it'e2 e5 9 dxe5 dxe4 10 i.xe4 "as
11 0-0 i.xe3 12 "'xe3 'it'xe3 13 bxe3
ltJd7 14 e6 bxc6
125
Mikhail Tal the 8th
17 l:tfc1 22 tiJel
Karpov places the rook behind his Such a superb endgame player as
own pawn instead of occupying an Karpov is now retreating. This made
open file. I felt good as he was me feel that I was on the right track.
defending the c-pawn before advancing 22 .. .llb4
it and generally it is reassuring when Exerting pressure on the c-pawn.
your opponent feels obliged to defend. After 22 ... l:tb2? 23 f3! White can
Karpov doesn't go for Rashkovsky's follow up with ltJd3 (but not 23 ltJd3
pawn fonnation with 17 tiJe5 when i-xg2+).
after 17 .. .iLb 7 18 tiJd7 lHc8 19 tiJxb6 22 ... .i.e4 is also met by 23 f3 and
axb6 the position is equal.
after 23 ... i-g6 24 e4. Later I will show
17... .li.b7?!
how I was misled in my judgment of
Since this game other players have
such a bishop. Alekhine and Euwe are
developed the bishop on d7, following
guilty for creating that impression.
up with l:tfd8 and Wf8. But I felt I was
Then 24 .. .l:tb7 25 ltJd3 nc8 enables
ready to start applying pressure.
Black to survive without losing
18 Wfl iLd5?!
Out of 17 games played since this material.
game, nobody has lost this position 23.i.dl
with Black. Perhaps they didn't know Karpov keeps going backward. I was
Tal's game and perhaps I was pulling his strings just as if he were a
overconfident. Tal also developed his puppet.
bishop on this square. 23 .. .ltb7
19 l:tb5 tiJd7 Here I got a bit confused - suddenly
Not 19 ... .li.xa2? when 20 c4 wins. I had to retreat as well.
Better was 19 ... l:tfc8 ! 24 f3 l:td8 25 ltJd3 g5
20 l:ta5
When Karpov made this move I
started to feel even better about life. He
had already used his other rook to
defend a pawn, thereby giving up the
open file. He must have been feeling
troubled.
20 ... l:tfb8 21 c4 iLc6
26.i.b3
Karpov seems to be in trouble.
Indeed he defends c4 with one more
piece, when it is not even attacked.
However the picture is not so rosy for
Black.
126
Taking the pawn with 26 4Jxc5? led 42 g3 .:le8 43 ':g7!

to equality. After 26 ... 4Jxc5 27 l:txc5 Classical principles. The rook stands
l:tb2 28 ':xc6 ':dd2 Black's rooks are well on the seventh.
good enough to draw. 43 .. Jlrs 44 %ba7 :f2 45 'Wfi'b4
26 ...tttrs 27 lLlxc5
Now the sobering reality of the game
made me forget about Tal's win. I wish
I had not known about that game at all.
27 ...lLlxc5
White is just a pawn up for nothing.
Very annoying indeed. Misha, Misha
what did your magician's spell do to
me?
28 ':xc5 l:td6 29 ~e2 ttte7 30 l:tdl
lbd131 ~xdl ~d6 32 ':a5 f5 33 ~e2
45 .. .lbh2
h5 34 e4!? fxe4 35 fxe4
Fischer also misled me with the
power of the edge-pawn due to his
game against Taimanov. Tal is not the
only one who can be blamed for my
loss in this game. However he played
the biggest role in it. Simplifying with
45 ... l1b2 did not help either as then
comes 46 c5+ ~c6 47 'Wfi'c4 ..ia2
48 ~xa2 l:txa2 49 :a6+ '1t>b7 50 .:lb6+
~c7 51 ':h6 l::txh2 52 ~d5 l:thl
53 l::th7+ Wb8 (53 ...'lt>c8 54 <iftd6!)
35 ... ~xe4
54 <ifte4! - a very strong switch.
Tal was able to take the g2 pawn. I
46 c5+ Iitc6 47 ..ia4+ 'Wfi'd5 48 :d7+!
remembered that - and hoped to do it as
In an adjournment it was easy for
well. But it never happens.
Karpov. He had so many seconds to
36l:lxg5 ~f5 37 ~e3 h4 38 ~d4 e5+
analyse for him.
39 ~c3
48 ... ttte4 49 c6 .:lb2+ 50 ~a5 l:[b8
Karpov sends a message that he is
going to play on the queenside again.
39 ... ~bl 40 a3 l:te7 41 l:lg4
In his analysis Tal preferred 41 h3 to
the game continuation but he did not
mention his Rashkovsky game - maybe
he did not dare to!
41...h3
I made the h-pawn push one of my
trademarks. The legacy of Karpov.
127
Mikhail Tal the 8th
51 c7 55 l:d1 it.a2 56 l::te1 +

That same c-pawn that I pressurised Now even the e-pawn falls.
with llb4 finally decides the outcome 56 ... ~f4 57 l::te4+ ~g3 58 :txeS
of the game. It was hard to take. 'iPxg4 59 l:re2 1-0
51...l:tc8 52 ~b6 ~e3 53 ~c6 h2 Tal refused to work for either of us in
54 g4 l:th8 our world championship matches. In
Here 54 ... e4 55 l:th7 wins. this game however he helped Karpov.
Let's have a closer look at Tal's win as Black with a particular queenside
pawn formation where Black has a potentially powerful passed d4-pawn.
Tal's game against Rashkovsky was played in 1973 when I was 10. Of course
that age is a very formative period for a young and ambitious player and I was
keen to pick up all available knowledge from the former champion's games. I
V.Saigin - M.Tal A.Kochyev - M. Tal
N.Andrianov - G.Kasparov
In the first two diagram positions

Tal was Black, whereas in the third
I played Black.
I guess you already know the

results of these games.
V.Saigin - M. Tal 8 0-0 lLlc6 9 e3 .i.e7 10 exd4 exd4

Game 8, match, Riga 1954
See diagram above.
1 d4 lLlf6 2 c4 c5 3 lLlfJ e6 4 g3 cxd4 It's a fine line between a pawn being

5 lLlxd4 dS 6 .i.g2 e5 7 lLlfJ d4 a strong point on d4 or a target.
128
11 ciJbd2 2S ... ciJe3!

Smyslov also won this position with Tal, of course, finds a sweet tactical
Black against Golombek in 1956. shot like this. He sweeps the white
Smyslov is covered in the next chapter. pieces away from the d-pawn, which in
11 ... ~e6 12 l:te1 0-0 13 b3 'ii'd7 turn decides the outcome of the game.
14 ~b2 ':ad8 15 a3 as An alternative sacrifice, 25 ... ciJxf2, was
also promising. 26 ~d5+ (26 Ildxf2
~xf2+ 27 Ilxf2 d2 wins.) 26 ... 'ti'xd5
27 cxd5 ciJe4+ 28 'ii'xc5 bxc5 and
Black should win with the extra pawn.
26 be3 ~xe3+ 27 cwthl ~xd2
28 'ii'xd2 l:te2 29 'ft'c3 l:txg2
29 ... 'ife7 wins as well.
30 ~xg2
16 ciJe5
If White can block the d-pawn and
exchange many pieces, it can become a
nice target. However it is not easy to
achieve both objectives.
16 ...ciJxe5 17 ':xe5 b6 18 ciJf3 ~c5
19 'iid2 ciJg4 20 :teel d3 21 1:0 'ti'd6
221fc3?
30 ... d2
This is too optimistic. After 22 h3
Tal's game plan works so well; his
comes 22 ... ciJxf2 (22 ... ciJf6 23 ciJe5)
d-pawn is irresistible.
23 I:txf2 'iVxg3. Tal was so good at
31 l:tdl ~g4 32 ciJf3 'ii'd3 0-1
playing positions with two pieces
versus a rook. He won many games like
that with both colours. A.Kochyev - M. Tal
22 ... f6 23 %tadl ':fe8 24 J:td2 ~fS Moscow 4-teams, 1981
25 ciJgS
1 d4 ciJf6 2 c4 e6 3 g3 cS 4 ciJf3 cxd4
5 ciJxd4 dS 6 ~g2 eS 7 ciJb3
Retreating the knight is the most
popular choice here.
7 ... d4 8 e3 as! 9 exd4 a4 10 ciJ3d2
exd4
Again Tal pins his hopes on the
d-pawn.
129
Mikhail Tal the 8th
11 0-0 iL.e7 12 ttJa3 ttJc6 13 ttJf3 25 ... d2!

iL.e6 14 iL.f4 0-0 White wins the d-pawn but overall he
loses material.
26 'iixd2 ttJd3
Black wins the exchange and his
position is winning.
27 l:.e3 ttJxc1 28 "xcI ttJe5 29 ttJb5
ttJg4 30 l:.e2 .i.xe4
Tal exchanges some pieces in order
to invade.
31 .i.xe4 lLlf6 32 l:.d2 'fie7 33 .i.f3
l:tad8 34 <iifg2 b6 35 l:.xd8 lbd8
15 ttJg5 36 lLlc3 l:.d4 37 b3 axb3 38 axb3 .-e5
The grandmaster from Leningrad 39 .-c2 -.el 40 lLlbl ttJd7 41 ""2
doesn't go for exchanges at all. ':d30-1
15 ... iL.g4! 16 iL.f3 .i.f5 17 l:.el ttJd7
18 h4 h6 19 ttJe4 ttJde5 20 iL.g2 iL.b4 N.Andrianov - G.Kasparov
21 ~f1 Azerbaijan Team Championship,
Baku 1978
1 d4 lLlf6 2 c4 c5 3 lLlf3 cxd4 4 lLlxd4

e6 5 g3 d5 6 .i.g2 e5 7 lLlf3 d4
When I played this move I hoped the
pawn would perform heroically like the
d-pawn in Tal's games.
8 0-0 lLlc6 9 e3 .i.c5 10 exd4 exd4

21. .. d3
Somehow Tal's d-pawn has become a 11 iL.f4 0-0 12 lLle5
powerhouse again. My opponent went for exchanges.
22 l:tc1 ttJg6 23 iL.d2 %:te8 24 iL.xb4 12 ... ttJxe5 13 .i.xe5 ne8
ttJxb4 25 liel
130
Mikhail Tal the 8th
14 ~xf6 My d-pawn gets to the third rank as

And he sticks to his plan. Tal drove well.
his opponents crazy with his knight 20 l:[el 'ifb4 21 ~e6 lUs
moves. Although I was left with no Trying to retain as many pieces as
knight, I did not feel there were any possible, but it gives White time to
drawbacks. organise his defence. After 21...d2
14...'ihf6 ISliJd2 'ifb6 16lL)b3 ~e6 22 l:[e3 ~d7 23 ~xd7 %:txe3 24 fxe3
17 ~xb7 l:[xd7 and the strong d-pawn secures a
draw, but no more. If 21.. .%:te7 22 ~d5
.i.xd5 23 lhe7 'ii'xe7 24 cxd5 l:txd5
25 'ifd2 White is a bit worse but with so
few pieces he may get away with
blocking the d-pawn.
22 l:e3
17... l:[ad8
I was still optimistic, Black can win
back the pawn and simplify to a
drawish endgame. I felt the d-pawn has
the same latent power as in Tal's game,
while White's extra queens ide pawn
won't start working at all. Tal probably
would have won even if he had been 22 ... d2
missing the b-pawn in those positions. This looks just as strong as it did in
After 17 .....xb7 18 lL)xc5 "xb2. Tal's game.
18 liJxeS "xeS 19 b3 23 'ife2 l:td6 24 ~e4l:[fd8 2S l:[dl
But the difference is that it is well
blockaded this time.
2S ... g6 26 h4
White makes room for his king in
case of mating threats and will perhaps
push his h-pawn all the way to h6 in an
attempt to create his own threats.
26 ... hS?
I just wanted to stop the further
19... d3 advance of the h-pawn.
131
Mikhail Tal the 8th
27 ~Ig6 46l:bd21-0
My last move was a blunder, which Now even the d-pawn falls.
gave away a pawn.
27 ... ~J:c4 28 l:te8+ l:be8 29 'ii'J:e8+ Let me just add that I finally had
~g7 30 'ii'e5+ ~J:g6 31 'ii'g5+ something to cheer about when I
I did not lose just one pawn but defeated a tough opponent the
several moreover Black's king knockout world champion Khalifman.
becomes exposed. This victory was sweet indeed and
31. .. ~h7 32 'ii'Ih5+ ~g7 33 "g4+ I went on to win the tournament as
~f8 34 'ii'c8+ ~e7 35 'ii'J:c4 "as well.
36 b4 'ii'e5 37 ~f1 ~d8 38 'ii'c5 lidS
39 'ii'f8+ cl;c7 40 'ii'd7+ ~b6 41
'ii'd6 42 'ii'e3+ ~b5 43 a3 ~a4
"f3 G.Kasparov - A.Khalifman
FIDE Grand Prix, Moscow 2002
44 ih:a7+
15 ... liJc6 16 b5 axb5 17 cxb5 liJb4

44 ... ~b3
18 ltJc4 "f5 19 l:te5 'ii'c2 20 ~f4
Other champions have won so many
'ii'xdl + 21 l:txdl ~a8 22 a3 f6 23 axb4
games with long king marches.
1-0
However I did not have as much luck as
they did. In the next game Tal's opponent
45 'ii'e3+ l:[d3 seems to have a dominating king in the
After 45 ... ~c2 46 'ife4+! l:td3 centre, but it can also become a target
4 7 ~e2 White also wins. as we will see ...
132
In both these diagrams White's king is better centralised than Black's and
surrounded by the opponent's pawns. First we will look at how Tal snares
Augustin's king - similar to the way he trapped Rashkovsky on page 125!
I was hoping that I might catch my opponent's king in the centre as Tal did.
In any event, no way could I lose with an extra pawn ...
J .Augustin - M. Tal S.Rublevsky - G.Kasparov
J.Augustin - M.Tal 19 h3 ~e6 20 'ifb3 'ii'd7 21 <i1th2 :lea8

European Team Championship, 22 ~f4 1:Ua3 23 'ii'c2 l:lc3 24 'iVd2 b6
Moscow 1977 25 ~h6 ~e5 26 ~f4 ~h8 27 %:ta2
l:tac8 28 l:tbl ~f6 29 ~g5 .e7
1 c4 e5 2 liJc3 liJf6 3 liJo ttJc6 4 d3 30 ~xf6 'iVxf6 31 "f4 'ii'xf4 32 gxf4
d65 g3 g6 6 ~g2 ~g7 7 0-0 0-0 8 ':'bl ':'c2 33 ':'bb2 %:txb2 34 :lxb2 l':tc5
a5 9 a3 liJd4 10 b4 axb4 11 axb4 c6 35 ~g3 <itf8 36 ':b4 'i;e7 37 lbd4
12 b5 ~g4 13 liJxd4 exd4 14 liJe4 ':xb5
liJxe4 15 ~xe4 ':'e8
38 f5
16 l:lb2?! With this imaginative pawn sacrifice
If 16 bxc6 bxc6 17 ':'b2 and unlike the Czech player opens a route for his
the game White doesn't have to defend king to the centre. It seems to improve
his b-pawn. his rook.
16... d5 17 cxd5 cxd5 18 ~g2 ':'al 38 ... gxf5 39 l:I.a4 ~d6 40 d4 l:tb2
133
41 .i.f1 ':bl 42 .i.g2 b5 43 :a6+ rj;c7 S.Rublevsky - G.Kasparov

44 ~f4 EU Cup, Izmir 2004
1 e4 c5 2 ll)f3 ll)c6 3 .i.b5
White's king can become menacing Rublevsky is quite an expert in the
in the centre. Rossolimo variation, but I also have
44 ... l:1b4 45 ~e5 pleasant memories with White. For
The king looks like a powerhouse on instance I beat Salov with it.
the e5-square. 3 ... e6 4 0-0 ll)ge7 5 c3 a6 6 .i.a4
45 ... l:ta4!
Out of the blue, Tal virtually traps 6 ... c4
White's rook and also exploits his This is an ambitious move. The pawn
unfortunate king. can be a target too.
46 l%d6 b4 47 .i.l:d5 7 'ife2 b5 8 .i.c2 ll)g6 9 b3 'iJlc7
10 bIC4 lLlf4 11 'ife3 bl:c4 12 .i.a3
.i.e7 13 .i.Ie7 lLlIe7 14 lLla3
This softens up the c4-pawn.
14... 0-0 15 l:tabl
This is Rublevsky's novelty.
15... f5 16 'iib6 'ifIb6 17 :bb6 fIe4
18.i.l:e4
47 ...1%a5!
A lethal pin shows just how
precarious White's king is on e5.
48ltc6+ ~d7 0-1
Tal's b-pawn will win the game.
I wanted to plagiarise the idea! And

ny punishment came not in court but
Iver the board ... 18... d5
134
Mikhail Tal the 8,h
Black can hold on tight to the All so classical getting to the

c4-pawn, but his central pawn chain seventh with the rooks.
becomes somewhat rigid. 27tOe6!
19 ~c2 tOeg6 20 ~Ig6 tOIg6 Forcing more exchanges.
21tOc2 e5 22 tOe3 ~f5 23 tOJ:f5! 27 ...:2f6
Sergei gets closer to the pawns by After 27 .. .1;[8[5 28 1:b8+ tOf8
exchanging the pieces around it. After (28 ... <iPfl 29 tOg5+) 29 l:txf8+ Ihf8
23 1:[e I ~d3 ! 24 tOxd5 e4 25 tOd4 30 tOxf8 lhf8 31 a4! White has decent
1:[fb8 Black has some compensation
winning chances.
according to Rublevsky. Alternatively
28 tOIfS 1:Ia6 29 tOxg6 hxg6
23 tOxd5 ~e4 24 tOe3 ~xf3 25 gxD
After 29 .. .l:hg6 30 ~f2! (30 l:tb5
1:[f4 26 1:[c6 tOh4 and Black has nice
1:d6) 30 ...:a6 31 ltb5 d4 32 cxd41ha2
play.
33 ~e3 Black is struggling.
23 ...1:[If5
24 1:[fbl! 1:[afS 30 <itfl!!

Getting to the second rank attracted Bringing the king into the centre. I
me. Not 24 ... tOf4? 25 g4! 1:[ff8 hoped I was going to catch White's
26 tOxe5 1:[ae8 27 tOd7 tOh3+ (27 ...:fl king the same way that Tal did.
28 1:[b8) 28 ~g2 tOf4+ 29 ~f3 1:[fl And in case there were no efficient
30 1:[b8 and White is clearly better as tactics against the king, I thought the
the winner pointed out. Better was extra pawn would still guarantee a
24 ... 1:[f6! although White still has the draw.
upper hand after 25 1:[xf6 26 :b7. 30 .. Jba2
25 1:[Ia6 e4 26 tOd4 :xfl After 30 ... d4 31 cxd4 lIa3 32 1:el
ltd3 33 ~e2 ltxd4 34 l:tc3 ~f7 35 a4!
wins according to Rublevsky.
31 <wPe3
The king is lured into a cage in the
centre of the board.
31. ..<it'f7 32 1:b7+!
Unlike Tal's opponent, White doesn't
step further into the danger zone.
32 .. .'1tf6 33 :b6+ ~f7
135
Mikhail Tal the 8th
See diagram on page 133. chances. Interestingly, Black should

now look for exchanges.
If 33 ... ~f5 34 l:d6 ~e5 35 l:lxg6 White could go back with 39 <it>d4!?
l:ta7 36 l:tg5+ ~e6 37 h4 Black has a and push his g-pawn before doing
very tough ending. anything else. (After 39 l:td7+ ~e8
34l:td6 ltaS 40 l:lxg7 [40 l:lxd5 e3! 41 dxe3 l:lxc3
Here I realised White's king could Black should be able to hold] 40 ... e3!
not be caught, but still didn't appreciate 41 dxe3 l:lxc3 42 l:lxg6 ltxe3+
the full scale of Black's problem. 43 ~xd5 c3 and Black saves himself.)
3S h4 Then 39 ... .:b2 40 ~e3 ':b5 41 g4 l:la5
42 g5 ltb5 43 'ii?f4 Rublevsky gets to
this position by a different move order.
He stops analysing here, evaluating the
position as a clear win. However Black
seems to be able to live with passive
defence. 43 ... lta5 44 ~e5 1%a3 45l:[d7+
~e8 46 ':xg7 e3 47 dxe3 ':xc3 48 ~f4
(48 ~xd5 1:txe3 49 ltxg6 lth3 and
Black survives.) 48 ... .:cl 49 l:[xg6 c3
50 'it>e5 c2 51 ltc6 d4 and Black is safe.
3S ... gS? 36 hIgS ~e7 37 ltc6 1%al
Aware of the trouble, I resorted to
desperate measures. However, with
35 ... .l:tb5 Black could wait and see -
and he has a hidden resource that
provides tough resistance if White
further improves his king. Then 36 g3!?
(White could also play 36 g4 at once,
which might create more obstacles for
Black.) 36 ... l:ta5 37 ~f4 nb5 38 ~e5
38~d4
The king is in the cage, but there is
no way to hurt it. Furthermore, I can't
prevent it from carrying out its threat.
All so tragic.
38 .. Jtdl
38 ... 1%a3? 39 ~xd5 e3 40 ':e6+ wins.
39 'it>IdS e3
Forcing simplification but not to the
38 .. J:tb3!! This move prompts desired extent. If 39 ... nxd2+ 40 <it>xe4
enough exchanges to create drawing lhg2 41 ~f5.
136
Mikhail Tal the 8th
40 l:le6+ ~d7 41 lb:e3 l:lxd2+ g7-pawn. On the other hand, giving up

42 ~Ic4 l:lxg2 43 l:e5 ~d6 the g6 pawn to advance the passed
If Black could exchange his last pawn would let Black off the hook, e.g.
pawn the game would be a draw. 55 ~b7? l:lg2 56 l1dl l:[xg6 57 c6 11g2
44 l:la5 l:lg4+ 45 ~b3 l:lgl?! and White is not far enough ahead in
According to my opponent I should the race.
have played 45 .. .'-fi'e6 and then 46 c4 g6 52 <it>c611g1
47 ~b4. After 52 ... <it>f5 53 ~d7 l:lcl 54 l:lc6
l:lal 55 c5.
46 ~b411bl+ 47 ~c4 ~e6 48 1:a6+

~f5 49 g6 11g1 50 ~b5 ~e5
After 50 .. .l:tb 1+ 51 ~c5 l:lg 1 52 c4 53 ~d7!
l:lcl 53 l:lc6 'it>e5 54 'it>b6 White's king White simply collects the g7-pawn
penetrates. and wins. According to my plan I was
51 c4 l:lbl+ supposed to hurt this king in the centre,
If 51...l:lg2 52 c5 :gl 53 l:ld6 just like Tal did - but instead it invaded
(on 53 ~b6 11xg6+ Black would not my territory with fatal results.
lose) 53 ... l:lg2 54 ~b6 l:lgl 55 ~c7! 53 ...l:ldl+ 54 ~e7 l:lbl 55 l:la5+
White's king walks over to collect the 'it>d4 56 <it>f8 l:lb7 57 l:lf5 1-0
G.Kasparov - Y.Anikaev
From Tal I picked up a way to

crack the Rauzer fonnation when
Black castles kings ide.
This time I show my position first

followed on the next page by two
misleading ones from the Magician
from Riga.
137
M. Tal - I.Platonov M. Tal - Y.Sakharov
We still start by looking at Tal's 15 %tafl liJa5 16 l:th3

games.
See diagram above.
M.Tal- I.Platonov
Dubna 1973
16... liJxb3 17 axb3 llac8 18 ~hl f5
1 e4 c5 2 ltJ13 d6 3 d4 cxd4 4 ltJxd4 19 exf5 'ifb4 20 f6 ~xf6 21 f51bc3?
ltJf6 5 ltJc3 a6 6 ~g5 e6 7 f4 'ii'b6 Black misses a win here by 21...exf5!
8 "d2 "xb2 9 ltJb3 Then 22 'i!rb6 (22 l:tff3 f4) 22 ... ~g7 23
Nowadays the popularity of this line
l:tg3 'ii'xc3! decides.
is increasing.
9 .....a3 10 ~xf6 gxf6 11 ~e2 ltJc6
22 l:tg3+ <it>h8??
12 0-0 ~d7 13 ~h5 ~g7 14 ':13 This loses to a wonderful fmesse.
Bringing the rook into play. With 22 ... ~g7 Black has a beautiful
14 ... 0-0 defence thanks to the weak back rank.
Portisch later put his king on the 23 f6 (23 l:txg7+ ~xg7 24 'ii'g5+ ~h8
queens ide and defeated Tal in a well- 25 :f4 :to!! forces White to accept a
known game in 1976. perpetual check.)
138
Mikhail Tal the 8th
23 ...:0!! 24 l:txg7+ <it>h8 25 it'cl 14 tLlb3

l:Ixfl + 261hfl 'it'c3 and the position is In this earlier game White wasted a
equal. tempo by putting his rook on b 1 - and
23 'ii'h6 lhg3 his attack still broke through. This
increased my confidence.
14 ...%1d8 IS f5 tLle5 16 ltg3
16... ~h8 17 lin 'ilrb4 18 ~hl d5

19 exdS exd5 20 tLld4 'ii'fB 21 lth3 b5
24 ~g6!! 1-0
White sets up an unstoppable
checkmate threat.
Here is his second game cracking the

Rauzer fonnation. 2211f4
Tal brings another piece into the
M.Tal - Y.Sakharov attack.
USSR Championship, Kiev 1964 22 ... .i.b7 23 lIfb4 'ii'g8 24 'ir'f4 l:tac8
2slOdl
1 e4 cS 2 tLlo d6 3 d4 cxd4 4 tLlxd4 If 25 l:th5 tLld7.
tLlf6 5 tLlc3 a6 6 ~g5 e6 7 f4 ""6 25 .. J~e8 26 tLle3
8 'ii'd2 'ii'xb2 9 ltbl 'iVa3 10 ~xf6 First it was the rooks and the queen,
gxf6 11 ~e2 ~g7 12 0-0 0-0 13 %10 now it is the distant knight that joins the
tLlc6 attack.
139
Now here is my game:
G.Kasparov - Y.Anikaev
USSR championship, Minsk 1979
1 e4 c5 2 llJf3 d6 3 d4 cxd4 4 llJxd4

llJf6 5 llJc3 llJc6 6 ~g5 a6 7 'ii'd2 'itb6
8 llJb3 e6 9 ~e2 "c7 1 0 a4 b6
11 ~xf6 gxf6
I was happy to have the typical
Rauzer pawn fonnation. One year
26 .. .l:lc3
earlier, back in my home town, I lost a
Black has not obtained enough play
game to Sideif Zade with reversed
on the queenside. He'll make Tal pay
colours. Here it is: 1 llJf3 llJf6 2 c4 g6
attention to that.
3 llJc3 d5 4 cxd5 llJxd5 5 'ii'a4+ ~d7
27 :h5 llJd7 28 ~d3 l:te5 29 'ii'h4
6 'iib4e6 7 llJxd5 exd5 8 'ii'd4 f6
llJf8 30 llJg4 :e8
9 'ii'xd5 llJc6 10 'ifb3 'ike7 11 d3 0-0-0
Giving up the exchange with
12 ~d2 ~g4 13 0-0-0 ~xf3
30 .. Jle4 removes some of the attacking
pieces and gives him more practical
chances after 31 ~xe4 l:txh3 32 'ii'xh3
dxe4.
31 c;t>gl
A careful move.
31 .. .l:lc4
14 gxf3 llJd4 15 "a4 llJxe2+ 16 <ifi>b 1

llJd4 17 "x a 7 "c5 18 'it'a8+ <ifi>d7
19 ~h3+ f5 20 'ii'xb7 'ii'c2+ 21 <ifi>al
~d6 22 "d5 %:a8 23 Jta5 'ika4 24 b4
32llJxf6
Tal has reached his optimum position
so starts the decisive operation which
leads to a win of the exchange with no
compensation for Black.
32 ... ~xf6 33 'ii'xf6+ "g7 34 'ii'xg7+
<itxg7 35 l:tg3+ cJilh8 36 ~xc4 dxc4
37 :h6 :d8 38 c3 ~e4 39 a3 :a8
40 :d6 as 41 llb6 b4 1-0
140
Mikhail Tal the 8,h
24 ... ltxa5 25 bxa5 liJc2+ 26 ""b2 12 f4 ~h6

l1b8+ 27 ~c 1 liJd4 28 ""d2 -"'4+ 0-1 This gains a tempo.
Kasparov Sideif Zade, Azerbaijan 13 0-0 ~b7 14 :f3 lId8 15 lth3 ~g7
Team Championship, Baku 1978. So I 16 f5 0-0
thought if he castles lon.g I was ready, Black's king could still walk over to
as I knew Tal's game. By the way the the queens ide, but I decided to look for
Rauzer fonnation with Black caused safety on the kingside. And I did not
me another painful memory. I held a even have to sacriftce for an attack that
record of winning or tieing for ftrst in looked very promising.
all tournaments for a period of 9 years
and 9 months. I did not make it a clean See diagram on page J37.
10 years because I came 2-3 rd in
17 "f4 tL'le5 18 ltdl 'ii'e7 19 "h4 h6
Linares 1991 where Ivanchuk was the
20 ..n
winner. He beat me in the very ftrst
I decided to improve the bishop as
round.
well.
20 ...'ii'c7 21 tL'ld4 ~c8 22 ikf1 ~7
V.lvanchuk - G.Kasparov
23 .i.c4 d5 24 exd5
Linares 1991
1 e4 c5 2 tL'lf3 d6 3 ~b5+ tL'ld7 4 d4

tL'lgf6 5 0-0 cxd4 6 'ifxd4 a6 7 ~xd7+
~xd7 8 ~g5 h6 9 ~xf6 gxf6
24 ... exd5
At ftrst it all looked similar to Tal's
game against Platonov, but by now the
pawn structure is the same as in the
10 c4 e6 11 tL'lc3 l::tc8 12 ~hl h5 Sakharov game. I knew that game as
13 a4 h4 14 h3 jie7 15 b4 a5 16 b5 well so I was still optimistic.
ikc7 17 tL'ld2 'ifc5 18 'ifd3 ':'g8
25 ~b3 l:fe8 26 .:tg3 ~h7 27 'iVa
19 l:tae 1 'ifg5 20 ':'g 1 'iff4 21 lIefl b6
Bringing the queen closer to the king
22 tL'le2 "ifh6 23 c5 l:txc5 24 lDc4 ~f8
25 tL'lxb6 ~e8 26 f4 f5 27 exf5 J:[xf5 by 27 'ir'f4, as Tal did, was preferable.
28 l:tc 1 ~g7 29 g4 l:tc5 30 ':'xc5 dxc5 27 ... tL'lc4 28 tL'lde2 b5 29 axb5 axb5
31 tL'lc8 ~f8 32 'ifd8 'ifg6 33 f5 "irb6 30 tL'lf4
34 g5 ~5 35 l:tg4 exf5 36 lDf4 1r'h8 Tal transferred the other knight to g4,
37 "f6+ ~h7 38 l::txh4+ 1-0 but I was happy with my choice.
141
Mikhail Tal the 8'11
Black takes the exchange.

51 'ilhd4 '1i'd8 52 lOb5 l:te8
52 ... lia6 followed by 'itb6 wa!
winning.
53 g5 hxg5 54 lbg5 l:tg8 55 l:txg8
After 55 'ilfh4 fxg5 56 lOf6+ ~g7
57 'ir'xg5+ ~f8 58 lOh7+ ~e8
59 'ihg8+ ~d7 60 "xfl+ ~c8 6 I ~g4
White has some practical chances.
55 ... ~xg8 56 lOb5 ':e6
30 .. .l:ieS 31 lOhS ~h8 32 l:tf3 lOxb2 56 ... l:ta6! gains a winning tempo.
33 l:td4 lOe4 34 ~xe4 bxe4 35 g4 57 'it'g4+
This is something Tal did not play, Unfortunately the queen stands on
yet I liked it. Sadly my position the same diagonal as the bishop. Best
gradually deteriorates as Anikaev was 57 'ifgl +! ~f8 58 lOd4 ':a6
59 lOe6+ l:txe6 60 fxe6 ~e7 6 I exfl
steadily improves his pieces.
and the position is equal according to
3S ... 'iWe7 36 '1i'd2 ~b7 37 ~f2 l:te8
Anikaev.
38 lOf4 ~g7 39 lih3 ~g8 40 lOee2 57 ... ~f8 58 lOd4 lla6
~f8 41 lOe3 'ir'eS 42 lOhS 'ifb6 43 l:tf3
~h 7 44 ~g3 l:td8
59 lOe6+ l:txe6 60 fxe6 q;e7 61 'ii'g7

~xe6 62 lOf4+
45 ~b4 Sometimes Tal caught the king in the
Steinitz liked to walk with his king ... centre in endgames even though he had
4S ... ~b4 46 'ir'f4 no queen. This time White has no
After 46 J:.f1 l:te2! 47 'ir'xe2 'ifxd4 chance even with the queens.
White is in trouble. 62 ... ~e5 63 ~g3 'ir'a5 64 'itf2 'ii'b6+
46 .....d6 47 l:tg3 'ir'e7 48 ~h3 ~eS 65 ~f3 d4+ 66 ~g4 d3 67 'ir'xn 'ir'gl +
49 'it'd2 l:td6 50 lOf4 ~xd4 68 lifrh4 'ir'g5+ 0-1
142
Vassily Smyslov the 7th
Smyslov won the world title in 1957 the champions and with that many
by beating Botvinnik 121h-9 1h. They games he had a stronger effect on me. I
had already played a match three years had the most games against the other
earlier, when they drew 12-12. In 1958 champions, but only played matches
Botvinnik won the rematch but with Karpov and Smyslov. Against the
Smyslov kept on playing successfully rest I just played a few games. Let me
for four decades. Incredibly he made it give you a few examples of Smysloy's
to the Candidates matches fmal, where influence on me.
I met him. The age difference between Firstly, here are a couple of positions
us is 42 years. He had the longest from Smysloy's games that caused me
career, playing the most games of all particular damage.
White's queenside pawn majority, spearheaded by the pawn on c5,

often occurs in the Alekhine defence and Panov attack versus the Caro-Kann.
V.Smyslov - W.Schmidt G.Kasparov - A.Yermolinsky
V.Smyslov - H.Azizi G.Kasparov - A.Dreev
143
Any loss is very painful, two losses 25 l:.b4

are even worse, but to lose two games It is quite an unusual role for the rook
with White is really pathetic. Making it to defend the d4-pawn from b4 in the
almost impossible to bear is the fact middlegame.
that one of the games was a regular 2s ...l:.eS 26 nO ne6?!
one. Even now, decades later, I need 26 ... e4 would lead to an unclear
time to calm down when I think about position after 27 ~b 1 lDc4 28 'fic 1
it. It still upsets me that much! ~g5 29 lhc4 dxc4 30 .i.a2.
V.Smyslov - W.Schmidt
Warsaw-Moscow, 1980
1 e4 lDf6 2 es lDds 3 d4 d6 4 lDf3

~g4 5 ~e2 e6 6 c4 lDb6 7 exd6 cxd6
S h3 ~hs 9 0-0 ~e7 10 lDc3 0-0
11 l:.el a6 12 b3 lDc6 13 ~e3 ds 14 cs
White often gains space like this in
the Alekhine.
14 ... ~xf3 15 ~xf3lDcS 16 l:.bl ~f6 27 e4!
17 b4 A nice riposte.
27 ... exd4?
See diagram 011 page 143. After 27 ... dxe4 28 lDxe4 .i.g7 29 na4
f5 (29 ... exd4 is met by 30 lDd6)
17... lDSe7 IS .i.g4 30 ~bl White is better.
Smyslov fights to prevent Schmidt's 2slDxds ~es
knights from taking up positions in the
centre.
IS ... g6 19 a4 hs 20 bs axbs 21 axbs
lDas
The knight never comes back into
play.
22 .i.e2 lDfs 23 ~d3 lDxe3 24 be3
es
29 'iVf3
With remarkable ease, Smyslov has
gained space on the queens ide and now
wins the game on the other side of the
board.
29 ... f6 30 lDf4 ~xf4 31 1Vxf4 ~g7
32 %1a4 l:.es 33 'iVa ltxcS 34 l:.xd4
144
Vassily Smyslov the 7,h
1Ife7 35 'ii'xf6+! 29 .1Lxc6 Ibe6

The sixth world champion finishes Even Tigran Vartanovich did not
with an elegant petit combination. come up with double exchange
35 .. :~xf6 36 Iid7+ ~h6 37 Ibf6 sacrifices too often.
l:te5 38 h4 1-0 30 liJxe6 'Wixe6 31 'ifg3liJd7 32 'ii'c7
liJhfS 33 l:tbl ~g5 34 ~c1 ~xc1
Smyslov had another win with this
35 Ilexc1 g5 36 11ft g6 37 f4 gxf4
pawn structure.
38 I1xf4 ~g7 39 l:.bft f5 40 'ii'xe6
V.Smyslov - H.Azizi bxe6 41 g4liJf6 42 ~f2 lLl8d7 43 ~e3
Rilton Cup, Stockholm 1998 liJb8 44 1:412 liJbd7 45 gxf5 exf5
46 I1g2 ~n 47 nfgl liJfS 48 b7 liJ6d7
1 e4 liJf6 2 e5 liJd5 3 d4 d6 4 liJo
49 h4 1-0
~g4 5 ~e2 e6 6 e4 liJb6 7 exd6 exd6
8 h3 ~h5 9 0-0 il.e7 10 liJe3 0-0 11 b3
Here are some of my own games.
liJ8d7 12 ~b2 liJf6 13 1:et l:le8
14 liJh4 G.Kasparov - A. Yermolinsky
A surprising decision. UI8 USSR Championship,
14 ... il.g6? Vilnius 1975
Better is 14 ... il.xe2 because you are
supposed to exchange when you have a 1 e4 liJf6 2 e5 liJd5 3 d4 d6 4 liJfJ
disadvantage in space. ~g4 5 ~e2 e6 6 0-0 il.e7 7 h3 ~h5
15 liJxg6?! hxg6 16 .1Ld3 8 c4 liJb6 9 exd6 cxd6 10 liJbd2 0-0
The bishop is strong now. 11 b3 liJe6 12 ~b2
16... a6 17 'if0 'Wie7 18 l:lac1 'Wid7 To retain the bishop. On e3 it would
19 a4 d5 20 e5liJa8 have been vulnerable to ... liJf5.
To put it mildly, this is not exactly 12 ...il.f6 13 a3 d5 14 e5 liJe8 15 b4
classical chess. Of course I was satisfied. I thought I
21 b4 would just copy Smyslov's play.
15 ... a6
21.. ..1Ld8 22 b5 liJe7 23 b6 liJee8
24 as liJb7 25 ~e2 liJef6 26 ~a4 l1e6 161lc1liJ8e7 17liJb3 il.xfJ 18 ~xfJ
27 liJa2 ~e7 28 lLlb4 Ilfe8 liJf5
145
19 il.g4 remember any game with this motive

My predecessor put this bishop on from any champion. 28 .. Jhc3 29 %bcl
g4. I reckoned that had to be correct. l:tc4 30 b7 %:tb8 31 bxa6 lOc7 32 lOa5
19 ... g6 20 'it'd3 il.g7 21 Jte3 1t'f6 lOxa6 33 lOxc4 dxc4 34 lhc4 l:txb7
22 il.xf5 'ilhf5 23 'ilt'xf5 exf5 24 a4 35 d5 and White has the upper hand.
lOd8 28 ... bxe5 29 b6 exd4 30 il.b2 Aab8
31 a5
After 31 ':xc8+ ':xc8 32 l:tcl White
is still no worse.
31. .. l:te4 321Od2 l:txc1+
If 32 ....:b4 33 Jta3 l:tb5 34 .i.d6 Ad8
35 Aabl.
33 %bc1 ~f8 34 lOb3 il.e5?!
After 34 ... ~e7 35 il.a3+ <it'd 7
36 lOc5+ lOxc5 37 l:txc5 .i.e5 Black
does better than in the game.
25 b5
All goes according to the Smyslov
concept.
25 ... lOe6 26 :a1
I must admit that for just a second I
lost my discipline and omitted ... b6,
thereby deviating from Smyslov's plan.
I felt I could improvise. But a single
improvisation and I ended up losing -
what a harsh punishment!
26 .. .l:He8 27 lHd1 35 il.a3+ ~e8?
In principle it is right to bring the
king to the centre, however in this
particular position it has its tactical
drawbacks.
36 'ue1 f6 37 f4! lOxf4 38 il.d6 l:[d8
39 il.e7 d3?
This is a bad move in a bad position.
Black could have played on the
exchange down with 39 ... lOd3, but his
position has to be lost. 40 %:te2 ~d7
(40 ... lOb4 41 b7 lLlc6 42 lLlc5)
27 ... b6! 28 l:tdc1 41 il.xd8 ~xd8 42 Adl lOb4 43 lOxd4
White should have taken with and the protected b6 passed pawn is too
28 cxb6, but I was not able to strong to live with.
146
Vassily Smysloy the 7th
13 ... a6 14 .i.a4
14 ... h6 15 .i.e3
Vassily also developed the bishop on
e3 in one of his games.
15 ... lLle8 16 'ii'd2 <i!tb7 17lLlg3
I provoked him into pushing his
pawns, but maybe it was not in my best
40 <i!tf2? interests.
I not only missed a forced win with
17... e6 18 lLlge2 g5 19 .i.c2 .i.xc2
this move but actually squandered the
20 'ii'xc2+ f5 21 l:tabl lLlf6 22 a4 'ii'e8
full point. I no longer remember
23 .i.f2 'ii'g6 24 b5
exactly but I suspect I was in time
trouble. Had Smyslov published
annotations to his game I may have
been able to memorise it and play
faster, thereby avoiding this time
trouble blunder. The winning
continuation was 40 l:txe5+! fxe5 41 b7
d2 42 lLlxd2 lLle2+ 43 <i!tfl lLld4
44.i.xd8.
40 ... d2!
Oh, no. White has to resign.
24 ... lLla5
41 l:txe5+ fxe5 42 lLlxd2 lLld3+
The knight may become strong on c4.
43 <i!te3 lLlc5 44 .i.xe5 <i!td7 45 lLln
25 bxa6 l:txa6 26 1:b5 lLld7 27 1:tb 1
l:te8 46 <i!td4 lLlb3+ 41 <i!txd5 lLlxa5 48
lLlb8 28 lLla2 lLlbc6 29 lLlb4 1:a7
.i.f4 <i!tc8 0-1
30 lLld3 f4 31 'ifdl 1:[17 32 lLl62 1:a8
Here is my second game with the 33 h3 b5 34 'ii'd3 l:ta7 35 b4 g4
same pawn structure.
G.Kasparov - A.Dreev
Moscow peA -Grand Prix,
Kremlin Stars, Moscow 1996
1 c4 c6 2 e4 d5 3 exd5 cxd5 4 d4
lLlf6 5 lLlc3 lLlc6 6 .i.g5 .i.e6 7 a3 .i.g4
8 f3 .i.e6 9 c5 g6 10 i..b5 i..g7 nlLlge2
0-0 12 0-0 .i.f5 13 b4
The pawn structure again reminds me 36 'tixg6+ <i!txg6 37 lLld3 .i.b6
of Smyslov's. 38 .i.elltJc4 39 1:al gxn 40 gxn <i!tf5
147
41 a5 Ag7+ 42 ~hl :a8 43 il.c3 ttJe3 draws.) 70 ':xc6 d4 71 ~g2 (71 Ad6?
44 ltgl?! l:hgl+ 45ltJxgl :a7 46ltJe2 ~xf3 wins.) 71...l1c2+ 72 ~fl <it>e3
ltJe4 47 cwPg2 ltJ4xa5 48 il.xa5 lOxa5 73 lte6+ ~d3 74 c6 and White holds.
49 ltJe5 ltJe6 50 lOxe6 bxe6 51 ':'b6 67... ~g3 68 Ag6+ ~h3 69 Ah6+
':'c7 52 ~fl il.g7 53 ~e1 il.f6 54 Cjf;d2 ~g3 70 l:tg6+ ~xf3 71 l:be6 ~g3
72 l:.g6+ ~f3 73 e6 l:te2
After 73 ... ~e4 74 l::tg2 ':al +
(74 ... l:ta7 75 ':'c2 l:.c7 76 ~xh2 ~xd4
77 ~g2 draws) 75 <it>xh2 ltcl 76 Ag6
~xd4 77 ~g2 White draws.
54 ... il.xh4
Losing a second pawn should be the
end, but it was a rapid game so you
never know.
55 :b8 il.f6 56 :f8 h4 57 ltJgl 'itg6
58 ~d3 ':a 7 59 lOh3 Aa3+ 60 ~e2 74 l:th6?
~f5? Cutting off the king with 74 l:te6!
60 .. J~e3+! simply wins. was correct.
61 ltJxf4 <itxf4 62 l:hf6+ ~g5 74 ... <ifi'e4 75 l:th4+ ~d3 76 ':xh2
63 lbe6 h3 64 cwPfl :a2+ 65 ~gl h2+ If 76 l::tg4 ':xc6 77 ~xh2 ltc4 wins
66 ~hl ~f4 as well.
76 .•. l::txe6 77 l::th4
67l:H6+
67 l:txc6 was another option, but 77 ..•Ae2?
more fun was 67 l:1e2!. This nice Here 77 .. J~c4 78 ~g2l:hd4 79 Ahl
stalemate finesse probably saves the ~e2 80 <itg3 ':'c4 wins, as does
pOSItIon. 67 ... ':a4 68 l:te6 ':xd4 77 ... Ag6.
69 ~xh2 ':c4 (69 ... ~xf3 70 l:txc6 78 <ifi'gl l::te2 79 ~n ':e4
148
Vassi/y Smys/ov the 7rh
ending. In the games Karpov-Kasparov

Las Palmas 1996, Kramnik-Kasparov,
Intel rapid 1995 and in my match in
2003 against Azmaiparashvili, I made
serious mistakes, some of which
changed or could have changed the
result of the game. I did not make these
mistakes in pawn endings in a simul
against an amateur but against my arch-
rivals, Karpov and Kramnik. Yes,
80 l:1h2?? Smyslov and Levenfish really should
What a dreadful mistake! Even if it have written a book on pawn endings.
was a rapid game this should never 80 ... ~xe4 (80 ... dxe4 81 d5 or 81 ~el
have been played. In a way Smyslov is is also an elementary draw.) 81 <iite2
a guilty party for influencing me to 'i!i>xd4 82 'i!i>d2 resulting in one of the
conduct the opening the way he did. I best known drawn positions in chess.
must say Levenfish is also partly to 80 .•.<it>xd4 81 l:1d2+ ~c4 0-1
blame for this. Together they wrote a I miss a successful frontal attack by
classic book on rook endings, which I three tempi. My rook should check him
studied deeply. But they neglected to from eland my king should be on f2. I
publish a book on pawn endings. invested so many hours on his rook
Simplifying to a pawn ending with ending book and yet here I was not able
80 l:lxe4! was the solution to the to gain three tempi.
problem. I tried to erase this weakness By the way, this tournament was run
from my play but even in the very last on a knockout system - I won the next
game of my career I made a losing two games and reached the next round
move against Topalov in a pawn of the competition.
Once Smyslov planted a knight on I also planted his idea in my mind

c4. Here is the position: and reached the following position:
W.Fairhurst - V.Smyslov Y.Seirawan - G.Kasparov
149
G.Kasparov - V.Kramnik
I wanted to encourage my
opponent to place a knight on c4.
Kramnik was oblivious to the
dangers of having such a knight -
and beat me!
W.Fairhurst - V.Smyslov 22 ...1i'a7! 23 It)e2 dxe4 2A fxe4

Hastings 1954/55 Jl.g4!
Suddenly White's centre falls apart.
1 d4 ~f6 2 c4 e6 3 ~c3 ~b4 4 e3 25 ~xc4 bxc4 26 It)g3 1i'xd4+
0-0 5 ~ge2 d5 6 a3 ~e7 7 ~g3 b6!? 27 ~e3 'ii'd3 28 'ifb2
Smyslov claims he is always
searching for hannony but he often
goes in for unbalanced fights. Here
7 ... c5 was better, just to equalise.
8 cxd5 exd5 9 ~e2 ~b7 10 It)f5 l:e8
11 It)xe7+ 1i'xe7 12 0-0 It)bd7 13 b4 c6
14 ~d2 14 ... a6 15 'ifb3 b5!?
Smyslov's knight is heading for c4.
16 .:tae1 It)b6 17 ~c1 as 18 f3 axb4
19 axb4 ~c8 20 ~d3 Jl.e6 21 1Ifbl
It)c4 28 ... lt)xe4
The rest is uninteresting.
29 It)xe4 lhe4 30 1i'f2 f6 31 ~c5
Jl.e2 32 :al Ibe8 33 :fel c3 34 'iff5
c2 35 h4 'ifd5 361i'f2 h5 37 :acl ~dl
38 lbe4 l:be4 39 :al 'ji'e5 0-1
This Smyslov game was against a

relatively unknown player whereas I
used his idea against a genuine
contender. I should add that it was not
only Smyslov who made me think that
22 e4? a knight on c4 would be almost
White pushes forward without proper decisive but also two other
preparation. champions misled me in one game.
150
A.Karpov - B.Spassky
Game II, Candidates Semifinal,
Leningrad 1974
I d4 It)f6 2 e4 e6 3 It)f3 d5 4 lCle3

~e7 5 ~g5 h6 6 ~h4 0-0 7 e3 b6
8 ~e2 ~b7 9 ~xf6 ~xf6 10 exd5 exd5
11 0-0 'ii'd6 12 lIe 1 a6 13 a3 lCld7
14 b4 b5 15lt)el e6 16lt)d3lt)b6 17 a4
.i.d8 18 It)e5 ~e8 19 a5 ~e7 20 g3 15 ... b5 16 It)c5 a4 17 'Wc3 ttJb6
It)c4 18 It)d2 l:tae8 19 lIfe1 l:le7 20 ~f3
lUe8
N ow I prefer to transfer the bishop to
g6 after 20 ... g5!? and start pushing the
f-pawn.
21 g3 ~h3 22 .i.g2 ~xg2 23 ~xg2
f5 24 h4
24 ... lt)c4
I managed to pOSItion the knight
just like Smyslov and at this point I was
21 e4 .i.h3 22 lie 1 dxe4 23 It)3xe4
satisfied and thinking appeciatively of
'ii'g6 24 .i.h5 ~7 25 'ii'f3 f5 26 lCle3
him.
g6 27 'it'xe6 gxh5 28 It)d5 f4 29 lte7 25 It)f3 ~f6 26 l:te2 l:Ig7 27 %:thl
'ii'f5 30 l%xe7 l%ae8 31 'iVx.h6 lif7 'ike7 28 %:tee1 h6 29 'ii'd3 l:lf8 30 It)d2
32 l:txf7 ~xf7 33 'ifxf4 l:te2 34 'ife7+ 'ife8
~fB 35 It)f4 1-0
Y.Seirawan - G.Kasparov
Dubai Olympiad 1986
1 d4 It)f6 2 c4 g6 3 It)c3 d5 4 lClf3

~g7 5 .i.g5 It)e4 6 cxd5 ttJxg5 7 lClxg5
e6 8 It)f3 exd5 9 b4 'ii'd6
This move was a novelty then. 31 It)xc4
10 a3 0-0 11 e3 c6 12 .i.e2 .i.f5 White would have done better not to
13 0-0 It)d7 14 It)a4 a5 15 'it'b3 have taken the c4-knight.
151
31. .•dxe4 32 'it'dl l:te7 33 l:tefl 'iffi Objectively this neither spoils nor
34 'iff3 'jj'd5 35 'it'xd5+ exd5 36 !itO improves the position, but puts Black
il.g7 37 ltdl l:tffi 38 lId2 l:te8 into a situation where he has to find a
39 lIddl il.f8 40 ltdgl il.g7 41 l:tdl very subtle plan in order to draw. The
~f8 42 l:td2 cj;e7 43 ':ddl ~d6 practical move 55 ... fxg4 56 ltxg4 ':h5
44 l:tb2 !ite6 45 lIhhl il.f8 46 lId2 offered an equal endgame.
il.d6 47 l:tddl 56 exf4 l:txa3 57 fxg5 l:a2+ 58 cj;f3
47 ... il.xe5?!
An impatient move. I should have 58 ... e3?
further improved the positions of my This natural move loses. The c-pawn
other pieces, for example by 47 ... J:tee7. is closer to promotion than White's g5-
Then after 48 l:td2 (48 l:h2? il.xc5! pawn, but Black's rook has less effect
49 dxc5 l:te4 wins.) 48 ... l:h7 49 l::tddl on it than White's on c3. And that
g5 Black can exert pressure. matters at this point. Better was
48 dxe5 l%e4 49 l:be1 lId7 50 l:td4 58 ...l:a3+! 59 'ifrg2 (59 ~i4 l:ta2)
g5 51 hxg5 bxg5 52 ltedl l:txd4
53l:bd4 l:th7 54 We2
54 g4 was interesting.
54 ... l:th3
If 54 ... l::thl 55 l:tdl.
55 g4
59 ... .:a2!!
This pin of the f-pawn is an
extremely difficult move to find. It
temporarily stops White pushing his
passe4 f and g-pawns as a team. It also
55 ... f4?! gains a tempo to help win the race.
152
(Not 59 .. J1al? 60 g6 l:tel 61 f4! Two 63 :tf7+ 'i!i>c6 64 <itif3 a3 65 na7 d4

connected passed pawns are often and this unusual position with two
advanced together. It is best not to connected passed pawns in both camps
separate them by sending only one out is probably equal.
in front. 61.. .c3 59 ndl d4
62 l1d 1!! I not only investigated the 60 g6 d3

champions' games but also those of Many champions won games by
other world class players. How hard it pushing two passed pawns all the way
was to balance my time between them! to the sixth rank. The c4 knight
Timman also liked to give up his rook on that I planted like Smyslov was later
the back rank as in his games against transfonned into this pawn - but in the
Ivkov at Amsterdam 1971 and Kramnik at end its slow pace is responsible for my
Belgrade 1994. The rook can't be taken as defeat.
White surprisingly checkmates Black's 61 'i!i>e3lhfl
king after queening his g-pawn first, After 61.. .d2 comes 62 g7 c2
as I first pointed out in my Informant 63 'i!i>xd2 cxd 1='iV+ 64 ~xd 1.
analysis. Alternatively: 62 g7 1-0
a) 62 ... lte2+ 63 <itig3 [63 <itif3 l1e8 I will not analyse in detail my
64 f5 d4] 63 ... d4 64 g7 l::te8 65 f5 c2 encounter with KIamnik as it is just a
66 :fl d3 67 f6 d2 68 f7 dl=fi' blitz game. I happily let my opponent's
69 fxe8='iV+ wins. knight go to c4 and when it arrived
b) 62 ... l1e8 63 f5 d4 64 f6 c2 65 l1fl there I could not resist a little smile.
d3 66 f7l::td8 67 g7 d2 68 f8=.+ dl=fi'
69 fi'f6+ <itic7 70 g8='ii' :txg8 G.Kasparov - V.Kramnik
71 'itb6+ <itic8 72 'ii'e6+ <;i>b7 73 nf7+ Champions Club, 5 minute game
wins. Or 63 g7 c2 64 g8='iV l:tgl + Kasparovchess.com 2001
65 <itih3 l:th 1+ 66 ~g3 l:tg 1+ 67 'it;lh4 1 d4 liJf6 2 c4 e6 3 liJo b6 4 tDc3
~c7 [67 ... cl='iW 68 it'c8 mate] ~b7 5 a3 d5 6 cxd5 tDxd5 7 e3 g6
68 'ii'f7+ <;i>c6 69 'iVe6+ Q;c7 70 'ii'e7+ SliJxd5 exd5 9 ~b5+ c6 10 ~d3 ~g7
~c6 71 'ii'd6+ <itib7 72 'ii'd7+ <itib8 11 b4 0-0 12 ~d2 ttJd7 13 nb 1 l::teS
73 c6 White catches Black's king.) 14 0-0 ttJf6 15 'it'c2 ttJe4 16 nfc1 l:lcs
60 g6 (60 <itig3 c3 61 g6 l:te2) 17 ~el ttJd6 18 a4 a6 19 tDd2 nc7
60 ... l:te2 61 l:tf4 neB 62 :f6+ ~c7 20 ttJb3
153
38 g4 'ilt'g5 39 ~xh3 ~xd4

40 ~hl?! lLle5 41 'ilt'g3 lLlxg4
42 :xe6?
20 ... b5 21 lOe5 ~e8 22 11al 1%ee7

23 ~e3 lOe4 See diagram on page J5 O.
24 :a2 'it'd6 25 axb5 axb5 26 1%a8
lOb6 27 ]:ta2 h5 28 l:teal h4 29 'it'dl
h3 30 g3 lOe4 31 ~f1 ~h6 32 :a8 42 ... ~xe5 43 ]:txe5 'ii'h5 44 ii.g2?
~f5 33 l1xe8+ 1%xe8 34 ~d2 'it'f6 lOf6? 45 f3?? lLle4 46 'fIe7 lOxe5 47
35 ii.c1 ~g7 36 :a6 ~h7 37 'iff3 ~g8 bxe5 ~h3 0-1
Smyslov and I both played the Griinfeld quite regularly. Assessing the
strength of the d6 passed pawn is not always a simple matter. I knew his win
against Euwe, so I also went for a variation in the Griinfeld with a d6-pawn.
Very sadly the result was not 0-1 as in Smyslov's game.
Actually the position I reached against Piket was virtually the same.
M.Euwe - V.Smyslov J.Piket - G.Kasparov
M.Euwe - V.Smyslov Nowadays players who develop the

Candidates Tournament, Zurich 1953 e2-knight on c3 go there at once with
10 ltJec3.
1 d4 lLlf6 2 e4 g6 3 g3 i.g7 4 ~g2 d5 10 ... ltJa6 11 ltJa3 exd5 12 exd5 ii.f5
5 exd5 lOxd5 6 e4 lOb6 7 lLle2 e5 13 tOe3 ltJb4
8 d5 e6 9 0-0 0-0 10 a4 Black bases his strategy on the fact
154
that he has lovely piece play. However Smyslov makes it so easy to remove
White's d-pawn can become dangerous the d6-pawn. A fascinating fight starts
as if he can push and then consolidate it in the centre.
on d6, it could stifle Black. 24 ttJf6+
14 .i.e3 lIeS If 24 ttJxd6 'ifa6+.
24 ... ~hS 25 .i.d4 .i.e5 26 ttJd7
15 d6 26 ... f6?!
I got the impression from this Smyslov takes a huge risk. After
particular game that the d6-passed 26 ... .i.xd4 27 'ii'xd4+ c;t>g8 28 ttJf6+
pawn is not something Black can't c;t>f8! (Euwe spotted that the natural
handle, especially in the Griinfeld.
continuation 28 ... c;t>h8, allowing a
15... .i.d3 16 ~xb7 lIbS 17 .i.g2
battery, gives more than just a
i.xfl IS c;t>xfl
perpetual: 29 ttJd5+ ~g8 30 ttJe7+ ~f8
White has reasonable compensation
for the exchange. 31 'iWh8+ ~xe7 3211el + ~d7 33'ti'd4+
lS ... ttJd7 19 ttJe4 ttJd5 34 .i.xd5 lIb4 35 'ti'e5 'ifa6+
36 c;t>gl iid6 37 'ii'c3) 29 ttJxh7+ ~g8
30 ttJf6+ ~f8 the players have to settle
for a repetition.
27 .i.xe5 fIe5
19 ...ttJe5
Smyslov exchanges to get closer to
the d6-pawn.
20 ttJxe5 .i.xe5 21 .i.xe5?!
After 21 ttJe4!? ~a5 22 ttJxc5 2S'ti'd2?
21..JWa5 22 .i.e3 llfdS 23 ttJe4 Going after Black's king with
i.xd6 28 'ii'd6 would force Black to return the
155
exchange and settle for a position a J.Piket - G.Kasparov

pawn down. 2S .. JlbcS (28 .. Jlxd7 Euwe Memorial, Amsterdam 1995
29 'ii'xd7) 29 'ilr'f6+ <ifi>g8 30 'iVe7 ':xd7
31 'iVxd7 and White should win this 1 d4 lOf6 2 c4 g6 3 lOc3 d5 4 ttJO
without too much of a problem. ~g7 5 'ifb3 ID:c4 6 'ii'xc4 0-0 7 e4 lOa6
28 ....:r.bc8 29 <it>gl? 8 ~e2 c5 9 d5 e6 100-0 exd5
29 1t'd6 was still better than the
game. After 29 .. :.a6+ 30 'ii'xa6 ttJxa6
3 1 ttJxe5 it looks balanced.
11 exd5 ':e8
In an earlier game - with 11 ... ~f5 -
the d-pawn did not move yet assumed a
29 ... 'ii'c5! great role just by threatening to move.
He makes sure White doesn't get out Sadly, I lost that one as well. 12 li.f4
of the pin. lIe8 13 l:tad 1 lOe4 14 lOb5 'iff6 15 ~d3
30 ~h3 'ii'e7 31 'ii'e2 lOb4 16 lOc7 lOxd3 17 lOxe8 l:he8
If31 ':dll::tc7. 18 'ifxd3 'ii'xb2 19 lIde 1 'iib4 20 ttJd2
31..Jlxd7 'ifa4 21 'ii'c4 'ii'xc4 22 lOxc4 ~c3
Black wins a piece. Euwe resists but 23 1Od2 ~xd2 24 ~xd2 ~d7 25 li.f4
in the long run he has no chance. ~b5 26 f3 g5 (26 ... ~xfl 27 <it>xfl lOf6
32 ~xd7 'i!fxd7 33 'iVxe5+ ~g8 28 l:lxeS+ lOxe8 29 ~e5 ttJg7 30 d6
34 1t'e4 a5 35 h4 'ii'd5 36 'ifg4 l::trs Now the d6-pawn wins.) 27 li.xg5
37 .:tdl'ifO 38 'ii'c4+ 'ifn 39 'ii'c5 'it'f5 ~xfl 28 <it>xfl lOd6 29 ~e7 lOc8
40 'ifc4+ 'iff7 41 1t'c5 'iff5 42 'ifc4+ 30 ~xc5 l:td8 31 .:te5 f6 32 .:tf5 b6
<ifi>g7 43 'ifd4+ 'ii'f6 44 1t'c5 %lf7 33 li.d4 lOe7 34 li.xf6 l:txd5 35 .:tg5+
45 ':d2 'ife7 46 'it'c3+ .:tf6 47 %ld4 %lxg5 3 6 ~xg5 ttJc6 37 We2 Wfl
ttJc6 48 ':d5 "e6 49l:[c5 h5 50 b3 <it>f7 38 ~d3 ~e6 39 ~c4 lOe5+ 40 <ifi>d4
51 l:[b5 'iWd7 52 ~g2 'ii'e7 53 'ii'c4+ lOc6+ 1-0 Karpov-Kasparov, London!
<ifi>g7 54 'it'd3 ~h6 55 .:td51If7 56l1d6 Leningrad 1986.
ttJe5 57 'iVe3+ ~h7 58 .:tb6 'ifc7 0-1 12 .:tdl ~f5

I suspected I would face the variation
in the next game and was hoping I 13 d6 b6
could prove I was able to neutralise the Piket has his own lines, which he has
d6-pawn at least as well as Smyslov. developed and refined to a high level. I
156
prepared this move to counter his match: 19 ... llJd3! 20 .i.g3 c4 21 "WIc2
preparation. Gulko defended the nc8 22 %:tadl 'ii'd7 23 h4 f5 24 Axd3
position differently with 13 ... llJe4. !h-Ih Karpov-Kasparov, Game 21,
Then 14 llJb5 ~d7 15 a4! llJb4? Seville 1987. And I had also come up
16 'tlfb3! 'tlfb6?! 17 iLe3 iLxb5 against 19 iLg3 in a quite different kind
18 ~xb5 llJc6 19 d7 l:ted8 20 'it'c4 llJf6 of event: 19 ... "WId7 20 a3 ~c6 21 'iWb5
21 iLxc5 "WIc7 22 ~xc6 bxc6 23 .i.d6 nc8 22 nad 1 ~xc3 23 bxc3 llJe5
'iixd7 24 llJe5 1-0 Piket-Gulko, 24 'ili'xd7 and I went on to win in
Groningen 1990. a simultaneous exhibition game, Rao-
14 ~f4 Kasparov, New York 1988.
I had already played against 14 h3?! 19...ne4?
when I managed to show that 19 ... ~xa4 deserves consideration.
Black's pieces work well and he can 20 ~g3
even take over the initiative with his After 20 ~e3 llJc6! 21 ttJxc5 l:tb4
piece play. 14 ... llJb4! 15 ~f4! ~d7 22 'iid3 llJc4.
16 nd2 a6 171ib3 b5 18 "WIdl c4 19 a4! 20 ... ttJc4
llJc5 20 axb5 llJbd3 21 ~xd3 llJxd3 If 20 ... llJa6 21 ttJxb6 'ifxb6 22 'ii'c2.
22 l:txd3? cxd3? (22 ... ~xd3!) 23 llJd5! 21llJxc5
axb5 24 llJe7+! Ih-Ih Karpov-
Kasparov, Game 15, World Champion-
ship, Seville 1987
14 ...llJd7 15 l:d2 ~b4 16 'it1J3 ~e6
Just like Smyslov I exchanged pieces
around the d6-pawn in order to weaken
it.
17 ~c4 llJb6 18 ~xe6 l:txe6
21. .. ~xd2
Everything goes according to plan.
Just like Smyslov I win the exchange.
22 llJxd2 l:te2
Here 22 ... nd4 23 llJf3 l:txd6!
(23 ... .J::.g4 24llJe6 wins.) 24llJxb7 nd3
25 'ifxb4 (25 llJxd8 l:txb3 26 axb3
19 llJa4!? l:txd8s) 25 ... ndl + 26 nxdl 'ihdl +
Tal recommended this and I lost 27 ..-e1 and White consolidates his
against it. This is one more example material advantage and has decent
where a world champion had a chances to win with the two pieces
controversial effect on my career. against the rook.
I had faced 19 a3?! in a world title 23 'iixb4 as 24 'iixb7 l:txd2
157
Here 29 ... .:.a7 30 ltJe4 wins.

Alternatively 29 ... 'irfS 30 ltJe4!
'ifxd6 (30 ... l:td8 31 ltJf6+ rJ;g7 32 'ii'd4
l:[f5 33 ltJe8+ ~h7 34 g4 l:b5
35 ltJf6+) 31 ltJxd6 :bb8 32 :cl l:td8
33 l:tc8 rJ;g7 34 l:txa8 wins.
30 h3 :bl 31 l:txbl l:txbl+ 32 rJ;h2
:b6
If 32 ...:b8 33 'ii'e5 ~f8 34 ltJe4 :b7
25 d7 35 'ii'h8+ q;e7 36 'ii'e8+ wins.
Here some doubt came into my mind 33 'ii'e5 ~IS
as Euwe's d-pawn never got this far, And here 33 ... J::tb8 is countered by
but I still did not worry. In reality Black 34 ltJb7!.
already has a lost position. 34 'ii'h8+ q;e7 35 'ire5+ litIS 36 f4
25 ...:xb2 26 'ii'd5 l:tb5 h5
After 26 ... R.f8 27 ltJe4! intending After 36 ... g5 37 f5 'iib8 38 'ii'xb8+
R.e5 is better than 27 ltJd3 l:d2 28 R.f4 l:txb8 39 f6 ':d8 40 rJ;g3 wins. And
l:ta7! 29 R.xd2 l:txd7 30 'irxa5 l:txd3 after 36 ... f6!? 37 'ii'd5 q;e7 38 h4!
31 'iVxd8 :xd8 32 R.e3. White keeps Black under pressure.
27 l:tdl Jtf8
After 27 ... :a7 28 'ii'e4! R.f8
(28 ... Jtf6 29 'ii'e8+ 1;g7 30 Jte5!)
29 'ii'e8 :a8 30 ltJe4 White wins.
28 R.d6 R.xd6
Not 28 ... l:ta7? 29 R.xf8 ~xf8
30 'ii'e5.
29 'ii'xd6
37 'ird5 h4
Now the pawn ending will not bring
victory:
a) 37 .. 5J;e7 38 'irg5+.
b) 37 ... g5 38 ltJd3.
c) 37 ... a4 38 ltJb7 l:hb7 39 'irxb7
~e7 40 ~g3 and Black can't enter the
29 ...:ab8 pawn ending.
Now the d-pawn is almost 38 'ireS!
suffocating me. Black treads a narrow After 38 ltJb7 comes 38 .. Jhb7
path but it is not enough to stay in the 39 'ii'xb7 ~e7.
game. 38 ... g5
158
Vassily Smyslov the 7'h
This is a desperate attempt, but it Here I resigned because I saw that I

can't loosen White's grip. There are had run out of plausible moves. I never
three other possibilities. got rid of that damned d6 pawn.
a) If 38 ... f6 39 'ir'e I!.
The conclusion of the game mght
b) 38 ... l:c6 39 '6b8+ ~e7 40 1h.h4+
have been 41...l::lg6 (If 4 1...l::tc6 42 g3.)
WIns.
c) 38 ... l:b4 39 '6b8+ <!;e7 40 'iVxh4+ 42 <!;hl (Black has no move.) 42 ... l:.b6
f6 41 'ii'e 1+ <!;d6 42 'iVe8 wins. 43 g6 lhg6 44 'ifh8+ rl;e7 45 'iVxh4+
39 'ilfb8+ <!;e7 40 'ilfeS+ <!;f8 41 fIgS nf6 46 ttJe4 and wins.
1-0
An attacking player like me often obtains posItions with unbalanced

material. Therefore it quite often happened that my opponent had several
connected pawns, while I preferred to have a piece. In this way Smyslov had
a very strong influence on me. Two of his games impressed me very much.
Here are a couple of critical positions from his games. Then come my games.
V.Smyslov - D.Bronstein G.Kasparov - B.Spassky
V.Smyslov - J.Timman G.Kasparov - J.Lautier
159
This is the third of my losses V.Kramnik - G.Kasparov

where there was unbalanced
material and a mass of pawns
coming in my direction.
No mistake, I lost three games

in this way! In fact I may have
lost even more, but (luckily for
me) I can't remember them.
V.Smyslov - D.Bronstein See diagram on page J59.

USSR Championship, Moscow 1951
15 ... ~e6
1 e4 cS 2 ttJc3
Smyslov played 56 Closed Sicilians
in his career. His first game and first
win with it came in 1946, his last was
in 1998. His first victim was Kottnauer,
his last Arakhamia. 52 years had
elapsed between these two games.
2 ... ttJc6 3 g3 g6 4 ~g2 ~g7 5 d3 d6
6 ~e3 ttJh6 7 'ii'c1 ttJg4 8 ~d2 ttJd4
9 h3 ttJe5 10 ttJce2 ir'b6 11 f4
16 ~c3
Smyslov prefers to face the three
pawns with a piece, rather than have a
rook against four pawns. After 16 :'xb2
comes 16 ... ttJxb2 17 ~c3 ttJd3
18 ~xh8 f6 19 ~g7 h5 20 g4 r;;;n
21 ~h6 .li.xa2.
16... ~xa2?!
After 16 ... .li.xc3 Bronstein avoids the
equation of three pawns against a piece.
11...ttJxc2+ 17 ttJxc3 .li.c4 18 ttJge2 (18 lhb7
This is a cute move, but not really a liJb4+ 19 ~e 1 .li.a6) 18 ... .J:tb8 and it is
good deal. hard to pick a colour. Everyone to their
12 'it'xc2 1rxb2 13 'ii'xb2 ttJxd3+ own taste.
14 ~fl ~xb2 15 libl 17 :xb2 liJxb2 18 ~xb2
160
18.. Jtg8 37 tLlxd6

White has three pieces against a rook Smyslov not only takes a pawn, but
and four pawns. This kind of position with his enonnous piece power also
with unbalanced material is really hard very quickly catches Bronstein's king.
to judge. I have already mentioned one 37 .. Jbb3 38 ~eS 1:a8
painful experience against Anand in He does not see a checkmate in two,
Tilburg 1991; in that game he had a but his position was hopeless anyway.
39 tLlc4+ 1-0
queen, I had three pieces.
i9 ~f2 ~c4 20 tLlo ~xe2
V.Smyslov - J.Timman
Black has invested two moves to
Hoogovens, Wijk aan Zee 1972
exchange the bishop for the knight. He
could have made two pawn moves 1 tLlo g6 2 e4 ~g7 3 d4 d6 4 tLlc3
instead. ~g4 S ~e3 tLlc6 6 ~bS a6 7 ~xc6+
21 ~xe2 ~d7 22 l:dl as 23 tLleS+ bxc6
Interestingly, Smyslov himself took
~c7 24 tLlxf7 a4 2S eS a3 26 ~al
on doubled c-pawns with Black against
1:ge8 27 llJgS 1:aS 28 tLle6+ ~d7
Timman in 1984. Even more
interestingly, several decades earlier in
1943, Smyslov also had a similar
position against Botvinnik. He lost the
first and drew the second game.
8 h3 ~xO 9 'il'xo e6 10 eS! tLle7
29 ~dS
White simply has too many pieces.
29 ... a2 30 g4 l:c8 31 tLlgS J:trs 32 fS
gxfS 33 gxfS b6 34 ~e6+ ~c7
3S exd6+ exd6 36 tLle4 11a3
161
11 ltJe4 26 ... e5 27 ltJe2 'it'd2?!

Smyslov plays powerfully. Timman adopts a risky approach. It is
11... ltJd5 12 .tg5 'iVb8 13 0-0 h6 dangerous to release the pressure on the
14 .tf6 .txf6 15 exf6 'ii'xb2 16 e4 f6-pawn.
'ii'xd4 17 exd5 exd5 28 <it7n 'ii'xs2
Black has three pawns for the piece.

If they start rolling they will be like an
avalanche. On the other hand the
choking f6-pawn can cause a lot of
problems for Black. Interestingly,
seven years later Smyslov, in the
9 ltJbd2 variation, introduced a very
important novelty which established a 29 'it'e3! ~h7 30 'iVb6
new variation. In that line Black also Of course White's queen is free to
sacrifices a piece for three passed invade.
pawns. I beat Shirov in that line (I had 30 ... 'irb3 31 ~g2 d4 32 'ifxd6 'ife6
the piece of course) in Linares 2001. 33 'iWxb4 'ifxf6
However the line is still playable for
Black nearly three decades after
Smyslov first played it.
18 ltJe3 0-0 19 l:1adl 'ife5 20 g4
l:tab8 21 l:tfel 'ilfg5 22 l:tbl
34 ltJg3!
The •three pawns for the piece'
balance is restored, but Black's pawns
are no longer in a mass, but separated.
The knight will really dominate and
22 ... c5 sooner or later White will penetrate.
Finally the first pawn of the herd 34 .. :ife6+ 35 liJe4 %;Ie7
moves. After 35 ... f5 36 gxf5 gxf5 37 'ife7+
23 <it7g2 :b4 24 l:txb4 exb4 25 ltJe2 wins.
l:1e8 26 ltJd4 36 'iff8 d3
Smyslov provokes Black into 36 ... g5 would have resisted for
pushing the pawn. longer.
162
37 Wh2 d2 If 15 ... fxg6 16 'iVh3 'ii'g5+ 17 <i!tbl

lhb2+ 18 'iii>xb2 ltb8+ 19 'it>a 1 ltJb4
20 ir'xe6+! (In my Informant analysis I
gave 20 a3 as winning. I can no longer
remember why, since it only leads to a
draw. 20 ... tLlc2+ 21 Wa2 iVa5 22 tLlb 1
[22 a4?? 'tlrb4] 22 ... ltJb4+ 23 'it>al and
Black has a perpetual.) 20 ... <i!th8
21 a4!! This move wins. 21...ltJc2+
(21..."1IVa5 22 l:ld2) 22 <i!ta2 ltJb4+
(22 .. .'ii'a5 23 'ti'xg6) 23 <i!tbl! and
38ltJxd2 White wins as there is not enough
The strongest pawn in Black's power in the battery.
position has now disappeared and with 16 ltd2 ne8 17 ltJg1! dS 18 tLlf3 as
it all hope goes too.
38 ... iff6 39 ltJe4 "f4+ 40 'it>g2 ':c2
41 iVd6 as 42 iVdS 1-0
Here are my exciting games featuring

the topic of a piece against three
connected passed pawns.
G.Kasparov - B.Spassky
Niksic 1983
1 d4 ltJf6 2 c4 g6 3 ltJc3 ~g7 19 eS?!

Spassky doesn't often play the Better was 19 ltd.h2! dxe4 (19 ... ltJe7
King's Indian. 20 lth8+!! is rather similar to Spassky's
4 e4 d6 5 f3 ltJc6 6 ~e3 a6 7 tLlge2 ... nh 1 against Larsen. 20 ... .i.xh8
l:lb8 8 'ii'd2 0-0 9 h4 bS 10 hS bxc4 21 ':xh8+! ~xh8 22 'irb6+ <i!tg8
11 g4 iLxg4!? 12 fxg4 tLlxg4 23 ltJg5 and White is about to
checkmate.) 20 ltJxe4 f5 21 ltJc3 liJxd4
See diagram on page 159. 22 tLle5 White wins.
19 ... l!Je7 20 iLh3
Black has three pawns for the piece 20 ltJa4 was not a clear win, though
and no real weaknesses - perhaps only White is better after 20 ... ltJf5 21 "f4
the h-file. I knew how much Smyslov fife7.
preferred to retain the piece. Black's 20 ... cS 21 dxcS
pawns are a long way from promotion. Black has only two pawns, but the
These two factors made me confident. path for his d-pawn is open. If
13 O-O-O! tLlxe3 14 'ihe3 e6 15 bxg6 21 l:ld.h2 ltJc6.
hxg6 21. .. 'ifc7 22 'iif4 ltJc6! 23 .:tel d4
163
31 'ifh7+ <it>f8 32 ltJxf7! wins.]

31 'iib7+ ~f8 32 ltxf7+ [32 'jixg7+
~xg7 33 l:txf7+ ~h8 34 l:th7+ Wg8
35 ltJf6+ ~f8 36 ltJxe6 mate] 32 ... ~e8
33 ltJf6+ ~d8 34 ltJxe6+ ~c8 35 l:tc7
mate) 30 'ii'h7+ ~f8 31 ltft ltc3+!!
This incredible move saves Black.
32 bxc3'ii'a3+ 33 ~dl l:td8+ 34 ltJd2
l:txd2+ and Black has a perpetual.
29 ... ~f8 30 ltJxe6+ fxe6 31 ltf1+
24lbd4 ~e8 32 ~g8+ ~f8
I knew Smyslov played 24 ltJe4, but
here Black gets nice play after
24 ... ltJxe5 25 ltJf6+ ~xf6 26 ltJxe5 c3.
24 ... ltJxd4
Black recovers the exchange.
Although the material no longer
favours White, he still has a dangerous
attack.
25 ltJxd4!
After 25 ii'xd4 'ifb7 26 ~f2 ~h6+
27 ~bl l:ed8 28 c6! ~xc6 29 ~g2 the 33 'ikxg6+
position is balanced. I did not see the possibility of
25 ... 'ihc5 26ltJf3! l:.ed8 33 ltJd5!! which forces a draw. Then
If 26 ... 'ilVb6 27 lte2 l1ed8 and I 33 .. .'~t(d7 taking the pawn would lead
preferred Black's position even a good to the same perpetual. (33 ... exd5
two decades ago. 34 lhf8+ 'ikxf8 35 'ii'e6+ ~e7
27 ltJg5! 'iie7? 36 'ti'g8+; 33 ... l:txd5 34 l:hfB+ ~xfB
White has no more than a draw after 35 ~xe6+) 34 ltJxe7 ~h6+ 35 ~c2
27 ... ltd7 28 ~xe6 fxe6 29 'it'h2 ~h8. lIxg8 36ltJxg8 lhh3 and the endgame
28 ~h4 ltd3 29 'ikh7+ should end in a draw.
Here I started to have problems with 33 ... ~d8 0-1
my time. The situation has changed, Here I lost on time. Maybe I spent
so, just like Smyslov, I might have too much time trying to work out the
considered 29 ltJce4! Sometimes I similarities between this game and
copied the champions too much, Smyslov's. But the position is lost
sometimes not enough. They should anyway.
have written instructions on their For example: 34 .i.xe6 'ifb4! (After
games. If you buy a television you get 34 ... l:tb6 35 'ikg8! l:txc3+ 36 ~bl!
a manual. 29 ... ~xe5! I missed this l:txb2+! 37 ~xb2 ~a3+ 38 ~bl l:tb3+
move in my 1983 In/ormant analysis. draws nicely.) 35 ltJa4 (If 35 ltJdl
(29 ... c3 30 ltfl! 'ilVb4 [30 ... l:tb6 'ii'd2+ 36 ~bl :tn! 37 lIgl [37 l:.hl
164
l':tf2] 37 ... l:lf2 Black wins.) 35 ......d2+ Black wins.) 19 ... ne8 20 'lifl and
(Or 35 ... iLe7! 36 ~f5 "'xa4.) 36 ~bl Black is doing fine here.
:f3 37 :c 1 .i.h6. 17 exf6 1i'xf6 18 ttJb3
G.Kasparov - J.Lautier See diagram on page 159.

Linares 1994
Remembering Smyslov's example
1 e4 e5 2 ttJO ttJc6 3 iLc4 ~c5 4 c3 retained the piece, but other
ttJf6 5 d3 d6 6 ~b3 h6 7 h3 a6 8 ttJbd2 continuations, like 18 'it'd3 g6, offered
iLe6 9 iLc2 i.a7 10 'ii'e2 'ii'e7 nothing but gloom.
Black intentionally holds back 18... ttJxb4
castling. Four pawns may not be one too
11 b4! d5 12 a4 b5! 13 0-0 0-0 many. White has chances to block them
14 axb5 axb5 as there are holes on Black's queenside.
19 iLbl
After 19 .i.e3 .i.xe3 20 fx.e3 ttJxc2
21 'ii'xc2 b4 22 ttJfd4 'ii'g5 favours
Black; 19 i.a3 ~xc2 20 'ii'xc2 nfd8
and the d-pawn can't be blocked.
19 ... d4!
Not 19 ... c5? 20 .i.a3.
15 d4
Opening up the centre doesn't favour
White, and other options offer nothing
either. If 15 .i.b2 :fd8; 15 exd5 ttJxd5
16 iLb2 ttJf4 17 'iie4 i.d5.
15 ... exd4
Not 15 ... dxe4? when 16 ttJxe5 is 201ba7?!
better for White. In such a complicated position it is
16 e5 natural that players cannot always fmd
Other moves were also harmless for the best moves. The best choice was
Black. 20 .i.a3! This extremely complicated
16... dxc3!? position could take pages of analysis,
I saw that Black has time to step but for now I'll just show the best
aside with 16 ... iLd7!. Then 17 cxd4 defence for White. 20 ... d3 21 i.xd3
(17 nel dxc3!) 17 ... ~xd4! 18 nxa8 (21 'iWe4 iLxb3 22 iLxb4 ~xf2+!
(18 'ii'd3 ~xe5 wins) 18 ... .J:[xa8 19 ne I 23 ~xf2 l:lxa 1 24 i.xd3 'iib6+ 25 ~g3
(19 ttJxd4 ttJxd4 20 'ii'd3 1i'xe5 and f5 26 'fIe7 ~g6+ 27 ~h2 White is still
165
in the game.) 21.. .liJxd3 22 ~xf8 liJf4! These kinds of positions are harder to
23 'ir'xb5 liJxh3+! 24 ~hl liJxf2+ play in a rapid game. Maybe that's an
25 ~g 1 and White is still alive. excuse for my loss. Because it was a
20 ... e2 rapid game our analysis was limited.
Good is 20 .. Jlxa7! 21 liJbxd4
(21 liJfxd4? ~c4 22 'iWe4 'if g6! 23 liJf5 V.Kramnik - G.Kasparov
c5!) 21..Jlal! 22 'iWe4 'ir'g6! 23 'iWxg6 peA Intel-Grand Prix, Moscow 1994
fxg6 24 liJxe6 l:tf6 and White is in
trouble. 1 liJf3 liJf6 2 e4 g6 3 liJe3 ~g7 4 e4
21 lba8 exbl='ir' 22 ':xf8+ ~xf8 d6 5 d4 0-0 6 ~e2 e5 7 d5 liJbd7
8 ~e3 liJg4 9 ~g5 f6 10 ~b4 h5
11 liJd2 liJh6 12 f3 liJn 13 'tie2 ~h6
14 0-0-0 e5 15 dxe6?!
15 ~b 1 came into consideration.
15... bxe6
23 'ir'xb5??
This is a bad mistake or should we
say that it misses the opportunity to
playa great move. The position was so
exciting that I forgot about Smyslov
completely. The surprise is 23 ~g5!! 16 ~bl as
'ii'fg6 (23 ... 'it'xfl + 24 'iWxfl hxg5 Better was 16 ... liJc5!.
25 'it'xb5 liJd5 26 liJbxd4!=) 24 'ir'xb5 17liJa4 c5 18liJc3 ~e3?
~g8 25 1ib8+ (25 l:lxb 1 'i!Vxb 1+ There is no need to think about
26 ~h2 'iWxb3 27 liJxd4 White is moving the bishop to d4.
a pawn behind, but it is not easy
19liJd5 ~d4 20 liJb3 ~b7 21liJxd4
to do anything with the extra pawn.)
exd4 22 f4 l:lb8 23 AbO liJb6 24 e5!?
25 ... ~h7 26l:hbl 'it'xbl+ 27 ~h2liJa6
~xd5 25 exd5 liJf5
28 'Wi'a7 ii'xb3 29 liJxd4 'il'c4 Despite
Black's extra material White can resist.
23 .. .'iixb3
White has only a rook against
Black's queen.
24 'iWb8+ rj;e7 25 'ir'xe7+ rj;e8
26 ~d2 'iWd8
Better was 26 ... liJd3!.
27 'ir'e5 ~f8 28 liJxd4 liJd3! 29 'ir'e3
'ir'e4 0-1
166
Vassi/y Smys/ov the 7,11
26 fxeS! 32 llg4+! ~f8

Kramnik sacrifices a piece. His
pawns are closer to promotion than
those of Smyslov's opponents.
26 ... tOxh4
If 26 ... tOxe5 27 l:txf5 gxf5 28 c6.
27 exd6 tOeS 28 lhd4
28 ... liJfS
After 28 ... l:tb4 29 llxb4 axb4 30 'ii'e4
tOf5 31 'ifxb4 the five queenside pawns 33 'ii'e6! l:1b7 34 c6
might be too much to handle even for White's central pawns are just too
Smyslov. much to bear.
34 ...l:txb2+
The rook sacrifice causes some
tension, but not much else.
35 'it>xb2 'itb6+
29 lhfS! gxfS 30 llhfS

Black has a rook for the pawns, but it
is hard to make a breach.
30 ...'it>g7
After 30 .. J:tb4 31 1li'e6+ <&tih8
(31 ... liJfl 32 l:txb4 axb4 33 d7 36 'it>a3!
[33 .i.x.h5 is also sufficient.] White The king moves up and finds a
must be winning, e.g. 33 ... 'it>g7 shelter near the central pawns.
34 .i.xfl .t:lxfl 35 c6) 32 l:txb4 axb4 36 ...'ii'cS+ 37 'it>a4 'ifc2+ 38 <&tibS
33 1li'e7! and White's pawns will soon 'ifb2+ 39 <&tia6 'tlfe2+ 40 <&tib7 l::th7+
move forward decisively. 41 d7 1-0
31 .i.xhS l:th8?
The rook shifts out of play. It was not After all these losses I gave up
at all hopeless for Black after 31 ... l:tb4! trusting in the piece against connected
32 l:td2! (32 l:txb4 axb4 33 'ife6 "a5!) passed pawns. The following position
32 ... 'iVc8 33 'ii'xc8 ltxc8 34 c6liJc4 and occurred in my game against Radjabov
he is still alive and kicking. at Linares 2003.
167
G.Kasparov - T.Radjabov 26 .•• ~d6

Linares 2003 The young Estonian grandmaster
centralises his king.
27 l:tdl ~e6 28 f4 ~fS
Black's king is very active.
Radjabov has just sacrificed a piece

on e5 and in the game I decided not to
take it but play 22 'ife3. However
22 ~xe5!? 4Jxe5 23 dxe5 'ikc7 29 g4+!!
24 0-0-0 would have given White a People who bum the candle at both
clear advantage. I had lost confidence ends live shorter lives.
so much in Smyslov's piece against 29 ... ~xe4 30 4Jf2+! ~xf4
connected passed pawns method, that I
did not opt for this possibility which
promised a winning position. Black has
only two pawns for the piece. Later I
even blundered and lost. This painful
game prevented me from winning
Linares one more time.
Smyslov also influenced me with a
lovely checkmate of his opponent's
centralised king, just like Tal did.
Here is that sweet finish!
31 lIg1! 1-0
V.Smyslov - L.OII
Smyslov allows Lembit to decide
Rostov 1993
whether he wants to be checkmated by
the bishop or the knight, but he has no
say as to how many moves it will take
to do the deed. Smyslov was 72 when
he set up this lovely checkmate.
Vassily is the oldest living world

champion. Though I have suffered from
his influence, I wish him an even
longer and very healthy life.
168
Mikhail Botvinnik the 6 th
Mikhail Botvinnik was the first - or winning return matches! But when
world champion who did not defeat his he lost to Petrosian in 1963, FIDE
immediate predecessor, Alekhine, in a denied him the right of a re-match and
title match. Botvinnik convincingly he was finally dethroned.
won the title of world champion in a Botvinnik retired from active play
5-player match-townament in 1948, in in 1970 but continued working on
which he played all his rivals four computer chess programs, something
times. He scored 14 points out of a he had started much earlier. He also
possible 20, beating each opponent opened his own school for teaching
in their individual match. Thereafter, in juniors and I was one of his pupils. He
duels with Bronstein, Smyslov and Tal, influenced my play not only as a great
he retained his title only by drawing player but as a trainer as well.
Botvinnik liked to play on the edge A.Ilyin Zhenevsky - M.Botvinnik

of the board, especially the h-file.
Though I won games with this
method I also lost some. On the right
is a position I remember so well.
The patriarch attacked on the h-file
while his opponent played along the
g-file and in the two positions
below I tried to copy Botvinnik's
method.
P.Svidler - G.Kasparov I.Sokolov - G.Kasparov
169
Mikhail Botvinnik the 6th
A.I1yin Zhenevsky - M.Botvinnik 21 'iWe2 h5! 22 ~hl h4 23 gxh4

USSR Championship, Moscow 1927 White ·opens the g-file for himself
and the h-file for Botvinnik.
1 e4 e6 2 liJc3 d5 3 g3 dxe4 23 ... liJf4 24 'ii'd2 J:h6 25 ~e3 lbh4
White applies little pressure in the
opening.
4 ~g2 ~d7 5 liJh3 ~c6 6 0-0 liJd7
7 liJxe4 liJgf6 8 d3 ~e7 9 liJf4 0-0
26 ~xf4 l:txf4 27 ltael 'iif7 28 'ilig2
10 ~d2 e5 11 liJxf6+ liJxf6
'iih5 29 l:te3 'ue6 30 l%gl
12 ~xc6 30 .. :iVh6
White finally doubles Botvinnik's Defending the g7 pawn while
pawns on the queenside. In exchange assisting his own attack on the h-file.
Black has a small space advantage. 31 b411h4
12 ..• bxc6 13 liJg2 'iWd7 14 ltJe3 liJd5 Botvinnik neatly brings up more
15 liJc4 f6 16 ~e3 l:tae8 17 a3 a6 fire-power to the h-file.
18 ~g2 ~d6 19 f3 f5 20 ~gl 32 'ilie2 'ilif4 33 'iWg2?
White hopes to attack on the g-file,
Botvinnik repulses the move nicely. ~
20 .. JH6!
Botvinnik slowly but surely builds up 33 ... ltg6!
an attack on the kingside. Cute and effective.
170
Mikhail BOfvinnik the 6th
34 "f2 e4! 35 lLlxd6 l:txh2+! P.Svidler - G.Kasparov

36 'Wlhh2 l:th6 37 :te2 'ili'xO+ 38 l:teg2 Tilburg 1997
1 e4 c5 2 lLlo d6 3 c3
This little move is not as harmless for
Black as it looks.
3 ... lLlf6 4 .li.e2 lLlbd7 5 d3 b6 6 0-0
.li.b7 7 lLlbd2 g6 8 d4!? cxd4 9 cxd4
lLlxe4
Black can simply develop, but I
could not resist taking the central pawn.
10 lLlxe4 .li.xe4
38 .. .l:hh2+
White survives the attack but has to
settle for a lost ending.
39 ct>xh2 'ili'h5+ 40 Cii?g3 cxd6
41 dxe4 'iig4+ 42 <it>f2 'iff4+ 43 <it>e2
iixe4+ 44 <it>d2 'ikd4+ 45 ~e2
11 lLlg5 d5
White's attack is very dangerous. The
text is probably an 'only move' Black
has no time to retreat the bishop with
11...~b7? as then comes 12 .li.c4 e6
13 ~xe6! fxe6 14lLlxe6 'iib4 (14 .. .'iic8
45 ...<it>f7 15 l:te 1 q;f7 16 ilb3 d5 17 'ir'f3+ lLlf6
Black's two cOlll1ected passed pawns 18 ~g5 ~e7 19 l%.ac 1 ~c6
are too much to cope with. Winning the
game takes time but is never in doubt.
46 ng6 'ii'c3 47 <it>dl d5 48 l:t6g3
'ir'd4+ 49 ~e2 iie4+ 50 <it>d2 'iif4+
51 <it>e2 iVh6 52 l:tlg2 g6 53 a4 f4
54 nc3 g5 55 b5 -.h5+ 56 q;d2 cxb5
57 axb5 axb5 58 l:tc7+ <it>f6 59 ':c6+
~f5 60 ':c5 'ir'f7 61 l:hb5 g4 62 c4 0
63 l:tgl f2 64 nn g3 65 nxd5+ ~g4
66 ':d4+ <it>h3 0-1
20 lLlg7!! nf8 21 ':e6 wins according
I also like to attack on the h-file. to Winants.) 15 l:tel ~e7 (15 ... <it>f7
171
16 .i.g5 '1Wh5 17 'iWb3 d5 18 h3!! traps that I should try and have some fun. I
the queen.) 16 .i.g5! .i.xg5 17 g3!! know that it doesn't sound the way
'ifh6 18 ttJxg5+ <itt] 19 'iVd2! ~g8 serious and professional chessplayers
20 l:te7 .i.c6 21 ttJe6! and White wins are supposed to think during the game,
according to Peter Svidler's remarkable especially if it's a game against the
analysis. world champion, but that's exactly
After II.. ..i.d5 12 .i.f3 .i.xf3 what I thought. And, after all, it
131Vxo ttJf6 14 'iVc6+ ttJd7 151Vd5 e6 worked."
16 ttJxe6! (16 "0 ttJf6 17 'it'c6+ ttJd7 13 ....i.fS 14 g4 b6
18 1V0 ttJf6 19 'ifc6+ ttJd7 20 'it'f] My plan was based on play along the
th-Ih Degraeve-Bacrot, France 1996) h-file.
16 ... fxe6 17 ~xe6+ .i.e7 White 15 gxfS bxgS 16 fxg6
stands better. Two of three possible
continuations lead nowhere. (17 ... 'ile7
18 1Vd5 l%b8! 19 .i.g5 ttJf6 20 ~3!
'it'f7 [20 ... 'ii'g7 21 l:tfel+ .i.e7
22 l%e6±] 21 l:tfe 1+ .i.e 7 22 lie6! 0-0
23 l%xe7 and White has won a pawn.)
16. .,a6!
I was not certain whether my
opponent saw this in advance.
17 gxf7+ ~xf7 18 .i.a4l%bS?!
18 .i.g5! ttJt] (18 ... ~f8 19 .i.h6+ See diagram on page J 70.
'ite8 20 life 1 and White has very nice
compensation.) 19 'ii'e4 lic8 20 :fel Trying to force matters on the h-file.
l%c7 21 l:tac1 lid7! Now all the nonnal 18 ... 'ii'c7! 19 ':'£2 l::th4 20 .i.e3 l::tah8
moves do not succeed: 22 'iff]!! Peter's 21 l:c I and White can force a draw if
move is very strong indeed. (22 .i.f6 he wants (Alternatively 21 'iVd2 b5
Wf7!!) 22 ... h6 23 .i.xe7 l:txe7 22 .i.c2!? ttJb6 23 b3 1i'g3+ also leads
24 lhe7+ 1Vxe7 25 ':'c8+ and White to a safe position for Black and in fact
wins. after 24 lig2 lixh2 25 lhg3 l::txd2
12 .i.bS .i.g7 13 f3 26 .i.g6+ ~xg6 27 lhg5+ 'itf7
Here I quote Peter's words from his 28 l:xg7+ ~xg7 29 .i.xd2 he has a very
Chessbase analysis: "Then I realized slight edge.) 21...lixh2 22 l:txc7 l:th I +
that 13 0 leads to some very interesting 23 ~g2 118h2+ 24 ~g3 ':'h3+ 25 ~g4
and promising positions and decided l:th4+ and White must settle for a
172
Mikhail Botvinnik the 61h
perpetual as pressing forward would 24 ... ttJh5 25 iLf5 ttJf4 26 iLxh3

land him in checkmate. ttJxh3+ 27 ~hl
19 .i.e3 ttJf6 20 -.d2 -.d6 21 ':12 Svidler, just like Ilin-Zhenevsky,
goes to hI with his king, so I had reason
for hope. In reality, Black simply has
insufficient compensation for the pawn.
27 ...1*'f6 2S .tlg3! _f5 29 i..xg5
ttJxg5
After 29 .. Jlxd4 30 -'g2 ttJxg5
31 l:xg5 ~7 (31.. ...f6 32 .tlgl) is
given by Winants, then 32 .tlgl i..f6
33 .tlh5! wins.
301bg5
21...11ahS
I directed all my heavy pieces against
the h2 pawn, actually more so than
Botvinnik, therefore I was optimistic,
22 l:1g2!
I was not worried about my
opponent's play on the g-file,
Botvinnik's opponent also had the
g-file but got nowhere with it.
22 .. .llh3?! 30 ... 'ii'h3
I kept attacking when I should have Black might have hoped for serious
been defending. Botvinnik won so I felt counterplay, but not in this case
obliged to play for a win as well. because of the weakness of his own
22 ... ttJh7 or 22 ... i..f6 should have been king.
tried. 31 l:lg2 i..f6
23110 l:lSh4? If 31.. ..i.xd4 32 "d3.
32 -.d3 .tlxd4 33 'ii'g6+ ~e6 34 'ii'eS
.tlc4
After 34 ... ~d6 comes 35 ~8+ ~c6
36 -'a8+ ~b5 37 a4+!
24.i.c2!
Here I had to realise that my rook on
h3 was trapped.
1"73
3S "d8! This is a sideline.

Black's king will stay where it is. 13 ... exd4
3S ..:.fS 36l:el+ ~eS 37 'irb8 After 13 ... c4 14 -ltxc4 exd4 15 cxd4
Black loses more material, therefore I ltJa5 16 -ltd3 'ifxc2 17 -ltxe2 tOxe4
resigned. 18 l:te1 ltJd6 19 -ltf4ltJac4 20 -ltb3 .tf5
1-0 21 ltJe5 -lte6 22 ltJd3 l:tad8 23 tOe5
-ltc8 (23 ... -ltd5?? 24 -ltxd6 wins)
This game was not the only time I 24 -ltxc4 tOxe4 25 -lte7 Rogers went on
to beat Solomon in Sydney 1999.
attacked along the h-file while my
14 cxd4 -ltg4 IS eS
opponent was looking for a result along
the g-file.
I.Sokolov - G.Kasparov
Hoogovens, Wijk aan Zee 1999
1 d4 ltJf6 2 c4 e6 3 ltJc3 -ltb4

I have never perfonned really well in
the Nimzo - my Psakhis game comes to
mind here - but before this game in the
tournament I had just won seven games
in a row. My opponents were strong 15 ...-ltxf3 16 exf6 ltJxd4 17 -ltxh7+
grandmasters, but only two of them Wh8 18 fxg7+ ~xg7 19 .tb2 %tad8
20 gxf3 I:th8
made it into the top 10. Among these
seven Topalov was the strongest. See diagram on page J 70.
4 e3 Of course I could select this game
I expected 4 "c2 as that is Ivan's under the motif of the doubled f-pawns,
main weapon against the Nimzo. I can't but I was hoping to attack on the h-file
be certain why he changed to the 4 e3 and I was not worried about the g-file
line. Maybe he did it because this was as Botvinnik dealt with this problem
Botvinnik's main line. well. There was no reason to think that
4 ... 0-0 S -ltd3 dS 6 ltJf3 cS 7 0-0 ltJc6 I was not going to handle things equally
8 a3 -ltxc3 9 bxc3 ~c7 10 ~c2 dxc4 as well.
11 -ltxc4 eS 12 ~d3 l:te8 13 e4 21 <it>hl
174
Mikhail Botvinnik the 61h
Just like Ilin-Zhenevsky. damage. Thus 24 ... 'ifh5 25 :lg2 f6

21...l:b:h7 (25 ... lte8 26 'ife4 f6 27 ~c3!) 26 'ifc4
Hiding the king with 21...~fB (26 'it'g6 'ifxg6 27 l1xg6 and White is
and, if necessary, walking over to the somewhat better in this endgame.)
queens ide, occurred to me. However 26 ..... f7 27 'ifxc5! (If 27 '6'd3 l:lh6
Botvinnik kept his king on g8, so gave 28 'ii'd2 'iih7 29 'it'f4 b6?? [29 ... lte8 or
me no hint what to do. He should have 29 ... ltfB should have been tried.]
told me this in his school. 30 lth3 and White was winning in the
22 ltgl+ ~h8 game Joshi-Shankar, Mumbai 1999.
I decided to step over to the edge. My But Sokolov is thinking of an exchange
brain was preoccupied with aggressive sacrifice.) 27 ... l£Je2 28 'ii'g5 l1dl+
ideas along the h-file, not my (28 ... l£Jxg3+ 29 fxg3 l:td6 30 l1c2 l1h5
opponent's play on the g-file.
is equally unclear.) 29 l::tgl l£Jxg3+
2311g3 'ilt'e5
30 fxg3 l:td6 White certainly has
On e5, the queen stops a threat. Not
compensation for the exchange and
23 ... b6? when we see White's threat:
Black must be careful. Nevertheless, to
24 ~xd4+ l::txd4 25 'ii'f5 and wins.
be objective, Black may well survive.
2411agl
25 'iicl?
Ivan wants to improve his queen, and
in the end he does, but he needs some
input from me. My opponent missed an
almost winning continuation in 25 f4!
when White opens the second rank
for the queen to get to g2! 25 ... 'irb5
(25 .....d6 26 'ilt'f5) 26 fJ! 'itb6
(26 ... l1h3 27 l1g4) 27 'ii'g2! 'it'fB
28 ltg7 White overwhelms Black on
the g-file. Then 28 ... l1h6 29 'ii'g5 wins.
24 ...l:lh4
Ftacnik writes: "The first
independent move is very unfortunate,
it seems that Black's position instantly
becomes hopeless." The Slovak
grandmaster did not know that it was
not fully independent. I just copied
Botvinnik who had put the rook in front
and the queen behind. When one
attacks a pawn the rook goes in front,
however when the king is the target, 25 ... ~b7??
then often the queen goes in front as an Botvinnik gave no instructions on his
invasion is feasible and it wreaks more game. I lost track of what to do with
175
my king. Anyway I wanted to attack 27 .. :iie6 28 1Vg2

Sokolov's king, not start defending my
own. Better was 25 ... 'ifh5! 26 l:tlg2
(26 l:l3g2 f6 (26 ... b6 27 'ii'e3 ~d5
28 "e7 wins.) 27 'ii'e3 l:te8 28 ~xd4
cxd4 29 'iVb3 fib7 and though Black is
living very dangerously, according to
Tsetsarsky, he will get away with it.)
26 ... l:le8 27 'ilfgl "e5 28 l:g8+ l:hg8
29 l:txg8+ ~h7 30 l:lf8 'ii'f4 (30 ... 'i!i'g7
31 'ifbl+! ~h6 32 ~c1+ ~h5 33 ~e3
and Ftacnik calls it a win at the end of 1-0
his line.) 31 'ifg8+ <ith6 32 'ifg2 ~h7 Both l:tg8+ and l::th3 !, threatening
33 1Vg3 White is better.
1Vg7 mate, are menaced. I could have
26 'i!Vbl+!
postponed the checkmate for another
A subtle check forces the king to h8.
six moves, but there was no point in
26 ... Wb8
doing so. It is remarkable that there are
After 26 ... f5 27 ~xd4 cxd4
28 'ii'xb7+ ~h8 29 'ii'f7! Black gets nine Kasparov versus Sokolov games
checkmated. in the database and only one draw,
271Vn when I was White against Andrei
Now Black can't ease the g-file Sokolov in the USSR Championship
pressure with ... l::tg8 as White would 1988 - the other 8 times the result was
take the rook with check. On other 1-0. In addition I lost to Andrei at the
moves Black's king will be caught on Reykjavik 1988 World Cup where I
the g-file. missed a battery and dropped a piece.
Botvinnik affected my play in many ways. I also picked up his idea in the
English Opening of allowing the opponent to push a black pawn to the
e3-square and letting him keep it there.
M.Botvinnik - V.Smyslov G.Kasparov - A.Karpov
176
M.Botvinnik - V.Smyslov 11 ....i.b7 12 1Vel tLlbd7

USSR Spartakiad, 1964
Ulllil my matches with Karpov these

two gfc!at players held the record of
playing the most games at the very top
of world chess. This game was their
first since their final match in 1958. In
the sixties they played seven games.
Botvinnik won two and the rest were
drawn. My score after my last match
with Karpov produced a somewhat
similar picture - relatively few games
13 g4
and one champion who won no more
Pushing the g-pawn was one of
games. I won 5 of the 12 regular games
Botvinnik's specialities.
we played together.
13 ... h6 14 h4
1 e4 lDf6 2 lDc3 e5 3 g3 ~b4 4 .i.g2
A pawn move that I also employed
0-0 5 a3 .i.Ie3 6 bIe3 e4
regularly.
This is quite ambitious.
14 ... lDlS 15 1Vg3 lDg6 16 lDh3 lDh7
7 lDh3 l:te8 8 0-0 d6 9 lDf4 b6
17 h5lDh4
10 f3 e3
It is quite unusual to put a knight on
See diagram on page 176. the edge like this, but Smyslov soon
makes sure he can rescue it.
Smyslov sacrifices the e-pawn In 18 ~hl f5 19 .i.b2 Vi'f6
return for the doubling of White's
pawns.
11 d3
20 f4
White can exploit the fact that the
Botvinnik doesn't take it - if he had queen is on f6.
done so, then the game would become 20 ... ~xhl 21 g5 bIgS 22 fIgS ~e5
unclear. Now the e3-pawn cuts White's 23 1iIh4 ~e6 24 .tlf4 g6
camp into two but at the same time it Opening the kingside helps White.
can itself become a target. 25 bIg6 lDlS
177
Mikhail Botvinllik the 6th
26 'it'h6 'it'g7 27 lbf5 tLlxg6 28 lIDO 43 e4 c5?

l:UB 29 ':f6 Black is lost anyway.
White has not only exerted pressure 44 ~f6 1-0
Black resigned as White will deliver
against Black's king but has cleared the
checkmate.
way to the e3-pawn.
29 ... ~xh6 30 gxh6 ':xf6 31 lbf6
G.Kasparov - A.Karpov
~h7 Game 2, World Championship,
Seville 1987
1 c4 e5 2 tOc3 ltJf6 3 ltJf3 ltJc6 4 g3

~b4 5 ~g2 0-0 6 0-0 e4 7 ltJg5 ~xc3
8 bxc3 l:te8 9 f3 e3!?
A novelty in this particular position.

It was Igor Zaitsev's idea.
10 d3!
32 ~c1 I could have taken but I knew
Botvinnik captures the e3-pawn. Botvinnik's game, he had beaten such a
32 ... .:g8 33 tLlg5+ <it>xh6 34 ~xe3 great player as Smyslov. Why not just
~h5 35:n follow him?
White is now winning easily. 10... d5
35 ... l:e8 36 :h7+ <it>g4 37 ~f2 tLle7 Karpov plays differently.
38 tLle6 tLlf5 39 tLld4 tLlxd4?! 11 ~3
Smyslov makes sure he loses. After I changed sides because Karpov was
39 ... tOxe3 40 tOxc6 tOd 1+ 41 ~e 1 playing on a different flank. I was able
tOxc3 42 ~d.2 tOxe2 43 ':xc7 Black to adjust.
struggles with his cut-off king. 11... ltJa5 12 'ira3 c6 13 cxd5 cxd5
40 cxd4 ':c8 41 d5 ~a4 42 ~d4 a6 14 f4 ltJc6 15 ':bl 'iic7 16 ~b2 ~g4
178
Mikhail Botvinnik the 6'h
17 c4 25 ... lLlxg3+! 26 ~gl! (26 hxg3?

After 17 lLlf3 'ifd7 18 ~al l:r.e7!? 'ifxg3 wins) 26 ... lLle2+ (26 ... lLlxfl
19 l:r.fc 1 ~h3 20 c4 White is a bit better. 27 l1xc7 e2 28 'ir'xa7 e 1='if 29 l:Ixf7+
17 l:r.fe I! and White has an edge <it>h6 30 lixf6+! ~g5 31 'ir'g7+ ~f4
according to Karpov and 1. Zaitsev. 32 'iVh6 mate) 27 ~hl lLlg3+.
20 ...l:r.ad8 2111b3 lLld4
17... dxc4 18 ~xf6
Doubling the f-pawn, as we know, is
a tricky matter.
18... gxf6 19 lLle4 ~g7
After 19 .. .lhe4!? 20 iLxe4 f5!
21 .i.f3 lLld4 22 dxc4 .i.xf3 23 exf3 e2
24 life 1 ~xc4 Black is safe.
20 dxc4
This is an inaccuracy. I thought I was 22 l:[xe3
getting closer to the e3-pawn - just like Finally I took the pawn, but there
Botvinnik. are too many pieces left on the
a) If 20 'ir'c3 'ir'd8! (not 20 ......e7 board, many more than in Mikhail
21 l1xb7). Moiseevich's game.
22 ...'ir'xc4
b) 20 h3 .i.xe2 21 lLlxf6 ~xfl
After 22 ... lLlc2 23 'ir'c3 ttJxe3
22 'ir'c3 lie5! and with this beautiful
24 'ii'xf6+ ~f8 25 'ifh6+ rj;e7
move Black takes charge.
they miss the precise check 26 'ir'g5+!
c) 20 lLld6 Karpov and Zaitsev show which covers the c5-square. (26 ~f6+
a very nice way to a perpetual, and I ~d7) 26 ... ~d7 27 lLlc5+ and White is
should have gone for it. 20 ... l1e6 in the game.
21 lLlxc4 l1d8 22 f5 lIee8 23 l1b2 lLld4 23 ~hl lLlf5 24 lId3 .i.xe2 25 l:bd8
2411xb7 lLlxe2+ 25 ~hl lbd8 26 ltel:e8
179
To make things even more annoying,

with Black I pushed my e-pawn all the
way to e3 in a 1982 game against
Romanishin. Do you know what
happened? I lost the following position
as well!
USSR 1982
27"a5
After 27 liJd6 liJxd6 28 'ifxd6 ..to!!
and the battery exploits the weakness of
the back rank.
"d3
27 ... b5 28 liJd2 29 liJb3
32 ... e3 33 f4 gxf4 34 lbf4 l:[xf4

35 gxf4 'ii'f5 36 lbd5 'ifxd5 37 "c7+
~g6 38 'ifc2+ ~f6 39 a6 "a8 40 'ifc4
"e4 41 'irc5 'ifbl+ 42 ~g2 "dl
29 .....tf3!! 30 ~xf3 'ir'xf3+ 31 ~gl 43 'ifg5+ cj;r7 44 'ifh5+ ~g7 45 a7
lbel+ 32 "xelliJe3 0-1 1-0
M.Botvinnik - A.Pomar
One of my specialties was to push

my g- and especially my h-pawn. I
won games with both, but sometimes
the idea backfired. First I show you
a diagram of the inspirational
Botvinnik game and then, on the next
page, my games.
180
Mikhail Botvinnik the 6lh
N.Short - G.Kasparov V. Anand - G.Kasparov
M.Botvinnik - A.Pomar
IBM, Amsterdam 1966
1 c4 c6 2 tiJc3 d5 3 cxd5
Though I played a few Exchange
Slavs and French defences, these
variations do not suit my style. By the
way I beat Dolmatov in an Exchange
Slav. I didn't select my loss against him
for this book but he did beat me in a
Youth tournament in the USSR in 1977.
I set up a battery but it very quickly
12 ... gS! 13 ~g3 h5 14 h3 g4 IS hxg4
lost. The opening of the Botvinnik
hxg4 16 lLle5 lLlxe5 17 ~xe5 f6
game did not catch my imagination but
18 ~g3 ~f7 19 l:.e 1 lth5 20 'ii'dl ~e4
the game did.
21 ~f1 .i.f3 0-1.
3 ... cxd5 4 d4 tiJf6 5 4)0 tiJc6 6 .i.f4
14 ... .i.g6 15 c5 lLle4 16 f3 lLld2
.i.f5 7 e3 e6 8 ~b5 ~b4 9 tiJe5 'ir'a5
17 l:[f2 lLlc4
10 ~xc6+ bxc6 11 0-0 .i.xc3 12 bxc3
Black saves the bishop, but it will
l:tc8 13 c4 0-0 14 g4
remain rather passive.
See diagram on page 180. 18 tiJxc4 dxc4
Going after the bishop also occurs in
this line, sometimes Black can even
do this to White, one example being
Seirawan-Beliavsky, Brussels 1988.
That game went like this: 1 d4 d5
2 c4 c6 3 tiJc3 tiJf6 4 cxd5 cxd5 5 .i.f4
ttJc6 6 e3 .i.f5 7 lLlfJ e6 8 ~b5 lLld7
9 0-0 .i.e7 10 .i.xc6 bxc6 11 l1c 1 l:tc8
12 lLla4
181
19 ~d6 34 ... e5 35 i.xe5 nb7 36 'ii'f4 as

The bishop targets nothing but still it 37 l:tfl ~b3 38 d5 cxd5 39 c6
is very useful as it keeps both black Botvinnik chooses to win with the
rooks very passive. c-pawn. He could have triggered an
19 .. .1He8 20 e4 f5 21 ~c2 fxe4 execution on the long diagonal as well.
22 fxe4 'ita3 23 ':el 'ii'h3 24 ng2 39 .. .11&7 40 c7 'ii'e7
':cd8 25 ':g3 'ii'h6 26 'ii'xc4 'itd2
27 'ii'c3 'ji'xa2
Black wins back the pawn, but
material often doesn't count in opposite
coloured bishops middlegames.
White's pieces are pretty active.
28 ':g2 'ifa6
41 ~d6 1-0
N.Short - G.Kasparov
Game 16, PCA-World Championship,
London 1993
29 h4 1 e4 c5 2 tOo d6 3 d4 cxd4 4 tOxd4

Pushing the h-pawn is a nice plan and tOf6 5 tOc3 a6 6 ~c4 e6 7 ~b3 b5
further restricts the bishop. 8 0-0 ~e7
29 .. J:td7 30 h5 ~f7 31 %:tal 'ifc8 I tried several set-ups against the
32 'ito 'iWd8 33 g5 g6 Sozin, but this was my final choice in
the match.
9 'if0 ~c7 10 jig3 tOc6 11 tOxc6
'itxc6 12 nel ~b7 13 a3
34 h6
Botvinnik's h- and g-pawns are
suffocating the Spanish grandmaster. 13 ... l:td8! 14 0
182
The standard sacrifice 14 lOd5? Black has fully equalised.

need not worry Black, e.g. 14 ... exd5 22 'ii'e3 ~g8 23 ~gl ~f8 24 'iifl
15 exd5 lOxd5 16 'iVxg7 ~d7 17 'ikxf7? .i..a8 25 lOe2 g6?
(17 'ii'g4+ Q;c7 18 ~xd5 'ii'xd5 I should have exchanged queens and
19 %lxe7+ ~b8 Black is very active.) settled for an equal endgame. However
17 ... l%de8 18 ~g5 :'hg8! 19 .i..xe7 I had won the previous game quite
convincingly and had not yet lost a
single game in the match. All of which
made me fall asleep.
26 lOd4 'tlfe5 27 l:tel g5
I advanced my g-pawn further - the

same way Botvinnik did so many
times.
28 c3 ~g7
19 ....:.xg2+!! Black finds an effective
way to destroy White's king. 20 ~xg2
ttJe3+ 21 ~g3 'ifg2+ 22 ~f4 iie4+
23 ~g5 (23 ~g3 'ikg4 mate) 23 ... h6+!
24 ~xh6 (24 ~h5 'ii'g4+ 25 ~xh6
:th8+ 26 'ifh7 lOf5 mate; 24 ~f6
'il'e5+ 25 ~g6 .i..e4+) 24 ... l:th8+
25 ~g7 'irb7+ 26 <i!7f6 'iib4+ 27 ~g7
ifh.6 mate.
14 ... 0-0 15 ~h6 lOe8 16 ~hl <it>h8
17 .i..g5 ~xg5 18 'iVxg5 lOf6 19 l:tadl 29 ~c2!
l:td7 20 l:td3 ':'fd8 21l%edl Now Nigel starts softening up my
queenside with a strong regrouping of
his pieces.
29 ...l:tg8 30 lOb3 ~f8 31 l:[d4 ~e7
My king presented problems on the
e7-square not only in this game, but
also when I was Black against Kramnik
in the Korchnoi-tribute tournament at
Zurich 200 I.
32 a4! h5 33 axb5 axb5 34 ltb4 h4
21. .••c5 351Od4
183
10 ... bS
Of course I advance my h-pawn.
11 ~xg4 ~xg4
I later played the simpler 11 ... hxg4.
12 0 ~d7 13 ~f2 ttJc6 14 'iWd2liJeS
3S ... g4
It was too late to back down
from Botvinnik's pawn onslaught. If
35 ... l:tb8 36 ~d3.
36lhbS
The unusual queen exchange with
36 ... 'ifh2 offered no hope either. 15 b3!
36 ... dS 37 'ii'xh4 'ii'hS If 15 0-0 then 15 ... g4! After all,
After 37 ... gxf3 38 liJxf3 'it'h5 39 'iVa sometimes Mikhail Moiseevich's ideas
White wins. really work. Two years earlier we had a
38 liJrs+! 1-0 play-off in the PCA Geneva rapid
tournament. In the blitz, improving on
V.Anand - G.Kasparov our rapid game from the same event, I
Frankfurt Giants 1998 pushed my g-pawn. I got a fabulous
1 e4 cS 2 liJo d6 3 d4 cxd4 4 liJxd4 game, yet I spoiled it. 16 f4 liJc4
liJr6 5 liJc3 a6 6 ~e3 liJg4 7 ~gS h6 17 'ife2 :c8! 18 b3 liJa3 19 liJd5 e6
8 ~b4 gS 20 liJb4 'ifa5 21 'ii'el
This is not the Botvinnik-effect as the
variation goes lik~ this, but maybe
indirectly there is an effect as I was
entering a g- and h- pawn pushing line.
9 ~g3 ~g7 10 ~e2
21.. .h4! I keep following Botvinnik's

play. 22 .i.e3 h3 23 g3 liJb5 24 l:ld 1
ttJc3 25 ttJd3 'ikc7 26 :c1 liJxe4 White
has very little for the pawn, yet Black
has to play carefully. 27 f5 e5 28 f6
liJxf6 29 liJf5 ~xf5 30 l:txf5 'ifc6
184
31 '6'e2 "'e4 32 1%f2 lDd5 33 1%e 1 Black's problem is that he can do

'iixe3?? A dreadful mistake. 33 ... 0-0 nothing as any move would expose his
wins simply. 34 ii'xg4! Suddenly White own king. But in the long run White
is winning. 34 ... 0-0 35 l:lxe3 lDxe3 will still open up Black's king.
36 ii'xh3 and though I played on I 27 ...... c8 28 ~b2 'iWb7 29 lDdc2 f6
no longer was in a position to save the 30 lDd4
game.
15... e6 16 lDde2! l%c8 17 ~d4! b5
18lDdl! l%g8 19lDe3! as
I decided to do what Botvinnik did
on the other side of the board as well.
30 ... h4
I have tightened my grip on
absolutely nothing! That's because
Anand has no pieces on the kingside -
200-0-0! and especially not his king.
Vishy has handled the opening in 31 g3 g4
great style, but that offers me little The same push but with a different
consolation. Maybe Botvinnik had effect.
mentioned that the opponent could 32 f4 lDo 33 lDIO gIO 34 f5 hIg3
castle on the other side, but he certainly 35 hxg3 1%xg3 36 'ii'h2 l:lg5 37 ~c1
did not emphasise it sufficiently. 1%g7 38 'ifh5+ l:lf1
20 ... a4 21 ~bl a:xb3 22 c:xb3 1%a8
23 ~c3 1%a6 24 lDc2 ~f8 25 lDb4 1%a8
26lDd4 ~e7
39 "':d3 1-0
I resigned as in a moment or two my
27 l::thel king will be caught.
1-85
Botvinnik's influence came from so many different directions. It is almost

impossible to summarise. I'll just show you one more example where,
without any fear of losing, I went for a queenless isolated pawn endgame.
G. Veresov - M.Botvinnik A.Karpov - G.Kasparov
G.Veresov - M.Botvinnik It is an interesting idea to exchange

USSR Championship Semi-Final when you have an isolani, but Black's
Leningrad 1938 quick play justifies it.
17 lbe3 lbe3 18 .xe3 'ifb6
1 d4 lbf6 2 e4 e6 3 lbe3 .i.b4 4 'ii'e2 19 'ifd41fxd4 20 lbxd4 l%e8
d5 5 exd5 exd5 6 .i.g5 .i.e6 7 lbfJ
lbbd7 8 a3 .i.e7 9 e3 h6 10 .i.b4 0-0 See diagram above.
11 .i.e2 e5 12 0-0 :te8 13 :tfel a6
14 dxe5 Black has solved his problems.
21 fJ <ifr'fS 22 ~n .i.d7 23 .i.d3
14 ... lbxe5
Botvinnik had quite a number of nice 23 ... g6
wins in isolated pawn middlegame I also played this it covers the
positions. f5-square.
15 b4 lbee4 16 'it'd3 lbxe3 24 ~e2 l%e3 25 ~d2 l:I.e8 26 ~e2
186
~a4 27 g4 lLle8 28 ~xe7+ ~xe7 29 h4

lLld6 30 ~d2
48 ... d4?!
Botvinnik pushes his pawns,
although there were other candidate
30 ... ~b5 moves. Possible was 48 ... g5!? or
Karpov also had a similar example, 48 ... gxh5 49 l:thl!? You will see that I
he beat Karasev brilliantly by have to face a problem like this when
exchanging pieces to obtain control
my opponent did not automatically
over the c4-square in a queenless
recapture after I took his h-pawn.
isolated pawn endgame.
Botvinnik did not mention this
31 .i.xb5 axb5 32 'it>d3 ~d7 33 lLlb3
lLlc4 34 lLld4 l1e8 35 lLlc2 possibility in his school.
After 35 lLlxb5 l:1xe3+ 36 ~d4 ltxf3 49 hxg6
37 ~xd5 lLlb6+ 38 ~e4 l:h3 39 l:cl Best was 49 g5! Interestingly,
White is not worse. Botvinnik missed a similar pawn
35 ... tOe5+ 36 ~e2 nc8 37 tOd4 breakthrough in his book on Karpov. I
discovered it and published it first in
the Predecessors book. 49 ... hxg5
(49 ... tOe5 50 gxh6l:1xf3+ 51 ~g2 :f5
52 ':hl the h- pawn is dangerous.)
50 h6 Ad2+ 51 ~gl lLle5 52 l1fl Ac2
53 tOxd4 :c8 54 'it>g2 White has an
edge.
49 ... fxg6 50 a4
After 50 l:e I! tOxa3 51 lLlxa3 l1xa3
52l:1e4.
37 ...l:1c3 50 ...l:ld2+ 51 ~g3 d3
Botvinnik's rook became annoying. The d-pawn becomes strong. On the
38 l:1a2?! tOc4 39 tOxb5 l:1xe3+ other hand I lost to Karpov when I had
40 ~f2 lId3 41 ~e2 lIb3 42 l:1c2 b6 a d6-pawn with the white pieces. That
43 lIa2 l:le3+ 44 'it>f2 lId3 45 ~e2 l:b3 was the sixth game of our first 1984
46 l:1a 1 ne3+ 4 7 ~f2 l:ld3 48 h5 match.
187
Moscow 1984
1 d4 d5 2 ('4 e6 3 lOn e5 4 cxd5

exd5
In the Predecessors book I looked at
chess culture in general. In this work I
do not go into details of the giants other
than the champions. Nevertheless do
In this game, however, the d-pawn keep in mind that this defence is named
becomes so powerful that it decides the after Siegbert Tarrasch.
outcome, whereas in the earlier game I 5 g3 lOf6 6 ~g2 ~e7 7 0-0 0-0
8 lOe3 lOc6 9 ~g5 cxd4 10 lOxd4 h6
just dropped it.
11 ~e3 l:le8 12 'ilfb3lOa5 13 'iVc2 ~g4
52 a5 bxa5 53 bxa5 lIb2 54 lOe3
14lOf5 l:lc8
ne2 55 ltJd5??
In my game the same move would
have been a mistake by Karpov. After
55 ltJe4! d2 56 ltJxd.2 lIxd2 57 a6 ltJb6
58 a7 ltJa8 59 nh 1 White holds.
55 ... d2 56 l:tdl l:[c1 57 ltJe3 ~c6
58 fitf4 ~b5 59 ~e4
15 ~d4 ~c5 16 ~xc5 lbc5 17 ltJe3

~e6 18 :adl 'S'e8 19 W'a4 %:td8
20 l:[d3 a6 21 ]:lfdl lOc4 22 lOxc4
lIxc4 23 "a5 l:tc5 24 'ifb6 lid7
25 l:td4 "c7 26 "xc7 l:tdxc7
Just like Karpov I strengthened the

59 ... h5 60 gxh5 gxh5 61 ~d3 h4
c-file, toying with idea of invading on
62 ~e2 h3 63 ltJg4 nxdl 0-1 the c-file.
27 h3 h5
I trusted Botvinnik so much that I Botvinnik paid a lot of attention to
even employed one of his ideas in my gaining space. For this reason, it is a
very first world title match against question whether I should put my pawn
Karpov in 1984. on the same colour square as Karpov's
188
bishop. If it is a mistake I will have to I should have kept the rooks on, but
work out who encouraged me to do his rook was a tower of strength on d4.
this. I hoped my d-pawn would work like
28 a3 Botvinnik's.
36 exd4 ~e7 37 ttJa2 .ltc8
Geller suggested 37 ... ttJe4 which is a
Karpov-style move.
38 ttJb4 ~d6 39 f3 ttJg8 40 h4 ttJh6
41 ~f2 ttJf5 42 ttJc2 f6 43 .ltd3 g5
44 .ltxf5 .ltxf5
Karpov beat Van der Wiel in

Amsterdam 1980 when they had a very
similar pawn structure endgame (the
28 ... g6 pawn islands were the same). Karpov
Another pawn goes to the white had a bishop versus a knight, but a pair
squares, but Botvinnik played this of rooks remained on the board. I
move as well. thought, okay, I will not win but there
29 e3 ~g7 30 ~h2 l::tc4 31 .i.f3 b5 was no way I could imagine losing it.
32 ~g2 l:t7c5 33 l:bc4 J. van der Wiel - A.Karpov
IBM, Amsterdam 1980
33 ... l::txc4
I had a chance to recapture with
But back to my game ...
either pawn. 33 ... dxc4 maintains the
balance by dynamic means, which 45 ttJe3 .ltbl 46 b4
comes to me more naturally, but Daniel King mentions in his analysis
Botvinnik's game was in my mind and that the position was reminiscent of the
I wanted to follow it. After 33 ... dxc4 game Saidy-Fischer, New York 1964.
34 ~d6 a5 35 l:tb6 (35 ttJa2 ttJd5) This was the U.S. Championship where
35 ... ttJd7 36 l::txb5 l::txb5 37 ttJxb5 ttJc5 Fischer made 100 percent. Maybe he
Black retains a material balance and spent less time investigating the games
has a secure position. of the world champions.
34 ~d4 ~f8 35 .lte2 l:txd4 46 ... gxh4?
189
I missed Karpov's reply, but at least I Taking the pawn is a mistake, as

contributed to the development of Black's king can cause problems by
endgame theory. Daniel King analyses approaching the queenside. On the
in depth the position after 46 ... ..tg6. other hand 55 liJh5!! wins. 55 ... ~xf3
His conclusion is that Black can hold.
56 liJxf6. From this point g4-e5-c6-a5-
b3-c5 attacks the a6-pawn; then with a
king move White will bring the
opponent into zugzwang. When Black
drops the a6-pawn, then the direct
manoeuvre itJb8-c6-itJa7 forces him to
defend the b5-pawn. After any casual
king move the knight returns to e3 via
c6-e7-f5. If White accomplishes these
things, he just has to march to a7 to win
47 liJg2!
the game.
This is against all principles, I did not
55 ... <iite6
think that when I have the bishop, and
he a knight, that he would open the Better according to Geller was
position. You can see how fair I am, 55 ... ~d6. He is right.
here I could blame my seconds for 56 itJc7+ ~d7 57 liJxa6 ~xf3
missing this move. Such a move 58 ~xf6 ~d6 59 ~f5 ~d5 60 ~f4
becomes part of our chess heritage, but ~hl 61 <it'e3
why does it happen against me? It
would take pages to show you all
finesses of this endgame. I'll just show
you some interesting points.
47 ... hxg3+ 48 ~xg3 ~e6 49 liJf4+
~f5 50 liJxh5 ~e6 51 liJf4+ ~d6
52 ~g4 ~c2 53 <ith5 ..tdl 54 ~g6
~e7
61 ... ~c4
Finally I do something active on the
c-file, just like Botvinnik did.
62 itJc5 ~c6 63 ttJd3 ~g2?
Averbakh and Taimanov suggested
63 ... ..te8! keeping the bishop on this
diagonal.
55 liJxd5+? 64 ltJe5+ ~c3 65 ltJg6 ~c4 66 ltJe7
190
66 ....i.b7 69 tLlxbS <it'a4 70 tLld6 1-0

The amazing move 66 ... .i.h 1!! still Botvinnik also set an example that
holds. I thought of this but could not sadly followed. He trained me and
remember any champion playing passed on his knowledge. I also gave
anything like it before. 67 tLlf5 ~d5 lectures to young juniors, including
This is the difference compared with Shirov and Kramnik. But Botvinnik
the game: here the .i.hl stops ~d3. undertook coaching only after he had
After 68 ~f4 .i.e4 69 tLld6 .i.c2 retired, whereas I made the mistake of
70 tLlxb5 <it'c4 Black draws. doing the same thing while I was still
67 tLlfS ~g2? an active player. So I helped to improve
That damned g2-square in this game! the play of the champions who
Key moves keep taking place there. dethroned me. What a mistake that
67 ... ~d5 was the best move, but the was! Had he worked on his own maybe
legacy of Botvinnik's game is to remain I would have stayed champion for
active on the c-file, therefore I did not longer. However, when all is said and
return with my king. done, overall I am of course very
68 tLld6+ <it'b3 appreciative of our' great first Soviet
world champion.
191
Max Euwe the 5 th
Max Euwe
Euwe was the last champion still
convincingly th
the 5regained his title with a
alive when I was born. He andfinal score of 15 'h-9',h.
Petrosian were able to form an opinion In 1946 Euwe still played very well,
on my playas they both died in the but the 1948 World Championship final
Euwe was the last champion still convincingly regained his title with a
1980s when I was already a decent showed that he had lost touch with the
alive when I was born. He and final score of 15 'h-9',h.
player. Not everyone knows that Euwe very best players of the world.
Petrosian were able to form an opinion In 1946 Euwe still played very well,
won the world title back in 1928. Nevertheless he continued to write
on my playas they both died in the but the 1948 World Championship final
But that was the World Amateur many fine articles and books. Though
1980s when I was already a decent showed that he had lost touch with the
Championship. Later, in 1935, he he was an amateur world champion
player. Not everyone knows that Euwe very best players of the world.
defeated Alekhine by the narrowest of I always considered him to be a
won the world title back in 1928. Nevertheless he continued to write
margins:
But that was 151J1-14'h. Though
the World their
Amateur
true world champion too and began
many fine articles and books. Though
rematch began with Euwe dominating,
Championship. Later, in 1935, he studying
he was hisan games
amateurwhen I waschampion
world young -
in the sixth
defeated game Alekhine
Alekhine commenced
by the narrowest of and not only for my Great Predecessors
I always considered him to be a
amargins: streak of three
winning 151J1-14'h. games their
Though and series.
true world champion too and began
rematch began with Euwe dominating, studying his games when I was young -
in the sixth game Alekhine commenced and not only for my Great Predecessors
I followed Euwe's play where he I show the position of my last rapid
a winning streak of three games and series.
cut the position into two with a game where I should have drawn
d5-pawn and paralysed the b7-bishop. - and thereby won the match - by
Then he gave up
I followed the strong
Euwe's play d5 pawn.
where he I show Euwe's
adopting the position
idea. of my last rapid
cut the position into two with a game where I should have drawn
M.Euwe
d5-pawn - A.O'Kelly
and de b7-bishop.
paralysed the Galway This game was extremely important
- and thereby won the match - by
Then he gave up the strong d5 pawn. to me because I wanted to prove my
adopting Euwe's idea.
superiority over Kramnik after losing
M.Euwe - A.O'Kelly de Galway theThis
titlegame
- andwassoextremely
there wasimportant
a lot of
to meatbecause
pride stake. I wanted to prove my
superiority over Kramnik after losing
G.Kasparo\7 - V.Kramnik
the title - and so there was a lot of
pride at stake.
G.Kasparo\7 - V.Kramnik
I played the same idea in a very

important last round rapid stage of a
match against Kramnik. We drew 4
I played
regular the Isame
games, won idea in a rapid
the first very
important last round rapid
and then we drew 4 games. stage of a
match against Kramnik. We drew 4
regular games, I won the first rapid192
and then we drew 4 games.
192
Max Euwe the 5th
M.Euwe - A.O'Kelly de Galway

Groningen 1946
1 d4 ttJf6 2 c4 e6 3 ttJc3 .i.b4 4 e3 b6

5 ttJge2 ~b7 6 a3 ~e7
22 ttJa4!
In the next few moves the Dutch
world champion increases the pressure
in a very subtle way.
22 ... ~8 23 'irc3 1IIa7 24 :a2 .i.e6
25 l%d2 l%ad8 26 c5 'fIc7 27 cxd6
7 d5 .i.xd6
The pawn chokes Black. It is quite
hard to undermine it.
7 ... 0-0 8 e4 d6 9 g3 c6 10 dxe6
White hopes to get play on the

e6-square which is why he gives up the
strong pawn.
10 ... fxe6 11 ttJd4 .i.c8 12 i..g2 e5
13 ttJf5 ttJa6 14 0-0 ttJc5 15 .i.e3 ~h8
28 ttJc5
By now the position has crystallised
thanks to very fine positional play by
Euwe. Black has two many weak
pawns.
28 ... ~xc5 29 .xc5 .J:1xd2 30 .xf8+
.i.g8 31 :a 1 'tlfb6
16 ~xc5
Rather a surprise. Euwe volWltarily
keeps on taking pieces.
16... bxc5 17 b4 cxb4 18 8xb4 .i.e6
19 ttJe3 a6 20 'tWd3 ttJg4 21 ttJxg4
.i.xg4
193
Max Euwe the 5 th
32 ~n l:tb2? 15 f4 'ike7 16 l:tadl l::tcdS 17 ~hl

A bad blunder in an inferior position.
After 32 .. J1d4 33l:ha6 'iixb4 34 'it'xb4
"f7 IS 'ii'c2 WhS
l:bb4 35 f3 Black would suffer with his

pawn structure.
33 ~c4 'iixf2+ 34 'it'xf2 .tt.xfl
35 ~xgS 1-0
G.Kasparov - V.Kramoik
Botvinnik Memorial 2001
1 d4ltJf6 2 c4 e6 3ltJc3 ~b4 4 e3 b6 19 b4

5 ltJge2 ~b7 6 a3 ~e7 7 d5 Just like Euwe. However, under
I happily followed the same different circumstances. I wish I knew
variation. why Krarnnik refrained from placing
7... 0-0 S ltJg3 his knight on c5 earlier. Because of this
I too did not allow the g3-square to I was prevented from taking it with the
remain vacant. bishop. You know, it's hard to adjust to
S... l:eS 9 ~e2 ~f8 10 e4 d6 11 0-0 new situations.
ltJbd7 12 ~e3 c6 13 'iid2 l:tcS 14 dxe6 19... e5
Playing 7 d5 was somehow a way of

following Euwe, but this is an opening
line. On the other hand here it is in
black and white that I am behaving like
a pupil fo1l9wing his instructions. But I
should have been more cautious as,
after all, White hereby gives up
his centre. But Euwe played this 20 ltJf5
because he appreciated the special Missing the best move because of the
circumstances. He did such a desire to obtain a certain type of
tremendous job with his writings, position against a particular player. It is
adding significantly to chess culture. interesting that before he beat me I
How unfortunate he did not publish his called him Vladi. But when he didn't
analysis here. dare to give me a rematch I referred to
14... fxe6 him as Krarnnik. We know language
I knew I would not be able to attack constantly changes, but Russian has not
e6 like Euwe, but I was able to gain changed that much in the last few
space. years.
194
Max Euwe the 51h
My mam weapon throughout my 22 .. JWg6 23 ~f3 l::tc8 24 'iibl e4

career was to stir up complications and 25 tiJh4 'ikfl 26 .i.e2 a6 27 tiJc3 dxc4
because of my special orientation 28 'iVb2
when there was a jWlgle of variations I
outplayed my opponent many times.
However after my game with the black
pieces against Kramnik at Linares 2000
I realised that I should play for open
positions. In our match I did not
succeed in getting those positions
because of the danmed Berlin defence.
Later, in the first rapid game of the
2001 Botvinnik Memorial, I managed
to beat Kranmik after obtaining an open
position, despite the fact that he had an 28 ... b5
edge early on. I was expecting to create some
Here, objectively, keeping the weaknesses, instead I had to live with a
position closed with 20 f5! was a better protected passed pawn. Slowly my
plan. I have never investigated deeply compensation for the pawn was
how far the Dutch world champion dissipated.
adapted his way of playing against
29 tiJf5 tiJd5 30 tiJxd5 ihf5 31 tiJc3
particular opponents, or whether he
tiJf6 32 h3 ':cd8 33 lbd8 lhd8
played the same way whoever he faced.
34 l:tdl
I see now that I had made a mistake
in this game, but why did I receive such
harsh pWlishment?
20 ... d5 21 exd5 cxd5
34 ....:xdl+
With an extra pawn it is not out of
22 tiJb5 place to ex.change. Tigran Vartanovich
Maybe White is still okay, but from might have played 34 ... .:d3.
now on Black is kicking. After 22 fxe5 35 .i.xdl 1fd7 36 ~e2 ~c6 37 'ii'c1
tiJxe5 23 cxd5 was nice for White and g6 38 1fgl ~g7 39 ~d4 <it'g8 40 ~e5
in little danger of losing. tiJd5 41 ~xg7 Wxg7
195
Max Euwe the 5 th
42 ciJxd5 it.xd5 43 'it'c5 ~a1 + 46 'ii'c7

Putting the queen on the edge of the I try to pin like Karpov, but for me
board is fraught with danger. Better here it did not work.
was 43 ... 'iid7 46 ...'ii'f6 47 iLd1 h5 48 a4 ~d4
44 <iPh2 it.f7 45 it.g4 49 it.c2 e3 50 f5 e2
If 45 'jia7 'ilff6 (but not 45 .. .'ii'xa3?)
45 ... ~g7
51 fxg6 e1='iV 52 'it'xf7+ ~h6 0-1
Time and again Euwe was happy giving up the e4-square in the King's
Indian. And unfortunately I too didn't mind giving up the e4-square - twice!
G.Fontein - M ..Euwe A.Veingold - G.Kasparov
Kleefstra - M.Euwe FRITZ X3D - G.Kasparov
196
Max Euwe the 5th
G.Fonteio - M.Euwe 20 ... lLlf4

Dutch Championship, Black's pieces are alive and
Amsterdam 1924 the f4-knight has special vitality.
Nevertheless White panics by taking
1 d4 tDf6 2 tDo g6 3 c4 ~g7 4 tDc3
0-0 5 e4 d6 6 b3 e5 7 d5 lLlbd7 8 ~d3 the f4-knight, thus opening up the g7
h6 9 ~e3 ~h7 10 "c2 b6 bishop and giving Black the e5-square.
Players no longer play ... b6 in such 21 ~xf4 exf4 22 tDc3 .i.xd3
positions. 23 ihd3 ttJe5 24 'iIfe4 O! 25 g3
11 0-0 lLlc5 12 lLlel lLlb5 13 ~e2
25 ...'ilfg5
13 .....b4 14 b4 lLla6 15 lLlo "e7 Euwe plays fluent chess.
161bbl f5 17 exf5 26 .1:.fdl l:tae8 27 lLlb5 "d8 28 .1:.b3
Or 17 l:.fel fxe4 18lLld2!? 'ii'd7 29 ~h2 .1:.f5 30 g4?
17....i.xf5 18 .i.d3 Once more Fontein panics. This time
See diagram on page 196. he creates an even bigger problem.
30 lLld4 lLlxc4 31 ttJe6 lLlxd2 32 l:txd2
18... lLlb8 c6 leads to a position in which he is just
Playing such a casual move on the a pawn down.
other side of the board shows that he
is oot paying much attention to the
e4-square at all.
19 lLld2 lLld7 20 lLlce4
30 ...l:.g5! 31 "f4 lLlxg4+ 0-1

This sacrifice wins against any reply
by White.
197
Max Euwe the 5 th
Kleefstra - M.Euwe
Amsterdam Chess Club Championship
1927
1 e4 g6 2 d4 d6 3 ~e3 lOf6 4 lOd2

~g7 5 ~e2 0-0 6 e3 eS 7 dS lObd7
8 'ife2 b6 9 h3 as 10 a4 lOeS
20 ... ~b7
Interestingly, Euwe did not mind that
both his bishops had very limited
prospects on their respective diagonals.
21 'ii'd3 1:[£8 22 lOe4 ~c8 23 lOg5
'ii'f6 24 lOe4 'tiff5 25 l:hel 'ifh5
26'ii'g3
H ~xeS?
This is clearly not a testing move.
H ... bxeS 12 ~bS lOhS 13 lOdf3 f5
14 lOe2 fxe4 15 'ii'xe4 lOf4 161Oxf4
26 ... l:.h4
Euwe puts his pieces on the edge. I
played something similar when I lost to
Ivanchuk at Linares in 1991. I did not
realise this game might have had an
16 .. Jbf4
effect in that respect as well. This was
There is no black knight to go to a
the flrst tournament I did not win for
vacant e5-square. This explains why he
almost a decade. What a pity I was not
captured this way.
17 1fe2 'itf6 18 lOd2 able to make it a full ten years.
27 'i!t'g5?
See diagram on page J96. White blunders a piece in a playable
position.
18 •• .'i!i'n 19 0-0 ~h8 20 ~e4 27 ...'ilhg5 28 lOxg5 lbe4 0-1
198
Max Euwe the 5th
A. Vein gold - G.Kasparov 18 ... tl)xf6

USSR Spartakiad 1979 My knight never reached h5 as
Euwe's had done.
1 d4 tl)f6 2 tl)f3 g6 3 e4 .i.g7 4 tl)e3 19 'ilr'd2 tl)cs 20 -te2 ~g7 21 :tae1
d6 5 e4 0-0 6 i.e2 eS 7 dS as 8 .i.gS h6
tl)b3 22 'ii'd3
9 -th4 tl)a6 10 0-0 'ii'e8 11 tl)d2 tl)h7
12 a3 fS
22 ... gS
13 exfS There is very little dynamism in
In 1977 I played a game against Black's position.
my trainer Nikitin in which he 23 -tg3 tl)eS 24 '6'd2 'ilf7 25 h4
replied 13 f3. That game ended in a tl)h7 26 i.xh7 gxh4 27 ..txeS+ dxeS
draw. 28 ~bl
13 ... -txfS
Maybe I should have taken with the
g-pawn.
14 g4 -td7
15 tl)de4 a4 16 f3 b6 17 ~d3 -tf6

18 tl)xf6+
28 ...'6'f4
Defending the king was also an
unpleasant choice.
29 'ihf4 l:txf4 30 :txeS l:tafS
It looks as though Black has achieved
some activity.
199
Max Euwe the 5 'h
FRITZ X3D - G.Kasparov

Match, New York 2003
I e4 e5 2 lOo lOe6 3 .i.b5 lOf6 4 d3

d65 e3 g6
This time the King's Indian pawn
structure arises from a Spanish.
6 0-0 .i.g7 7 lObd2 0-0 S l:tel l:teS
9 d4 .i.d7 10 d5 lOe7
31 l:te7+
But this check hurts.
31...l:tsn 32 :'xn+ ~xn 33 lOe4
lOb3 34 ~fl .i.xg4 35 ~e3 l:tf5
Stepping into a battery.
36 l:tfl .i.h5
II ~xd7
Without the light-squared bishop
Black's attack develops more slowly.
1l ... lOxd7 12 a4 h6 13 as a6 14 b4
f5 IS e4 lOf6 16 .i.b2 "d7 17 l:tbl
37lOd6+ exd6 3S .i.xfS ~f6 39 .i.e2

lOe5 40 l:th2 ~g5 41 .i.d I 1-0
What a miserable defensive game
that was!
The Veingold accident happened

when I was only 14. Sadly I lost one
game like this when those nwnbers 17... g5 IS exf5 "xf5 19 lOn 'iWh7
turned around to 41! Everything has 20 lO3d2 lOfS
advantages and disadvantages my
memory has served me well during See diagram on page J96.
my career, but in these games
influenced by Euwe it helped only my 21 lOe4 lOxe4 22 l:txe4 h5 23 'it'd.3
opponent. l:tfS 24 l:tbe I l:tn
200
Max Euwe the 5th
Euwe also put his rook and queen on

the g-file, however in this particular
position I drop a pawn.
33 nxe5!
Chopping off my pawn! If the
electricity had gone off maybe the
blackout would have driven the chips
of the computer crazy and they might
have missed this tactical shot. But no
such luck for me.
251Ue2 33 ... dxe5 34 'iVxfS lLld4?!
This is not very natural, yet it defends In a bad position this merely hastens
f2 satisfactorily. the end.
25 ... g4 26 'W'b3 l1afS 27 c5 1i'g6 35 .i..xd4 exd4 36 ne8 ng8 37 'iVe7+
28 cxd6 cxd6 29 b5 axb5 30 "xb5 %:tg7 38 "d8 .tlg8
~h6 31 'iib6 ~h7 32 'iib4 .tlg7?
39 'iVd7+ 1-0
201
Alexander Alekhine the 4th
Alekhine is the last champion to precise to say Alekhine) avoided each

be born in the 19th century. He beat other for nine years and their next game
Jose Raul Capablanca by 6-3 in 1927. took place only in 1936. In fact they
Maybe that match generated the played only three more games against
greatest interest of all matches until each other. In 1935 he lost his title to
World War Two. Looking back, that Euwe, however two years later he took
duel still attracts my interest like few his revenge and-beat him 15 1;2-9 112. He
other matches in the whole history of died in 1946 in Portugal. As Fischer
chess. Alekhine won the all-Russian played a rematch with Spassky, I am
Championship in 1909 for the first the only player who retired as world
number 1, while Alekhine is my only
time. I won it in a tie with Karpov in
predecessor who died as the defending
1988 and on my own in 2004.
champion.
In the late 1920s and early 1930s he Maybe there is a similarity between
enjoyed one of the strongest periods of his play and mine. Both of us are very
domination in chess history. Alekhine hard to follow. Even so I tried to use
and Capablanca (probably it is more some of his ideas.
Alekhine was such an imaginative player, playing some stunning attacks,

but, interestingly, the strongest effect he had on me was positionally: freezing
the Slav bishop on the kingside, opening the position in the centre and then
forcing a win on the queenside where Black misses his Slav bishop.
A.Alekhine - E.Bogolyubov G.Kasparov - GENIUS
202
A.Alekhine - E.Bogolyubov
Gennany/HoUand 1929
1 d4 dS 2 c4 c6 3 lLlfJ lLlf6 4 lLlc3

dxc4 5 a4 ~f5 6 lLleS e6 7 .i.g5 ~e7
8 fJ h6 9 e4 ~h7 10 .i.e3
10 ...lLlbd7 lllLlxc4 0-0 12 ~e2 16 'ii'xd8!

White is happy to keep on
exchanging - nonnally that would just
help the opponent develop.
16... l:tfxd8 17 lLla2 lLlb8 18 ~fl
lLlc6 19 l1hdl
Intending to exchange even more
pieces.
19... lLld4 20 l:Iac1 ~f8 21 .i.n lLle8
22 ttJc3 f6
Black's last three moves indicate that
12 ... cS he cl.early wants to bring the bishop
This is a difficult decision. In a way into play.
it helps, as Black exchanges the d4- 23 lLlaS nab8
pawn, plays e5 and f6 and the bishop
may come back. Then, with the control
over the b4 and b3 squares, the flow of
play could go Black's way. On the
other hand this may prompt the
following plan: by opening the
queens ide and exchanging a number of
pieces he will effectively have an extra
piece with which to invade the
queenside. Let's have a look to see
whose strategy prevails. 24lLlb5!
13 dxcS ~xcS 14 ~xcS! Further exchanging.
The fewer pieces on the queenside, 24 ... ttJxbS is lhd8!
the easier it is to invade. Enviable dedication to the task.
14... lLlxcs 15 b4 ttJa6 25 ...l:txd8
203
he r~eived for his win over Petros ian,

but it may have just been a copy, almost
plagiarising Alekhine when he took a
bishop with his knight on d7! If there is
a reprint of My Great Predecessors I
may add an extra comment on that.
39 ... rJilxd7 40 .i.n l'.lb6 41 c;t>eS %:tb8
42 hS cwt>d8 43 .i.e6 ~e7 44 :a3 ~f7
45 .i.e4 ~e7
26lUxb7!
He is not only playing a great
strategic game, but the tactics are on
Alekhine's side too.
26 .. J~b8 27 lUeS rJile7 28 axbS
It was not necessary to accept the
doubled pawns. After 28 .i.xb5 lUd6
29lUa6 :b7 30 iLc6 wins.
28 ... lUd6 29 ~al lUe8 30 ..te4 .i.g8
31 f4 ..tn 32 eS
Alekhine pushes his opponent back. 46 cwt>c6
32 ... fxeS 33 fxeS l::tb6 34 rJile3 iLe8 Finally the world champion invades
35 l::taS ~d7 nicely with his king.
The bishop finds another diagonal 46 ... Wd8 47 l:td3+ ~e7 48 We7 1-0
but it is not too active here either.
36 rJild4 ..te8 37 h4 iLd7 38 iLe2 G.Kasparov - GENIUS
1:tb8 PCAlIntel-Grand Prix rapid 1994
1 e4 e6 2 d4 dS 3 lUn lUf6 4 'ii'e2
dxe4 5 'it'xe4 .i.fS 6 lUc3 lUbd 7 7 g3 e6
8 iLg2 .i.e7 9 0-0 0-0 10 e3 lUe4
11 ~e2 ~b6 12 l:tdl l::tad8 13 lUel
lUdf6 14 lUxe4 lUxe4 15 n lUd6 16·a4
~3
39lUxd7!
In order to invade, he keeps on
exchanging. Now, when I analyse this
game, it occurs to me that maybe
Fischer too was copying the other
champions! How many congratulations
204
17 e4 29lbd8+
I was optimistic about this game as The first element was not in my plan,
the bishop on g6 is really out of play. I wanted to exchange queens not the
17....i.g6 rooks but I thought, okay, it is after all
See diagram on page 202. an exchange.
18 nd3 "'b4 19 b3 ~c8 29 .. Jlh:d8 30 ~f1 b6 31 'it'c3
20 ttJc2 1fb6 21 .i.f4
The computer's 24 ... e5 was strong as
it gained space.
31. .. f6 32 ~c4+
21. .. c5
All goes according to the super
instructive Alekhine game!
22 .i.e3 cxd4 23 ttJxd4 .i.c5 24 %tad1 32 •.. ~ti
e5 25 ttJc2 lbd3 Black offers another exchange which
I am not at all against exchanges. fits into my plan, but I was no longer
26 "'xd3 ttJe7 27 b4 .i.xe3+ happy as the computer's bishop was
supposed to be frozen on the kingside.
33 ttJe3 'it'd 4
Black keeps exchanging!
34 .i.xti+ ~xti 35 1ib3+
Maybe entering an equal knight
ending would have been more practical
against a never tiring opponent.
35 ...~f8 36 ~g2
If 36 ~fl!? 'it'd2 37 ttJc4.
36 ......d2+
28 'ii'xe3
Further swapping just like Alekhine
and an invasion will not be long
coming as well. I was already thinking
how nice it would be to penetrate with
my king.
28 ...%td8
The computer doesn't know those
classical games. It (or should I say he
or she) seems to play into my hands.
205
37 <it>h3 Going a pawn down is already a very

Alekhine had activated his king. I gloomy prospect.
wanted to start a march as well, but on 46 'it'd2 'it'f4+ 47 ~g2 'ii'd4 48 'it'xd4
the edge the king remains isolated. The knight ending gives little hope of
37 ... 'it'e2 38 4Jg2 h5 39 'it'e3 'it'c4 survival.
40 ~d2 'it'e6+ 41 g4 hxg4+ 42 fxg4 48 •.. exd4 49 4Jc4 4Jc6 50 b5 4Je5
White's pawns have been separated 51 4Jd6
and e4 is now a target.
42 .. .'ifc4 43 'it'e1 'iVb3+ 444Je3 'it'd3
45 <la>g3
51...d3 52 <it>f24Jxg4+ 53 ~e14Jxh2

54 <iti>d2 4Jf3+ 55 <it>xd3 <la>e7 56 4Jf5+
wn 57 <la>e4 4Jd2+ 58 Wd5 g5
45 ... 'ihe4 59 4Jd6+ <iti>g6 60 ~d4 4Jb3+ 0-1
Alekhine caught some of his opponents on their back rank. I was really
impressed by those exampfes and it is natural that I wanted to do something
similar to that...
E.Bogolyubov - A.Alekhine A.AJekhine - E.Colle
206
... so here are a couple of my attempts at checkmate on the back rank.
G.Kasparov - N.Sbort DEEP BLUE - G.Kasparov
E.Bogolyubov - A.Alekhine 21 :lac1 l:ad8 22 d6 liJbxd6 23 liJxd6

Game 22, World Championship match lbd6 24 'ii'xa6 'iVd7
GennanyfHolland 1929 The position is equal.
25 :te2 e5 26 a4 f4 27 ~d2 g5
1 e4 e5 2 liJfJ liJc6 3 ~b5 a6 4 iLa4
d6 5 c3 ~d7 6 d4 g6 7 ~g5 f6 8 ~e3
liJb6 9 0-0 ~g7 10 b3 liJf7 11 liJbd2
0-0 12 dxe5 dxe5 13 ~c5 lIe8 14 .i.b3
b6
28 "fibs
Bogolyubov panics unnecessarily, as
Black's attack is not yet that dangerous.
28 ..."fixb5 29 axb5 1:d3 30 1:alliJd6
15 ~e3 "fie7 311:a6
Alekhine stabilises his position. The rook leaves the first rank, but it
16 'it'e2 liJcd8 17 .i.d5 ~e6?! 18 e4 is not yet a matter of decisive concern.
~xd5 19 exd5 f5 20 liJc4 liJb7 31. ..llb8
207
d4-pawn. However only one move

had entered my head and as the time
control had been passed I was able
to spend 17 .minutes checking the
variations before writing down ...
41 liJd7!!
Even today, this move still pleases
me. It threatens 42 lOf8+ ~h6 43 l:tb8!
followed by 44 ltJxg6 and 45 lIh8+
mating the boxed in king.
32 .i.e3?
41 .. .l:hd4
White opens the way to his back
Karpov takes the pawn in the hope of
rank. He should exchange the active
swapping queens, but...
rook by 32 ':c3!
43 lIb4!!
32 ... lOxe4 33 .i.xe5
Ruling out his intended 43 ... 'it'f4,
In a bad position he goes for a direct
while upon 43 .. .lhb4 I had planned the
loss. Better was 33 ~e I.
beautiful idea 44 axb4 d4 45 b5 d3
33 ....i.xe5 34 lOxe5 lIdl + 35 ~h2
46 b6 d2 47 b7 dl='it' 48 b8='ii' 'ii'cl
lOd2!
49 ltJxg6 'ifxg6 50 'ii'h8+ 'ifh7
See diagram on page 206. 51 'ifgxg7 mate.
Of course I found this move because 43 ... lIc4 44 l:xc4 dxc4 45 '1IVd6!
of my talent not because of Alekhine 's Now I have a mate from a different
influence! Incidentally I also caught angle. If 45 ... f6? 46 'it'd2+.
Karpov like this in our 1986 world title 45 ... c3 46 'ii'd4
match. And Karpov resigned as the c3 pawn
is lost and with it game, set and match!
World Championship, Leningrad 1986 But back to Alekhine's game:
36 b4 l:e8 37 ltJo ltJxf3+ 38 gxf3
In Game 22 of the match, I had to

seal my move at adjournment. My 38 .•.:eel!
seconds and all the commentators were Setting up a lovely mating net.
now expecting 41 lIb4, covering the 39 ~h3 b5 0-1
208
Alexander A lekhine the 4th
A.Alekhine - E.eolle
Paris 1925
1 d4 dS 2 c4 ~c6
Sill) :ilov played the Chigorin against
me once in our Candidates match fmal.
3 ~O .i.g4 4 .a4 .i.IO 5 eIO e6
6 ~c3 .i.b4 7 a3 .i..xc3+ 8 bIC3 ~ge7
9 l%b 1 l:tb8 10 c.xdS 'ii'IdS 11 .i.d3 0-0
12 0-0 '6'd6 13 'ii'c2
White has got little from the opening. 30.Id7!!
13 ... ~g6?! 14 f4 ~ce7 15 g3 l:tfd8 A wonderful and unusual way of
16 l:tdl b6 17 a4 ~dS 18 .i.d2 cS 19 fS exploiting the weakness of the first
edS 20 .i.IfS c.xd4 21 cId4 lLlde7
rank.
22 .i.b4 'ii'f6 23 .i.Ie7 'ii'Ie7 24 l:tbc1
30 .. Jbd7 31 :e8+ <it>h7 32 l:tcc8
l:tdS 25 .i.e4
Black can do nothing.
White only has a small advantage.
32 ... l:td8 33 l:texd8 1-0
2S ... l:td7 26 dS .f6 27 l:tel l:tbd8
28 'ii'c6
G.Kasparov - N.Short
Game 3, exhibition match,
London 1987
1 lLlo dS 2 d4 ~g4
The bishop develops in a similar
fashion to the Alekhine game.
Naturally I did not anticipate any back
rank chances.
28 ...•gS? 3 lLleS ~fS 4 c4 f6 5 lLlo c6 6 ~c3
Black must be able to hold with e6 7 g3 .i.b4 8 ~g2 lLle7 9 0-0 0-0
28 ... ~e7! 10 'lib3 as 11 a3 .i..xc3
29 .i.Ig6!! hIg6
29 ... fxg6! would have prolonged

the game and delayed resignation.
Nevertheless after 30 'it'e6+ l:[f7
31 l:tc8 l:lxc8 32 .xc8+ l:lf8 33 .e6+
.J:tf7 34 d6 White's advantage is
overwhelming.
209
12 bxc3 It is worth giving up a pawn in order

White has emerged from the opening to open the position.
with a small advantage, but it will 39 ... 4Jxd5 40 'WWcS JtxfJ 41 1:[xbS
evaporate. l:tc7
12 ... 4Jd7 13 4Jd2 84 14 -'a2 Jtg6
IS e4 Jtn 16 ':bl l:tb8 17 'WWc2 b5
18 cxdS cxdS 19 -.d3 "as 20 l:tel
':fc8 21 l1b4 4Jc6 22 :b2
42 :b8+
Suddenly Alekhine's back rank
checkmates came to mind. I should
have just taken the a-pawn after 42 l:ta5
22 ... 4Je7 and pushed my own a-pawn. 42 ... 'iVc8
By now Nigel has equalised. There 43 'itd6.
follows a long manoeuvring phase. 42 ... ~h7 43 'WW1'8
23 1:[c2 liJb6 24 h4 :b7 2S Jth3 :c6 See diagram on page 207.
26 ':b2 4Jc4 27 :b4 'itc7 28 liJxc4
:xc4 29 Jtd2 -.c6 30 eS fS 31 Jtn I was trying to catch the h7-king just
JthS 32 'WWe3 h6 33 :ebl ~n 34 ':lb2 like Alekhine.
lPg8 3S fJ 'WW86 36 :bl 43 •••'it87+ 44 cwt>n :e7 4S111b2
45 ... ~g6!
36 ... 4Jc6 Nigel had a similar win against
Nigel sacrifices the exchange. The Timman in an Alekhine defence in
position is very closed, so the move is Tilburg 1991! In that game he set up a
justified. mating net, here Nigel escapes with his
37 ~xc4 dxc4 38 1:[4b2 4Je7 39 d5 king intact.
210
46 ~el <i!i>h5 47 11a8 'ii'e5? 3 exd5 'ii'xd5 4 d4 tt)f6 5 tt)f3 ~g4

Here 47 ... tt)e3+ wins. 48 ~el 6 ~e2 e6 7 h3 ~h5 8 0-0 tt)c6 9 ~e3
(48 ~xe3 'iVxe3 49 .xe7) 48 ... ~xa8. cxd4 10 exd4 ~b4! 11 a3 ~a5 12 tt)c3
48 %te8? 'it'd6 13 tt)b5 'ile7 14 tt)e5 ~xe2
White could keep Black busy with 15 'iVxe2 0-0
defensive duties. 48 l:te8 should have
been played.
Black has equalised.

16 :ac1 :ac8 17 ~g5 ~b6 18 i..xf6
48 ...•xa3 gxf6 19 tt)e4 l::tfd8 20 tt)xb6 axb6
Suddenly Black's pieces can attack 21 l::tfdl f5 22 'ife3 iff6
White's king, leaving him lost.
49 g4+ ~xg4 50 Ibc4 .al 0-1
I trusted the back rank attack of

Alekhine so much that I even played
for it against the calculation monster
super computer Deep Blue. Normally
knowledge is an asset but, though I
knew Alekhine's back rank tactics,
after my experience in the next game
they left me feeling blue. Even today, a 23 d5!
decade later, whenever I think of that I was taken completely by surprise.
computer, it makes me deeply blue! This is the kind of positional sacrifice
computers are not supposed to play.
DEEP BLUE - G.Kasparov Later we found that by sheer brute
Match, Philadelphia 1996 force Deep Blue had calculated that it
could win back the pawn after 23 'ir'g3+
1 e4 c5 2 e3 d5 <i!i>f8.
I beat Sveshnikov with the other 23 ...:xd5
main line in the USSR Championship Friedel pointed out that White is
at Minsk 1979. My opponent missed a better after 23 ... exd5 24 .xb6 .xb2
very interesting blockade of my king in 25 1t'xb7 :b8 26 'ii'xc6 11xb5 27 ':c3.
a bishop ending. 24 :xd5 exd5
211
Alexander Alekhine the 4 rh
291tJxb7
Sadly White has time to grab a pawn.
29 ... ~e5 30 'iVd5 f3 31 g3 ~d3
Forcing my way through on the g-file
with 31...'iVf4 did not work.
25 b3!
Deep Blue adopts my style! A quiet
move after a sacrifice. Maybe I have a
way to stay in the game, but it is very
hard to find among the many
complicated variations.
25 ...'it'h8
I have already shown games in which
32 l:tc8!! (32 ~h2? l:txg3!! and Black
I tried to force a checkmate on the g-
file. I think Botvinnik passed on this mates) 32 .. :ifg5 33 ~d8! leaves Black
idea to me. After 25 ...IiJe7 26 'it'g3+!? in trouble. White can control matters
(This is Nunn's suggestion. After with 33 l:tc5! as well.
26 l:Ixc8+ ~xc8 27 'ii'e8+ rj;g7 28 But not 33 h4?? when 33 ...1txc8!!
'it'xc8 'ii'a 1+ 29 rj;h2 'ife5+ 30 g3 'ii'e2 34 bxg5 l:tc I + 35 ~h2 ~g4+ 36 ~h3
31 'ii'xf5 'ii'xb5 Black should hold.)
~xf2+ 3 7 ~h4 l:th 1 is mate.
26 ... 'it'f8! and Black stands his ground.
Alternatively, after 25 .. J~d8 26 'ifxb6 32l:tc7 lIe8
.J:[d7 White has a small edge. After 32 ... ~f4 33 'iVxf3 .
26 'it'xb6 %:tg8 27 'iVc5 33 ~d6!
Not 27 'it'xb7?? 'ii'g5.
After 33 1ixf7 the Alekhine-like back
27 ... d4
If 27 ... 'iVg5 28 g3 'ii'd2 29 ~d6 rank play 33 ... l:tel + 34 <iPh2 'ii'xf7
Black's pawns are all separated. 35 lhf7 occurred to me and some
28 ~d6 f4 commentators thought it would win as
well. But after 35 ... ~e5 36 :f8+ 'it'g7
37 ltd8 ~d3 38 g4 ~xf2! (Playing on
the back rank, just like Alekhine.)
39 ~g3 (On 39 ltxd4 ~hl! Black is
better. But not 39 ... ~e4 40 l:txe4 lhe4
41 rj;g3) 39 ... ~xh3 40 <it>xf3 lte3+ the
game ends in a draw.
33 ... lte1 + 34 <iPh2
212
36 tiJg5+
We have arrived at the next motif I
learned from Alekhine and this position
is an example of his effect on me. I
underestimated the power of the
discovered check arising from the
battery.
36 ... <iii'h6 37 :'xh7+
34 ...tiJxfl
I set up a mating net just like
Alekhine. But there is a small
difference between our games - my
checkmate can be parried.
35 tiJxti+ ~g7
If35 ... 'ihf7 36 'ii'd8+ ~g7 37 l:txf7+
~xt7 38 'ifd5+ ~e7 39 'ii'xf1 wins.
1-0
I resigned as I drop the f1-pawn
which cages in the king. Without it I am
just desperately lost.
After 37 ... ~g6 38 'ii'g8+ ~f5
comes 39 tiJxD and now Black's
mating threat has disappeared and I am
hopelessly behind on material.
I continue to show Alekhine's effect on me when it came to handling

batteries. In the next game he rightly ignored the power of the opponent's
battery.
M.Euwe - A.Alekhine V.Kramnik - G.Kasparov
213
Most chessplayers know that 14 ... cxdS IslDxdS ~d7 16 lDd4 f4

Alekhine and Euwe played two
matches for the world title. However
they also had a third or should I say a
first match in 1926. Later on Euwe also
played a match against Capablanca,
who beat him 6-4 with no losses.
Bogolyubov also beat him 5 1/Z-4 1/z in
two different matches.
In the Alekhine-Euwe match of 1926

Alekhine won two games early on but 17 l:del?
lost the seventh and the eighth, and This is a bit too subtle. Winning a
they drew the ninth. We look at their pawn for nothing with 17 b4 was the
decisive last game with the match simplest. 17 ... 'ifd8 18 lDxf6+ 'jixf6
standing at 41/z-41/z. 19 ~xb7.
17 ... lDxdS 18 ~xdS+ Wh8 19 lDe6
1:(f6 20 lDgS
M.Euwe - A.Alekhine
Game la, match, Amsterdam 1926
1 lDf3 e6 2 c4 fS 3 g3 lDc6 4 d4
~b4+ S ~d2 ~xd2+ 6 'ifxd2 d6
7 lDc3 lDf6 8 ~g2 0-0 9 :1dl lDe7
10 0-0 lDg6 11 'if c2 c6
20 ...:afS
Black's pieces come into the game.
21 'iib3 fxg3 22 'ifxg3
White has compensation for the
pawn after 22 bxg3 l:xf2 23 'ilfe3.
Black's pieces have now become
threatening.
23 l:1e7 l:1g6 24 :In??
12 e4 Even after 24 lDf7+ l:xf7 25 l:1xf7
White occupies the centre. lDe2+ 26 ..t>hl lDxg3+ 27 fxg3 h5
12 ...'ifaS 13 exfS exfS 14 dS! 28 :Ixd7 White is worse, but it is far
Euwe cuts Black's camp into two. from over.
214
Alexander Alekhine the 4tiJ
24 .. .lhti 25 ttJxti+ ~g8 In the match the Griinfeld did not

work well. It is a pity Griinfeld did not
See diagram on page 2 J3. become a world champion, then I could
blame him for losing the second game
Alekhine intentionally steps into the of this match.
discovered check of a battery. I knew 4 e3 0-0 5 ~d3 d5 6 ttJf3 c5 7 0-0
this example when I played my last cxd4 8 exd4 dxc4 9 ~xc4 b6
World Championship match. I employed Karpov's variation, but
26 ttJe5+ Alekhine is also not 100 percent free
from guilt for this loss.
10 ~g5 ~b7 11 nel ttJbd7 12 nc1
nc8 13 'ii'b3 ~e7 14 ~xf6 ttJxf6
15 ~xe6
26 ...'ii'xd5! !
Euwe must have missed this.
27 cxd5 ttJe2+
And Euwe resigned.
0-1 15 ... fxe6
I've already showed you that I was
You might say Euwe was unlucky. not worried by the battery, and in fact
Yes and no. Looking at this match only, this time the battery is not dangerous
yes, but probably he learned from the after 15 ... ltc7!. Then 16 ttJg5 (16 ~c4
extremely tense situation. When he .i.xf3 17 gxf3 ltd7 and Black is safe.)
played the last game of the 1935 match 16 ... 'ii'xd4! 17 ttJe2 (17 ttJxf7 .i.c5
he was able to handle the pressure 18 ttJe5+ 'it>h8 is okay; 17 l:tcd 1 'irb4
and he was successful. He probably 18 ttJxf7 .i.c5 and Black can move
became wiser because of this painful despite the discovered check.) 17 ... 'ii'd2
experience, whereas I derived no 18 ttJxf7 'ii'xe2 and Black is doing all
benefit from it because I got no chance right. In this game all the motifs are
of a return match. here that I picked up from Alekhine -
the back rank included.
16 'ii'xe6+ 'it>h8 17 'ii'xe7 .i.xf3
18 gxf3
London 2000
After 18 W'xd8 ncxd8 19 gxf3 comes
1 d4 ttJf6 2 c4 e6 3 ttJc3 ~b4 19 ... l::txd4.
215
18 ... "xd4 19 tObS 24 tOg5+ ~h8 25 'iff5 'ifxa2 (25 .....c3

26 :e6 [26 l:te7 'ifc5] 26 ... 'ii'c7
[26 ... rJ8>g8 27 h4!] 27 1i'g6 ':'f8 28 ':'xf6
gxf6 29 'jjb6+ lfrg8 30 'ii'xf8+ a nice
check follows.) 26 "g6 'iVa3 (26 ...:f8
27 :e7) 27 l:te6 ~g8 28 h4 White wins.
If 23 ... h6 24 ltJxh6+ ~h7 25 'ii'f5+!
~h8 (25 ... ~xh6 26 'ii'f4+ ~h7
27 1i'xb8 This is why putting the
rook on a8 creates a bigger obstacle.)
26 tOf7+ 'lPg8 27 tOg5 'iVxa2 (27 ......c3
28 ':'e7 'iVc5 29 Ihg7+) 28 ':'e6 ':c8
19 ..... xb2 (28 ....:.e8 29 ':'xe8+ fOxe8 30 ~g2
The excellent Hungarian junior wins; or 28 ... l:td8 29 'lPg2) 29 "'g6
trainer Hazai, playing the White pjeces, wins.
had a game where he was faced with 24fOd8+
19 .....f4. He did not mind taking a The battery looks innocent as the
walk in the centre. 20 ':'xc8 l:txc8 knight wins nothing, however it still
21 tOd6 'iVxD 22 tOxc8 'ii'g4+ 23 ~f1 blocks the eighth rank.
'jjb3+ 24 ~e2 'ii'xc8 25 ~d2 and went
on to win in Hazai-Danielsen, Valby
1994.
20 ':'xc8 lbc8 21 tOd6 l:tb8
21 ... %:ta8 was better as the rook would
be less vulnerable if Black's king goes
in front of its pawns. 22 tOf7+ ~g8
23 'ii'e6 h6 24 tOxh6+ ~h7 25 tOf7!
(25 tOg4 11£8) 25 .. .l:te8 (25 ... 'ifd2
26 :e4!?) 26 'iff5+! and White has
good winning chances.
22 tOf7+ Wg8 23 'iVe6 24 ... ~h8
I did not have the same luck as
See diagram on page 2 J3. Alekhine and was not able to take the
piece at the base of the battery.
This is the battery I did not mind. 25 'ii'e7 1-0
Thanks to Alekhine. I resigned, because the d8 knight
makes it possible to trap my rook or,
23 ... l:tf8 if it moves away, there is a back rank
Trying to include the rook in the checkmate or even worse a nightmare
defence, but in vain. Other moves did smothered mate. All Alekhine's motifs
not help either. Kramnik showed how that I wanted to employ played a role in
White wins with 23 ... h5. Then this game. I knew them all and still lost.
216
Alexander Alekhine the 4'h
After this game started to match but, unlike Euwe, I was

understand that the crown won't stay inexperienced when I played this game
with me till I die as happened with against Kramnik.
Alekhine. It was a very important Indirectly, Alekhine affected me as
game.
well, since Kramnik learned from him
The world champions have had a how to avoid a rematch. He never gave
strongly negative effect on me, me a chance to prove my superiority
something which has probably escaped over him in a match, the same way
everybody's attention. I learned from that Alekhine denied Capablanca. So
them never to be behind at a decisive Alekhine had an especially marked and
final stage of a World Championship controversial influence on my career.
217
Jose Raul Capablanca the 3 rd
Capablanca beat Emanuel Lasker in strong player who beat him and never
1921 to become the third world gave him a chance for a rematch. What
champion. Capablanca won 4 games did I pick up from him? Well, there are
and 10 were drawn, therefore the result a few indications of his style in my play
was 9-5. The match took place in Cuba, . but these are rather superficial. First I'll
where the weather favoured him, but show you the games that inspired me -
anyway the great Cuban was destined and also cost me dearly.
to become world champion. Actually
he may have been the best player J.CapabJanea - L.Molina
earlier or, more precisely, during World Casual game, Buenos Aires 19 I I
War I. At that time there was hardly
1 d4 dS 2 e4 e6 3 ltJc3 ltJf6 4 ~gS
any chess activity in Europe and
ltJbd7 S e3 e6 6 ltJo ~e7 7 exdS?!
Capablanca was performing at his best
ltJxdS 8 ~xe7 ltJxe7 9 ~d3 eS 10 0-0
in America. In a way, among the world
0-0 11 dxeS ltJxeS
champions he is the most dissimilar
type of player to myself. He had a very See diagram below.
positional style, whereas I prefer
complications. He was a laid back 12 ~xh7+!?
easy-going person, whereas I am a hard Capablanca had a positional style,
worker and have conflicts. Of course but also very sharp tactical vision. It is
there are similarities too, He was very very hard to foresee all the components
talented and had better results in of this sacrifice.
individual tournaments against a very 12 ... <iti>xh7 13 ltJgS+ ~g6
Because Capablanca had won with the bishop takes h-pawn sacrifice,
I settled for a draw when I was faced with it - instead of trying for more.
J.Capablanea - L.Molina G.Kasparov - DEEP JUNIOR
218
Jose Raul Capablanca the 3Td
13 ... c;t>h6?? loses to 14 lLlxf7+; and

there is no way back with 13 ... <it'gS as
after 14 '6b5 neS 15 ltadl ~d7
16 'ii'xf7+ ~hS 17 'ifh5+ ~gS IS b4
wins.
14 'it'g4 f5
This is the only move.
If IS ... e5 19 nad 1 'iVb6 20 'ifh4 i..e6

21 4Jxe6+ <it>xe6 22 exf5+ <it'fl
23 'ifc4+ cj;eS 24 b4 4Ja6 25 4Je4 and
White has nice prospects.
19 exf5 exf5 20 l'%adl 4Jd3 21 'ir'h3!
21 'ifh5 could lead to a perpetual.
21.. .nhS 22 tiJce4+ fxe4 23 4Jxe4+
cj;fl 24 4Jg5+ cj;f6 25 tiJe4+ rJ;fl.
15 -'g3 <it'h6 21. .. 4Jdf4
After 15 ... f4 16 exf4 lLlf5 17 'ir'g4 There's nothing better. If 21...lthS
4Jh6 IS 'ifh4 White has two pawns 22 1r'e3 4Jgf4 23 g3 and White wins
back the piece. Or 21...4Jgf4 22 'ir'g3
and play for the piece enough
l:.hS 23 h4 i..e6 24 :fe 1 l:teS 25 4Je2
compensation for the material deficit.
and Black's position has deteriorated.
And after IS ... l:thS 19 l:tfdl 'ir'f6 22 1r'g3 1kc7
20 l:.ac 1 he will play b4 with an edge. After 22 ...•a5 23 h4.
If 15 .. .'it'f6 16 nad1 'itb6 (16 ... 'ir'a5 23 l:tfe! 4Je2+
leaves White two ways to look for play. After 23 ... l:tdS 24 l:lxdS 1r'xdS 25 h4
17 4Jh7+ cj;fl IS 4Jxf8 ~xf8 19 l:td4 or Black is in trouble. And if 23 ... nhS
17 e4 4Jxe4 IS 4Jcxe4+ fxe4 19 4Jxe4+ 24 h4 tiJxh4 25l:td6+! .xd6 26 4Jce4+
White wins.
r/;fl20 b4) 17 e4 (17 4Jh7+) 17 ... lLlxe4
24 l:txe2 'ir'xg3
IS lLlcxe4+ fxe4 19 'ir'f4+ tiJf5
20 4Jh7+ cj;fl 21 4Jxf8 ~xf8 22 'ir'xe4
with easier play for White.
16 'ir'h4+ ~g6 17 'ir'h7+ cj;f6
Not 17 .. 5~xg5?? IS 'ir'xg7+ ~h5
19 lLle2! f4 20 exf4 tiJf5 21 'ifh7+ tDh6
22 4Jg3+ ~g4 23 'ir'xh6 and checkmate
follows on the next move.
18 e4! 4Jg6
219
Jose Raul Capablanca the 3rd
25 lLlh7+ 14 i(.d2 'Wi'h2+ 15 ~f3

This is an effective intermediate
move. It crushes Black's position.
25 ... ~f7 26 hxg3l:th8 271L1g5+ ~f6
28 f4 1-0
G.Kasparov - DEEP JUNIOR

Game 5, Man v Machine,
New York 2003
1 d4 lLlf6 2 c4 e6 3 lLlc3 iLb4 4 e3

0-0 5 iLd3 d5 6 cxd5 exd5 7 lLlge2 l:te8
15... 'Wi'h4
8 0-0 1l.d6 9 a3 c6 10 'Wi'c2
The pieces are placed in a similar
See diagram 011 page 2 J 8. fashion to the Capablanca game.
10 ... 1l.xh2+?
Seeing this move on the screen had
the effect of a cold shower. It struck me
that perhaps it was my tum to lose to an
opening trap, the same way that Karpov
once lost horribly to Korchnoi. Then,
with the Capablanca game rapidly
flashing across my mind, my heart
nearly missed a byte's worth of beats.
11 ~xh2 ltJg4+
16iLxh7+?
Here I was -virtually settling for
a draw because of the game of the
Cuban champion. I thought it would
be dangerous to play on because
Capablanca had won with the sacrifice.
However I should have continued with
16 g3!! as the centre is not as open as in
the Capablanca game.
Then 16 ... 1L1h2+
12 ~g3 a) 16 ... ~ This is surely the move
The only move - and just like the Capablanca would have chosen. 17 f5
Capablanca game. (17l:tael) 17 ... 1L1d7 (l7 ... h5 18 e4 or
12 ... 'Wi'g5 13 f4 17 ... 'ii'h3 18 l:hl 1Llh2+ 19 ~f2 wins.)
Things are developing in a very 18 ~xg4 'ilt'g2 19 e41L1f6+ 20 ~f4 dxe4
similar way to the Capablanca game. 21 .i.xe4 lLlxe4 22 lLlxe4 'ilt'xe2
13 .. .'1!i'h5 23 l:tae 1 1%xe4+ 24 'ilt'xe4 'ifxd2+
This represents a slight difference. 25 'ife3 and Black runs out of play.
220
Jose Raul Capablanca the 3rr1
b) 16 ... 'irb5 17 l:hl liJxe3+ 18lhh5 Maybe White is somewhat better, but
.i.g4+ 19 <ii?f2 liJxc2 20 l:ah 1 .i.xh5 this is by no means certain and things
21 l:xh5 liJal 22 .i.xh7+ wins. can easily go wrong for White.)
20 ... tLld7 (threatening tLle5 mate) 21 e4
17 ~f2 liJg4+
c5 (Black is not worse in the line
starting 21 ... dxe4+ which might end in
a particular perpetual check. 22 tLlxe4
[22 .i.xe4 tLldf6] 22 ... gxf5 23 tLl4c3
l:r.e3+ 24 .i.xe3 liJde5+ 25 dxe5 liJxe5+
26~f4
Now White's king can run away with

18 ~el!.
Before we continue with this line

let's see the continuation if White's
king stays in the area with 18 ~g2,
which allows Black to hold. There
follows 18 ... 'ifh2+ 19 ~f3 g6!! (19 ... f5 26 .. :iib6+ 27 ~xe5 .g7+ leads to a
20 .i.xf5 'irb5 21 .i.xg4 wins.) draw.) 22 .i.g5 (White should settle for
a draw. It is too risky to play for a win
by 22 ~xg4? cxd4 23 .i.g5 [23 l:th 1??
liJe5+ 24 ~f4 h6 - 24 ... g5+ 25 ~xg5
h6+ 26 <ii?f4 'it'f2 mate - 25 tLlxd4 .f2+
and checkmate on the next move.]
23 ... dxe4 24 lLlxe4 'irb5+ 25 ~f4 gxf5
26 tLlxd4 .g4+ 27 ~e3 .xg5+ 28 ~f2
fxe4 29 .i.xe4 tLlf6 and Black takes
over.) 22 ... gxf5 23 liJxd5 cxd4 24 exf5
1in5 25 <ii?g2 'irb2+ is another
20 f5 (20 e4 dxe4+ [Black can force perpetual.
a draw by 20 ... 'irb5 21 <ii?g2 - 21 l:h 1
liJe5+ 21 .. :ifh2+ 22 <it>f3 ~5] Meanwhile back to 18 ~e I! when
21 .i..xe4 liJf6 22 f5 'irb5+ 23 <it>f2 play continues 18 ...1in3 (18 ... 'iib2!?
lLlxe4+ 24 tLlxe4 'it'xf5+ 25 <it>e3 'it'g5+ 19 tLld 1 tLld7 20 e4 dxe4 21 .i.xe4 1ib5
26 'it>d3 [26 ~f2 'it'f5+ 27 ~e3] 22 l:h 1 tDh2 23 tLle3 tDf6 24 ig2
26 ... iLf5 27 tLl2c3 'it'xg3+ 28 .i.e3 'irb3 liJfg4 25 'it'c5 leaves White an edge.)
221
Jose Raul Capablanca the jrd
16... ~b8 17 ltJg3 ltJh2+ 18 ~f2

ltJg4+
19 :tg 1 (Had not known

Capablanca's game I might have tried
19 ltJd 1 !? I am not saying this wins but
19 ~f3 ltJb2+ l;S-th
I knew he had won with the rook move
And the game ended in a perpetual.
and I trusted his idea.) 19 ... ltJd7 20 e4
But I should have won this game or at
dxe4 21 ltJxe4 'ifh2 (21...ltJdf6 22 ltJd6 least pressed harder. So, planted deep in
lte6 23 ltJc4 and White keeps his my mind in my junior years was
position together.) 22 ~dl ltJdf6 Capablanca's game - and it returned to
23 ltJxf6+ ltJxf6 24 l:tel ~g4 25 'ifc4 me much later when I played Deep
and White is tied up. Junior.
Capablanca beat Alekhine in an Here is my position - also from a

endgame in which he was a pawn up world title match - and with similar
and with a particular pawn structure. characteristics.
J.Capablanca - A.Alekhine G.Kasparov - A.Karpov
222
J.Capablanca - A.Alekhine
Buenos Aires 1927
1 d4 dS 2 c4 e6 3 lDc3 lDf6 4 i..gS

tiJbd7 5 e3 c6 6 tiJf3 'iVaS 7 tiJd2 i.b4
S 'iVc2 dxc4 9 jlxf6 lDxf6 10 lDxc4
'fIe71I a3
21 ~34:Jd5
Here Alekhine missed a chance to
reduce his disadvantage with 21...b5.
Then 22 tiJc5 i.xc5 and Black is only a
fraction worse.
22 b5 cxb5 23 'ilt'xb5 l:taS 24 ':c1
l:ta5 25 1Vc6
White soon wins the b6-pawn.
25 ... i.a3 26 l1bl i.f8 27 i.xd5
l1xd5
1I ...i.e7
There are problems assessing this
type of middlegame. This bishop may
be worth as much as White's extra
space. For example, players once
considered the Moscow variation to be
slightly better for White. By the end
of the 1990s masters began to sacrifice
the c4-pawn instead. I also beat
Dreev with this idea in an extremely
important game in the 2004 Russian 2SlDxb6
Championship. White has excellent chances of
12 g3 0-0 13 i.g2 i.d7 14 b4 b6 converting his advantage.
15 0-0 as 16 tiJe5 axb4 17 axb4 l:txal 2S .. Jtd6 29 ~7 h5 30 4:Jc4 l::td7
31 'il'e4 .J:tc7 32 tiJe5 'ii'cs 33 ~g2 .i.d6
If 17 ... i.xb4 18 lDb5 'fIc8 19 lDa7
34 ltal l::tb7 35 ttJd3 g6 36 l:ta6 .i.fB
wins.
37 ltc6
IS1hal l:tcS It is interesting that Capablanca
If 18 ... i.xb4 19 lDb5 'fIc8 20 i.xc6! chooses not to attack with h3 and g4.
Capablanca liked this continuation and 37 ... l:tc7 3S l1xc7 "xc7 39 ttJe5 i.g7
he had a point. 40 1i'aS+ ~h7 41 ttJf3 .i.f6 42 'ii'a6
19 tiJxd7 'fIxd7 20 lDa4 'iVdS ~g7
223
Jose Raul Capablanca the 3rrl
bishop along both diagonals was better

so perhaps b6 is the best square for the
bishop. Then Black has real chances of
survival.
56 ltJe5! .i.d4
If 56 ... f5 57 d6! fxg4+ 58 <it'g2! wins.
Alternatively, 56 ... .i.a3 57 d6 ~f6
58 d7 rJ;;e7 59 ltJxn is decisive.
57 ltJxf7+ ct>f6 58 ltJd8 .i.b6 59 ltJc6
.i.c5
43 'iVd3
Capablanca starts creating a passed
d-pawn.
43 ... 'iib7 44 e4 'ifc6 45 h3 1r'c7
46 d5 exd5 47 exd5
Had the Cuban not exchanged rooks
ten moves earlier, and if the game had
proceeded in a similar way, by now his
advantage would have been greater.
47 .. :ifc3 48 'ii'xc3
Another small surprise when it was 60 ~f4!
possible to keep the queens on. Capablanca can now use his king and
48 ....i.xc3 returns a pawn for a winning endgame.
60 <ite2 .i.d6 (60 ... <it'g5 61 liJe5)
See diagram on page 222. 61 <it'd3 !itg5 62 ll)d8 ~xg4 63 ~e4
also wins.
49 ~n !itf6 50 <it'e2 .i.b4 51 ltJd4 60 ....i.xfl 61 g5+
.i.c5 52 ltJc6 <itf5 53 ~f3 <itf6 As Capablanca played g5, I too opted
for the same idea - see the next game!
61 ... ~f7
If 61...<it'g7 62 d6.
54 g4 hxg4+ 55 hxg4 <it'g5

An unfortunate move, as it fixes
White's king. However, it does provide
freedom for the knight. Moving the 62liJe5+ q;e7
224
After 62 ... <itg7 63 d6 wins. 18 0-0 a6 19 lLla3 ':e8 20 lLle2 l:lxe2

63 ~xg6+ 21 'il'xe2 ~bS 22 l:lIbS axbS 2311hbS
Winning a second pawn. The rest is l:lIa2 24 ~e3
simple.
63 ...<it>d6 64 <it>e4 ~g3 6S ~f4 q;e7
66 ~eS ~el 67 d6+ ~d7 68 g6 ~b4
69 <it>dS
Not 69 g7?? ~c3+ draws.
69 ... ~e8 70 d7+ 1-0
This game looked very convincing to
me.
Game 40, World Championship, 24 ...11aS?
Moscow 1985 Karpov hangs on to the pawn, but
soon he has to relinquish it. Geller
1 d4 ~f6 2 c4 e6 3 ~f3 dS 4 ~c3 recommended 24 ... lLla6!? which loses
~e7 S ~gS h6 6 ~h4 0-0 7 e3 b6 the pawn but might hold the game.
8 ~e2 ~b7 9 ~If6 ~xf6 10 cIdS For example, 25 l1c 1 (25 ~xd5 l:la5
exdS 11 b4 eS 12 bxcS bIeS 26 tLJe7+ ~f8 27 ~c6 l:bb5 28 ~xd8
~b8) 25 ... g6! 26 ~xd5 (26 ~7 'it'f6)
26 .. .1:135 27 ~e7+ ~g7 28 ~c6 ':xb5
29 ~xd8 ~b4 30 l:lbl l:lb8 and Black
seems to escape.
2S 1fiJ7 'it'e8 26 lLlxdS l:lbS 27 "a8
'it'd7 28 ~e3 ':b4 29 dS 'it'e7 30 tLJdl
11bS
There are 126 games with this

particular position in the database.
Karpov and I played this position five
times in our World Championship
matches, all ending in draws. These
games do not belong to the most
exciting pages of World Championship
history but as far as I was concerned 31 ~e3
they were justified because of my White has consolidated his extra
standing at the time in those matches. pawn. However it is not so simple to
13 l:lbl 'it'aS 14 'it'd2 cxd4 IS lLlId4 convert it.
~Id4 16 exd4 ~e6 17 ~bS 'it'd8 31. .. 'it'aS
225
I had no choice, but it was not 42 l:td4

against my wishes as Capablanca had After 42 gxf6 comes 42 ... ~t7!
exchanged queens as well. 42 ... ~fi 43 lOc4 ~e6 44 ~f3 ltc5
32 'if'xa5 lha5 33 ~d1 45 ~e3 l:tb5 46 'it>d2 ':d5 47 l:txd5
~xd5
See diagram on page 222. We too have simplified to one piece
against one piece. But I was not so
There are rooks on the board and
lucky as Capa.
Black has a knight. But I thought the
48 lOe5 ~xd6
rooks could be swapped and the
difference between having a knight
instead of a bishop is not significant.
Capablanca won his game, therefore I
was hoping to win my endgame as
well.
33 .. .lod7 34 g4 g6 35 ~g2 l:ta4
36 h3 ~g7 37 d6
Capablanca advanced his pawn at a
much later stage.
37 .. .lh6 38 f4 l:tc6 39 h4 ~f8 49 lOxg6
My simplification is different from
the previous game. White's king is
unable to improve its position and so no
progress can be made. Oh, it's all so
sad!
49 ... lOc5 50 lOh4 ~e6 51 ~e3 lOe4
52 1Of3 q;fi 53 ~d4 ~e6 54 ~c4 1Of2
55 ~d4 [55 lOgl!?] 55 ... lOe4 56 1Oe1
~d6 57 lOc2 lOc5 58 ~e3 lOe6
59 lOd4 lOg7 60 ~d2 ~c5 61 <itd3
~d5 62 lOe2 lOh5 63 'it>e3 lOg7
40 g5
64 lOg3 q;d6 65 q;f3 ~e7 66 lOe2 lOe6
The great Cuban pushed the g-pawn
67 lOg3 lOg7 68 1Of1 ~fi 69 lOe3 'it>g6
later. But he pushed nevertheless and
70 lOd5 lOe6 VI-VI
why shouldn't I do the same? That's
what I thought. On the other hand But not all my memories were sour
Geller preferred 40 h5. as regards the particular pawn structure
40 ... hxg5 41 hxg5 f5! with three pawns on the kings ide and
Karpov does not allow my king to one extra d-pawn. Here is my game
penetrate. against Shirov:
226
A.Shirov - G.Kasparov
Astana 2001
22 ... 'ii'c4 23 lIael ~c6 24 ~cl
"xd4 25 ~e2 'ii'a4 26 'ir'g3 l:tfd8
27 ~c3 'ifb3 28 l:te3 d5 29 ~h6 ~f8
30 ~e4 ~2 31 ~c5 ~xc5 32 l:tc3
~xf2+ 33 l:txf2 l1al+ 34 l:tfl 'irb6+
35 lIe3 l:txfl + 36 ~xfl 'ird4 37 'ikc7

'il'c4+ 0-1
I have sacrificed a piece on b5 a few times, of course this occurs most

frequently in the Sicilian defence. So my main recourse to employing this idea
came from other games and I very clearly remembered one of Capablanca's
wins where he sacrificed a bishop on b5. He played it in his first European
tournament at San Sebastian 1911 after which he was universally regarded
as a world-class player.
Here is the position where And here you can see the moment
Capablanca had just sacrificed on b5: when I unleashed the same move.
J.Capablanca - O.Bernstein G.Kasparov - J.Lautier
227
Jose Raul Capablanca the 3 m
J.Capablanca - O.Bernstein
St. Petersburg 1914
1 d4 d5 2 lOn lOf6 3 c4 e6 4 lOc3

lObd7 5 .i.g5 .i.e7 6 e3 c6 7 .i.d3 dxc4
8 .i.xc4 b5 9 .i.d3 a6 10 e4 e5 11 dxe5
22lOxc8
This is a surprising solution. He gives
up his well placed knight for an
undeveloped bishop.
22 ...'ifxc6 23 1Wd8+ 'jie8
If 23 .. .'.pn 24lOd6+.
24 .i.e7+ q;n 25 lOd6+ Wg6
26lOh4+ ~h5 27lOxe8 l:txd8
I have had two games in which my
opponent had doubled e-pawns in the
opening and I lost. One was only a blitz
game again Kramnik, but the second
one was a regular game against Hubner.
Tal, Euwe and Steinitz all won games
against such a pawn structure.
11. .. lOg4 12 .i.f4 .i.c5 13 0-0 'ii'c7
14 l:[cl f6 15 .i.g3 fxe5
Against Hubner I did not mind him 28lOxg7+
taking back the e-pawn, as I based my Capablanca had luck with such
play on my queenside pawn majority. knight-saving intermediate moves.
28 ... ~b6 29lOgf5+ ~h5
16 b4 ~a7 17 ~xb5!
In return for the bishop Capablanca

obtains three pawns and prevents
Bernstein from castling.
17... axb5 18 lOxb5 'ifd8 19 lOd6+
~f8 20 ':xc6 lOb6 21 .i.h4 'ifd7
228
30 h3
White is a rook down but has plenty
of pawns for it and, more importantly,
far too many pieces around Black's
king.
30 ... liJe8
If 30 ... liJh6 31 liJg3 mate.
31 hxg4+ ~xg4 32 ~xd8 l:txd8
Black has avoided direct loss, but he
has given back the rook. Now he is 12 a3 b5
absolutely constrained with his three One year earlier Ivanchuk castled
pawn deficit. against me in this position. Joel
obviously had time to prepare.
13 Ji.xb5!?
See diagram on page 227
I also had time to prepare and felt this

sacrifice would wreak psychological
damage.
33 g3 l:td2 34 ~g2 lte2 35 a4 liJb6 13 ... axb5 14 liJdxb5 'ifb6?!
36 liJe3+ ~h5 37 as liJd7 38 liJhf5 After 14 .....c6 15 ~xc5 dxc5 16 e5
liJf6 39 b5 ~d4 40 ~f3 1:[a2 41 a6 iLa6 17 a4 liJd5 18 liJxd5 exd5
~a7 42 ltct l:tb2 43 g4+ ~g6 44 l:te7 19 liJd6+ ~e7 20 l:te1 h5 21 'fi'f3
l:txf2+ 45 ~xf2liJxg4+ 46 ~f3 1-0
Lautier evaluates his line as giving
enough compensation.
This b5 sacrifice is nice indeed and I
15 ~xe5 dxe5 16liJd6+
have also tried it.
Also after 16 e5 ~a6 17 a4! ~xb5
G.Kasparov - J.Lautier (17 ... liJd5 18 liJd6+ ~e 7 19 ttJxd5+
Euwe Memorial, Amsterdam 1995 exd5 20 'it'xd5 wins.) 18 liJxb5 ttJd5
(18 ... liJe4? 19 'it'f3) 19 c4liJe7 20 "d6
1 e4 e5 2 liJf3 e6 3 d4 exd4 4 liJxd4
White has decent compensation for the
liJe6 5 liJe3 ..e7 6 ~e3 a6 7 Ji.d3 liJf6
piece.
8 0-0 liJe5 9 h3 ~e5 10 ~hl d6 11 f4
liJed7 16... ~e7
229
Jose Raul Capablanca the 3 rd
17lDxc8+ 22 g4!
Capablanca captured the bishop after By now I had already cleared my
the d6-check. I did not do the same head of the Capablanca game. I just
even though I had had the controversial used my own brains and made a
experience when the computer reasonably good move, which creates
sacrificed the bishop on h2. I still fully chances. However, sadly, it is not
trusted Capablanca's way of attacking. enough. If 22 l:lg3 "f2.
After 17 e5 ~a6 (17 ... lDe8 18 lDc4) 22 ... fxg4 23 'it'xg4
18 l:lf2 l:lhd8 19 l:ld2 the position is After 23 hxg4 'it'c6.
unclear. 23 ....:a5!
17... l:lhxc8 Not 23 ... 'ii'xb2? 24 f5.
It is a bit worrisome that Lautier's 24 lDe4 'ii'c6?!
Once again taking the b2 pawn was
rook is not stuck on h8 as Bernstein's
decisive: 24 ... 'it'xb2! 25 lDd6 lDxd6
was. But I still felt relaxed.
26 exd6+ ~xd6! 27 .xe6+ Q;c7
18 e5 lDe8 19 _h5 b6 20 l1ae1
28 ltdl lDb6 and Black is too far ahead
Not 20 f5? lDxe5 21 fxe6 1fxe6.
in material.
20 ... f5!
25 lDd6 lDxd6?!
Lautier consolidates his king and
Impatiently and prematurely parting
now I started to realise things would with the d6'-knight. After 25 ... l:b8
not necessarily go my way. Not 26 ~gl l:ld5! 27 l:lf2 lDxd6 28 exd6+
20 ...•xb2? 21 f5! 'ifxc3 22 fxe6 fxe6 ~d8 Black wins.
23 l:lf7+ ~d8 24 l:ld 1 and White has 26 exd6+ ~f8 27 l:tgl?
compensation even for the double This move is a bad time-trouble
knight deficit. mistake. 27 l:lxe6 was necessary. Then
21 l:.f3 after 27 .. J1e8 (27 ... lDf6?? 28 d7!)
After 21 exf6+ lDexf6 22 'ifg6 cwt>f8 28 l:lxe8+ (28 l:le7? l:lxe7 29 dxe7+
Black wins. ~f7) 28 ... ~xe8 29 ~gl 'ii'xM
21. .. c4?! 30 'it'xg7 Black has just a few pawn!
Knaak's move 21...~xb2 wins. After left and his king has no shelter. All 0
22 l:ldl (22 g4 l:lxa3) 22 ... lDf8 it is all which makes it very hard to win witl
over. the extra knight.
230
28 ....:f5 29 ~h5lLlf6 30 'iixh6+ ~f7

31 !iii>gl
27 ... g5!
Black saves the g-pawn which is a
very important achievement for him.
His queen stands well on the long 31 ....:g8 0-1
diagonal.
28 ngg3 Had Capablanca actually observed
Defending the vulnerable f3-rook in my position collapsing in ruins like
advance. But White's king remains this, he would surely have been
precariously placed. For example, after embarrassed at his negative influence
28 fxg5+ l::lf5 29 l:tgg3 lLle5 wins. on me.
Capablanca played a very famous G.Kasparov - J.Timmao

game where he froze his opponent's
bishop on g3 with his g5 and e5
pawns.
W.Winter - J.Capablanca
No doubt you have already noticed

that while Winter had a frozen bishop,
Timman has a knight. It's so annoying
that this could happen. Go through the
game and you can see why I say this!
231
W.Winter - J.Capablanea
Hastings Victory Congress 1919
1 e4 e5 2 tOf3 tOe6 3 tOe3 tOf6

4 R.b5 R.b4 5 0-0 0-0 6 ~xe6 dxe6
7 d3 .i.d6 8 .i.g5 b6 9 .i.b4 e5
23 ... e4! 24 axb5 exb3 25 exb3 lbb5

26 l:ta4 l:hb3
With such a frozen kingside White is
hopelessly lost.
27 d4 l:tb5 28 l:te4 -l:tb4 29 l:he6
lIxd40-1
10 tOd5? g5 11tOxf6+ 'iixf6 12 R.g3
.i.g4 13 h3 .i.xf3 14 'ifxf3 'it'xf3
15 gxf3 G.Kasparov - J.Timmao
Wereldhaven Festival, Rotterdam 1999
1 e4 e5 2 tOf3 tOe6 3 .i.e4 lLlf6
4 d3 R.e5 5 e3 d6 6 R.b3 0-0 7 ~g5
White's bishop on g3 is dead.
.i.e6 8 tObd2 a6 9 b3 .i.a7 10 .i.b4
15 ... f6 16 ct>g2 a5 17 a4 ~n 18 l:[hl
~b8
~e6 19 b4 l:tfb8 20 bxg5 bxg5 21 b3
11 g4
21. .. e6
When I was really young I also
Having centralised his king,
gained space like this against Petrosian.
Capablanca now opens the queenside.
11 ... tOe7 12 .i.xf6 gxf6 13 tOh4 tOg6
He can afford a lot of things on that
144:)g2
side as White is virtually a piece down.
22 l:ta2 b5 23 l:thal See diagram on page 231.
232
In the later part of the analysis you Now we return to my game:

can see why I thought Tirnman had a
bishop on g6. But here I had some 14 ... c6 IS 'iko dS 16 ttJn as 17 ttJg3
chessic doubts when a Steinitz game ~cS 18 a41Le7 19 ~a2 1:1a6 20 tLlhS
suddenly occurred to me.
A.Schwarz - W.Steinitz
Vienna 1873
20 ... l:b6
Tirnman's rook play is interesting.
21 'iVe2 'ifd6 22 0-0 l:1d8 23 1:1fdl d4
24 1:1d2 'ikcs
Jan should have played 24 ... dx.c3
16... ltJxg3
25 bxc3 'ita3 26 l:lad 1 "xc3 27 tLle3.
This confused me. Steinitz captures
White has some play for the pawn but
the bishop which seems to imply that a
B lack should be better.
bishop like this may not be so bad after
2S 1:1c2 'ifd6 26 ltJel 1:1g8 27 tLlo
all. Steinitz went on to win the game.
'ikd7 28 'it'hl cS 29 1Lxe6 fxe6 30 tLld2
On the other hand Capablanca won by
1:1a6
saddling his opponent with this bishop.
I was aware of the contradictory
messages these champions were
sending but thought, as the Cuban was
the later champion, he must have
played better than Steinitz, therefore it
was his principle I followed.
17 lhh8 ttJxe2+ 18 ~xe2 lbh8
19 dxcS dxcS 20 l1g1 lIg8 21 'it'c2
ltJd8 22 'ii'dS 'ike7 23 lIdl c6 24 'ifb3
bS 2S c4?
Giving up the d4-square was a huge 31 ttJc4
mistake of course. Somehow I have gained a small edge.
2S ... ltJe6 26 'it'bl ltJd4 27 'ii'e3 1:1h8 31. ..1Ld8 32 ncc1 l:[f8 33 0 1Lc7
28 'ii'd2 l:1h2 29 ~n b4 30 'ife3 'ii'f6 34 J::ta3 1:1aa8 3Sl:1b3
31 l:[d3 g4 32 bg4lIxfl 0-1 It is my turn to use the rook the way
White resigned in this lost position. that Tirnman did.
233
35 ...1ha4 36 l:.xb7 'ir'c6 37 ':b3 f5 cranes. Their shapes were somewhat

38 lOd2 l:I.f7 39 c4 a4 40 l:b5?! similar and I mixed them up. I was so
[40 lta3] 40 ... il.a5 41 lOft happy that I could copy Capablanca's
freeze technique. He is not solely
responsible for my loss here but shares
the blame with the heavy metal
designer. Maybe chess events should be
separated as in tennis indoor and
outdoor - so this game would not count
against my indoor record.
47 'iifJ?
I was so stunned that I made a losing
move, but Timman stood better
41...il.b4 anyway.
Timman's bishop is very much better 47 ... lOxh5 48 ..-xh5 'iif2 49 lOe2
than Winter's as it traps the b5 rook. 'iifJ+ 50 'ir'xfJ
42 lOfg3 fxg4 43 fxg4 l:afS
Black takes over the f-file.
44 ltft 'ilie8 45 lbf7 'iixf7 46 g5?
50 ...l:I.xfJ
That's it.
51 ~g2 l:xd3 52 lOg3 ct>g7 53 lIb6
46 ... lOf4 il.e1 54 lOft ~f7 55 lOh2 ':d2+
This came as a huge shock. I thought 56 cjw>h 1 :te2 57 lOg4 l:I.xe4 58 lOf6 l:I.e2
this piece was a bishop and now 59 lOxh7 0-1
suddenly it moves like a knight! The
piece on g6 has not moved for so long, Alekhine praised Capablanca's talent
and in my mind it remained like a so highly after the Cuban died and
slumbering bishop on g6. How could Kramnik said Leko was a tougher
this happen? This game was an outdoor opponent in the final than me. I think
exhibition in which pieces and pawns his true thoughts lie in the fact that he
were large steel containers moved by never gave me a chance of a rematch.
234
Emanuel Lasker the 2 nd
The second world champion beat Lasker had to win the last game to save
Wilhelm Steinitz 10-5 with 4 draws in the match. In the same year he beat
Philadelphia 1894. It was 2-2 with two Janowski again. This event was even
draws after the sixth game, then Lasker more convincing as he dropped only 3
raced away with five consecutive wins. draws out of 11 games. In the final of
Two years later Lasker started with four the 1914 St Petersburg tournament
consecutive wins. In the first eleven he scored 7 out of 8 and won the
games the ageing Steinitz made only tournament ahead of both Capablanca
four draws. Then Steinitz won two and Alekhine.
games in a row. In the last four games
During World War I Lasker lost his
Emanuel scored three more wins,
wealth, so consequently he had to put
winning the match 10-2 with 5 draws.
his title at stake in 1921. It was here
Lasker held the title for the longest that he lost to Capablanca in Cuba.
period - 27 years in all. However he
However he still scored some fine
was the champion who played the least.
tournament results, including a victory
After the second Steinitz match at New York 1924 ahead of Capablanca
Lasker played very little. He firstly and Alekhine. In 1935 he took third
defended his title in 1907, demolishing place in an extremely strong event in
Marshall 111h-3 1h, winning 8 games Moscow - at the age of 67.
and drawing 7 with no losses. One year
later he beat the ageing Tarrasch The chess he played is quite different
IOlh-5 1n, He won 8 games, lost 3 and from the style of the late twentieth
drew 5. century. So his effect on me is less than
say Smyslov's. Nevertheless, he still
In the famous St. Petersburg
tournament Lasker and Rubinstein both played games that influenced me.
scored 141h out of 18 games, although
He won many games with the
Rubinstein beat him in their individual
Alekhine defence pawn structure
encounter. Rubinstein also showed his
b4-c5-d4 against Black's b7-d5-e6, but
class in other tournaments. Sadly a
those we have discussed in the Smyslov
match between the two never took
chapter.
place. In the same year he destroyed
Janowski 8-2, winning 7 games and In the following examples we see
losing one with two draws. how Lasker gained space on the
In 1910 Lasker again defended his kings ide with g4 and then backed up
title with a 5-5 score against Schlechter. his attack with the aid of his knights.
235
Emanuel Lasker the 2nd
I found Lasker's pawn and knight setup rather attractive so I gave it a try.
E.Lasker - W.Steinitz G.Kasparov - R.Hiibner
E.Lasker - W.Steinitz
Hastings 1895
1 e4 e5 2 It:)f3 It:)c6 3 i..b5 a6

4 i..a4 d6 5 0-0 It:)ge7 6 c3 i.d7 7 d4
It:)g6
Though this setup is a bit passive it is
steady and was still being played more
than a century later. Timman won a
crucial game with it against Motylev at Masters still play moves like this in
the European Team Championship in the King's Indian. I also played liJg8 a
2005 where Holland went on to win the few times in a King's Indian position
event! (this game is similar). I won with this
8 l:tel i.e7 9 liJbd2 0-0 10 liJn 'ir'e8 move against Korchnoi in Barcelona
1989, and Khali&nan and Gelfand were
11 i.c2 ~h8 12 liJg3
my victims in the Paris Immopar rapid
White is a slightly better.
in 1991. The position seems to be like a
12 ... i.g4 13 d5 liJb8 14 h3 i..c8
Fischer Random chess position. Only
15 liJf5 i..d8 16 g4 liJe7 17 liJg3
three of the eight Black pieces are on
See diagram above. their starting squares of a conventional
chess game.
White's g4-pawn provides a nice 18 Wg2 It:)d7 19 i.e3 liJb6?
space advantage on the kings ide and Players no longer develop their
the white knights provide proper queen's knight in this way.
support. 20 b3 i..d7 21 c4lt:)c8 22 'ifd2lt:)ce7
17... liJg8 23 c5!
236
A classical manoeuvre. White cracks confident because of the plus I had on

Black from both sides of the board. the kings ide.
a very effective method when it 13 ... d4 14 ttJf5 ttJc5 15 ltgl ltfi
comes to exploiting a space advantage. 16 cxd4 ~xf5 17 gxf5 ttJxd4 18 ttJxd4
23 ... g6 24 'iic3 f5 'il'xd4 19 ~n ~h8 20 l:Ibl
A desperate attempt to get some play.
Waiting passively was also hopeless.
20 ... l:.d7
I was already gaining the impression
25 ttJxe5! that my kings ide play was possibly no
This takes Black apart; the rest is not stronger than Hubner's initiative on the
very interesting. other side of the board.
25 ... dxe5 26 'iixe5+ ttJf6 27 ~d4 21 l:Ig3 as 22 ~e3 iVd6 23 l:Idl ttJa6
fxg4 28 hxg4 ~xg4 29 'iig5 'iid7 24 ~b3 ~d4 25 ltg4 ttJc5 26 ~c4 '6'f8
30 ~xf6+ ~g8 31 ~dl ~h3+ 32 ~gl 27 l:Ih4
ttJxd5 33 ~xd8 ttJf4 34 ~f6 'iid2 Putting the rook on the edge doesn't
35 l:Ie2 ttJxe2+ 36 ~xe2 1i'd7 37 ndl achieve enough to gain an advantage.
'iif7 38 ~c4 ~e6 39 e5 ~xc4 40 ttJf5 27 ...'iie8 28 iVg4 .i.xe3 29 fIe3 h6
1-0 30 ~e2 c6 31 a3 b5 32 .i.a2 a433l:.g1
ttJb3
G.Kasparov - R.Hiibner At the end of his plan Hubner nicely
Game 4, Cologne TV blitz 1992 cuts off the dangerous bishop.
34 ~xb3 axb3 35 ':cl l:.ad8 36 l:Ic3
1 e4 e5 2 ~c4 ttJc6 3 d3 ~c5 4 ttJf3 'iig8 37 'iig6 ltd6 38 J:tg4 l:t8d7
d6 5 c3 ttJf6 6 ~b3 0-0 7 h3 ~e6 39ltgl 'iid8 40 nxb3
8 ttJbd2 a6 9 'fi'e2 ~a7 10 g4
I intentionally postponed castling.
10 ... ttJd7 11 .i.c2 d5 12 ttJn f6
13 ttJg3
Hubner's position is stronger on the

queenside than Steinitz's, but I was still
237
40 ... cS! so nice I really want to share them.

This was a blitz game, so I had little Maybe I was enjoying calculating
time to think. Nevertheless I did them. 49 %1f7 ltxe3+! (49 ... 'fIg8
remember that Lasker managed to 50 Acc7 l:tbb3) 50 ~f2 (50 ~xe3?
attack Steinitz's king. For a while I was 'ii'a3+ 51 ~d2 'ifb2+ and Black
hoping to do the same, but [Tom now on delivers a checkmate in 8 moves.)
I had to divide my attention between 50 ... l:tb2+ 51 <ifi'fl :0+ 52 ~gl :g3+
attacking and taking care of my own 53 'it'xg3 'it'xf7 54 l:tc8+ <if;h7 and it
king. looks like Black is winning.
41 l:txbS l:.xd3 42 <it>f3
After 42 ltxc5 Axe3+? there is no
perpetual, but 42 ... ltd2 transposes back
to the game.
42 ... ltd2 43 ltxcS Axb2 44 l:tgc1
Ab8 45 84
Bringing back the queen with
45 'fIg2 was not effective. After
45 ... Ad2 46 l:.c8 Axc8 47 ':xc8 ,*xc8
48 'iixd2 'fIa6 49 'fIc 1 "'a4 the queen is
tied to the defence of the pawn and the
idea of a perpetual. White probably 55 l:tc7!! ltbl+ 56 ltrh2 'fIa2+
can't make any progress. 57 1fg2 'iixg2+ (57 ... l:tb2? 58 Axg7+!)
4S ......f8 46 as lta7 47 a6 l:ba8! 58 <itxg2 ltb4 Black can press on, but
Hubner's feel for chess doesn't let probably White holds.
him down. He patiently takes back the
pawn, or should I say tries to take back
- as you will see. Recapturing at once
with 47 ... l:txa6? was clearly a worse
option. 48 ltc7?! (White is a pawn up
with some winning chances after
4811xe5!)
48 l:tc6 0-1
In playing my move I overstepped
the time limit! The position is equal as
Black could take the a6-pawn.
Luckily this Lasker-affected game

was only blitz, so it was not important.
48 .. .11a3!! It is rather strange to However the next one hit me at a
analyse a blitz game, but the lines are crucial moment.
238
In the first two examples the second world champion allowed his opponents
to have a strong rook on the seventh and yet he still won. Recalling his games
I also allowed this - but regrettably with a different result.
Heydebreck - E.Lasker D.Janowski - E.Lasker
G.Kasparov - A.Karpov V.Topalov - G.Kasparov
Heydebreck - E.Lasker 1 l:ld7

Berlin 1889 This time the rook on the seventh is
not particularly dangerous. However it
looks no fun for Black either.
See diagram above.
l. ..l'.tbl!
What a nice riposte!
2 l:lddl
Losing without much resistance. But
taking the queen with 2 lhb7 leads to a
beautiful win.
239
2 .. .ltJf2+!! 3 ltJxf2 l:txgl +! 4 ~xgl 19 .....a6?!

%:tel mate. Staying closer to the centre with
2...:txdl 3 %hdl ltJxc3 0-1 19 ... 'iVc6 seems better. Then 20 ltJxe7
White resigned. .i.xe7 21 1r'h6 .i.e6 22 f5 gxf5 23 'iVe3
l:ta6! and Black keeps his position
together.
D.Janowski - E.Lasker 20 'ir'h6 .i.e6 21 ltJxf6+?!
Cambridge Springs 1904 Capturing a pawn but allowing
Black back into the game. Better
1 e4 e5 2 ltJf3 ltJc6 3 ltJc3 ltJf6 was 21 ltJxe7! .txe7 22 d5 .i.f8
4 .tb5 .tc5 5 ltJxe5 ltJxe5 6 d4 .td6 23 'iltb4 .i.xc5+ 24lt>h 1 and Black is in
7 f4 ltJg6 8 e5 c6 9 .tc4 .tc7 10 exf6 trouble.
'ii'xf6 11 0-0 d5 21 ...rvt;f7 22 ltJe4 ltJf5 23 'iVh3 .te7
24.i.c3
12 iLxd5
An interesting piece sacrifice. 24 ... .td5
12 ... cxd5 13 ltJxd5 'ii'd6 14 'iVe2+ After 24 ... ltJxd4 25 'ith6 l:tad8
ltJe7 15 l:te1 .td8 16 c4 f6 26 :tad 1 ltJe2+ 27 ~hl ltJxc3 28 IIxd8
Black wants to play 17 ... ~f7. :txd8 the position is equal.
17 .td2 as 18 'iWh5+ g6 19 c5 25 g4 ltJh4 26 ltJd6+ ~fiI
240
27 l:txe7 34 liJe8+ ~h6 35 h4 g5! [35 ... hxg4

Janowski plays in optimistic fashion. 36 f8='tW+ l:txf8 37 .i.d2+ ~h5
After 27 <itf2 the position is unclear. 3811h7 mate] 36 hxg5+ ~xg5 37 l:te5+
27 .••l L'lf3+ 28 'iVxf3! <ith4 38 b3 'ii'xa2 39 .i.e I + ~h3
White sacrifices his queen based on [39 ... ~xg4 40 liJf6+] 40 lle3+ ~xg4
the strength of a rook on the seventh. 41 tDf6+ ~f5 42 liJxd5 and at the end
28 ... ~xf3 of this long tactical line the position
remains very complicated.) 30 11f7+
~g8 31 l1ee7 .i.d5 32 ':'g7+ ~f8
29 l:tti+?
33 b3! White has compensation for
Janowski probably misses Lasker's
the queen and the game should
32 nd move. Bringing his other rook
probably end in a draw. It looks like
across with 29 l:tae I! leads to some
Black can't do anything useful in the
fascinating tactics. 29 ... 'ir'c6 (29 ... .i.d5
ensuing ending. (33 f5 ir'a4!! wins.
30 f5 h5! [30 ... g5? 31 f6 'iVc6 32 f7 33 l:tgf7+ .i.xf7 34 l:txf7+ ~g8
'ifa4 32 ... ~g7 33 l:Ue6!! ~xe6 35 l:tg7+ ~xg7 36 d5+ ~g8 37 dxc6
34 d5+ ~f8 35 l:txe6 wins - 33 l:tc7 bxc6 and Black should be a bit better.)
~g7 34 f8='if+ ~xf8 35 l:tfl + ~g8 33 ... h5 (33 ... a4 34 ':'gf7+ ~xf7
36 lLlf5 White wins.] 31 f6 [31 fxg6? 35 l:txf7+ ~g8 3611g7+ ~xg7 37 d5+
:th6 32 l:t7e5 l:txg6] 31...'iVc6 32 f7 ~g8 38 dxc6 bxc6 39 b4 White is no
'ii'a4 33 ':'c7 ~g7 longer worse.) 34 f5 gxf5 (34 ... hxg4
241
35 f6 lth3 36 lien+ iLxn 37 lhn+ 10 iLg3 'iVe7 11 a4 as 12 h4 ltg8

Wg8 38 l:tg7+ <ii:>f8 is equal, but not 13 hxg5 hxg5 14 ~3lL\a6
38 ... Wh8?? 39 l1Jn and checkmate)
35 l:tc7 fxg4 36 lhc6 bxc6 37 l:tb7 and
the position is still unclear.
29 ... ~g8 30 d5 iLxd5 31 ltg7+ 'it;f8
32 %:tel
15%:tb1!
I made sure Karpov would not castle
long.
15 ... ~f8
He got the message.
32 •. :~Vc6! 16 ~d1 iLe6 17 %:th2! ~g7 18 e5
The only move to Win. It stops bxe5 19 iLb5 lOb8 20 dxc5 d5 21 ~e5
33 l:tc7 'it;f8 22 l:th6 lOe8 23 'ii'h5 f6 24 %:th7
33 b4 l1d8 34 iLd4 l:txd6 35 cxd6 lOg7 25 'ii'n ~f7 26 'ii'h5+ ~f8
iLh1 0-1
G.Kasparo\' - A.Karpov
1 d4 l1Jf6 2 c4 e6 3 lL\n b6 4 lL\c3

iLb4 5 iLg5 iLb7 6 e3 h6 7 iLh4
Lasker's spirit was alive in this game.
One game earlier I had a 4-1 lead
against Karpov and wanted to finish
him off once and for all. So even 27 'ii'n
though he had won the previous game, With the help of imaginative play I
I still wanted to clinch the match have forced him to defend doggedly.
with a decisive result. What does this My problem was that it took too long to
have to do with Lasker? Well, he was calculate and I was already short of
the champion who won the most time.
one-sided world title matches, doing 27 .. !:J;;f7 28 lth6 lL\e8 29 e4 g4
comprehensive demolition jobs on both 30 'ii'f4 iLxb5 31 lbb5 lL\d7 32 il..xe7
Janowski and Marshall. lOxe5 33 'it'e3 lL\xe4 34 lL\xe4 dxe4
7... iLxe3+ 8 bxe3 d6 9 lL\d2 g5 35 iLxa5 f5 36 iLb4 'ii'd7
242
43 ... l:tal 44 .i.c3! ltcl 45 ~e5 l:lc2+

46 ~el l:lxc4 (46 ... l%dd2 47 l:th6+!!
~g5 48 l:th7 :e2+ 49 ~dl tiJe8
50 l:tb8 Black has no more than a
perpetual.) 47 :b6 e3 48 fxe3 l:te4
49 l%d6 (49 i.xg7? l:txe3+ 50 ~f2
l:ldd3!! wins.) 49 .. Jbe3+ 50 ~d2
l:txe5 51 l:txd8 l:txa5 White holds as I
gave in my book of the match.
44 a6?
37"d4 After 44 .i.c5 tiJh5 45 g3 l:txc4
Karpov has done well to stay in the 46 ~e3 l:ta4 47 l:tb6 White is okay.
game, but I still had the preferable
position.
37 ...l1a7 38 l:th7+
38 i.c5 was strong.
38 ... tiJg7 39 as?
Better was 39 i.c5!
39 ... ~g6?
After 39 .....xb5! 40 "xa7+ ~g6
41 ':h4 l:td8 42 'ii'e3 tiJh5 White is in
trouble.
40 "Id7 l:lId7 41 l:th4 44 ...l1c2+!
This was my sealed move. Karpov gets his rook to the second
rank but I was not worried as Lasker
coped with it even when his opponent's
minor pieces were backing it up. If
44 .. .lhl 45 .i.d2 l:ta2 46 l%h6+ <itf7
41. .. l:.gd8!
Karpov goes after my king.
42 c4 l:.dl+ 43 ~e2l:tc1!?
Karpov controls himself so well
when his opponent has passed pawns. 4 7 ~e3!! (White's king creates
Perhaps he would play this move sufficient counterplay. After 47 l:tb7+
anyway but the must-win situation ~g8 48 l%g6 :axd2+ 49 ~e3 l:t2d7
helps to prompt a move like this. After 50 c5 the subtle intennediate move
243
50 ... ~h7!! wins. [On the other hand However, as played, he is really hurting
50 ... e5 is met by 51 ~d6.] 51 ~f6 g3 the king.
52 fxg3 ttd3+ 53 ~f4 e5+ 54 ~xe5 54 ':h3 f4 55 l:[b4 <ifi>f5 56 Ab5+ e5
:e8+ 55 ~f4 e3 56 ~fn ~e4+ 57 ~g5 57 :a5 ':dl
:g4+ 58 Wf6 l:tg6+ Black wins.) Makarychev says this squanders the
47 .. J:1axd2 (47 ... e5 48 :d5) 48 l:tb7+ win. I was right - it makes Black find
~g8 (48 ... l:t2d7 49 c5) 49 ~f4 l:t2d7 more good moves.
(49 .. J~txf1+ 50 ~g5!) 50 c5 e5+ 58 a7?
(50 ...11n 51 c6) 51 ~xe5 g3 52 fxg3 e3 I went down without putting up a
53 ~xd7 l::txd7 54 l::thl and White fight. Pushing 58 c7 would have given
holds. me some practical chances, but in
45 ~el 11a2 46 nb6 l:td3! 47 c5 reality it was losing as well. After
After 47 i.c5 g3! (47 ... f4 48 :bl) 58 ... e3
48 fxg3 l:txg3 49 :d6 Agxg2 50 a7 e3
51 :hl ttJh5 52 l:td8 ttJf4 53 a8='iV
Axa8 54 l:ha8 l:te2+! 55 <ifi>dl l:td2+
56 <it>cl ttJd3+ 57 ~bl ttJxc5 Black
wins.
47 ...11al+ 48 ~e2 .:1a2+ 49 ~el g3
49 ... ~g5! was even stronger.
50 fxg3 l::txg3 51 ~f1
59l:th2!! 1:.ccl! is the simplest move.

(If 59 .. .l::tx.h2 60 c8='iV+ ttJe6 61 'fIVc3
':f1+! [if 61 ... ttJd4? 62 'fIVc8+ ~e4
63 'ifb7+ White has a perpetual check
at his disposal.] 62 ~gl itJd4 63 'fIVc8+
<ifi>e4 64 'ifb7+ [64 ':xe5+ ~xe5]
64 ... ~d3 65 l:.a3+ ~e2 and Black
51 ...Agxg2 wins.) 60 l::te2 ttJe6 61 a7 itJxc7 wins.
At this stage I was not certain
whether I should have allowed both
rooks to go to the second rank.
However, I was still relaxed. Lasker did
not mind things like this either.
52 i.el
As a junior Karpov liked to double
on the seventh.
52 .. J:gc2 53 c6 1:.al
He gives up the idea of doubling.
244
58 ... e3 Taking the b2-pawn was more

I had to resign as the rooks and the consistent.
knight deliver a checkmate. 16 .i.xd6 tOxg4
59 :10 tOb5 60 a8=. tOg3+ After 16 ... ttJd4 17 e5!? .xb2 18 exf6
61Ibg3l:[f2+ 62 ~gll:bel mate 'ii'xa 1+ 19 ~f2 'iVxc3 20 .xc3 :1xc3
21 fxg7 l1g8 22 ttJe4 Speelman's line is
V.Topalov - G.Kasparov winning for White.
Moscow Olympiad 1994 17 it.xg4 'iVxb2?
If 17 ... it.xg4! 18 ttJa4 ~5 19 hxg4
1 e4 e5 2 tOn d6 3 d4 exd4 4 tOxd4 .xa4 20 'ii'd5 .a5+ 21 c3 White has a
li:)f6 5 tOe3 a6 6 .i.e3 e6 7 g4 h6 small edge.
In those days the Perenyi variation
had not yet been exhaustively analysed.
8 f4 ttJe6
In Wijk aan Zee 1999 I played 8 ... e5
like this game. Timman forced a
perpetual after 9 ttJf5 h5 10 gx.b5 exf4
11 .i.xf4 ttJxh5 12 tOxd6+ .i.xd6
13 .i.xd6 ~4+ 14 ~d2 'iWg5+ 15 ~el
'iib4+ 16 ~d2 l,h-lh. I did not dare to
play on as I knew Steinitz liked to walk
into the centre with his king. But we 18 e5!?
have one more chapter to deal with This temporary rook sacrifice allows
that. White to maintain the initiative.
9 it.e2 e5 10 tOf5 g6 11 li:)g3 exf4 However 18 tOge2! was even better,
12 .i.xf4 .i.e6 13 :1n :1e8 14 h3 perhaps just winning for White after
18 ... 'ii'xal+ 19 ~f2 iVb2 20 :'bl.
18... tOxe5 19 l1bl 'ir'xe3
Bansch's move 19 ... li:)c4! underlines
Lasker's play in the previous examples
- and Alekhine's play in relation to
batteries. 20 lbb2 tOxd2 21 it.xe6 fxe6
22 :'xb7 ~xc3
14.. Jib6
Since Fischer's time we Najdorf
believers play this kind of move.
Bansch suggested 14 ... d5 as a standard
kind of response.
15 'iVd2 it.g7
245
23 ':ff7 (The doubled rooks on the

seventh can force no more than a
draw). 23 ... tOe4+ 24 <it>d1 tOxd6
25 l:tbe7+ and the game ends In a
perpetual.
20 'ihc3 ':xc3 21 .i.xe6 fxe6
After 21.. ..:xg3 22 ':xb7 (22 .i.xn+
~d7!) 22 ... fxe6 Black is living
dangerously, but I see no win for
White.
26 tOd4!
Moving the rook away from the
f-file with 26 l:tfe7+! was more precise.
Then 26 ... ~d8 27 tOd4 .i.g3+ 28 ~e2
and Black can't even sacrifice the
piece.
26 ...l:te3+
If 26 ... .i.g3+ 27 <it>e2! l:th2+ 28 ~f3
lH2+ (28 ... .i.h4 29 :lfc7 tOd2+ 30 <it>f4
22 l:txb7 .i.d8 31 :b8 wins.) 29 ~xg3 ':xn
From what I had learned from Lasker 30 ':b8+ litd7 31 l:txh8 and Black has
I was confident that there is no point only two pawns for the piece.
panicking just because one rook 27 ~n l:te4 28 l:tfe7+
reaches the seventh rank. Checkmate comes very soon.
22 ... tOc4 28 ... ~d8 29 tOc6+ 1-0
If 22 ... tOd7!? 23 l:ta7 .i.f6.
23 .i.b4 l:te3+ The last motif is not really a motif at
After 23 ... l:txg3 24 ':xg7 a5 all, more a variation. Maybe it is best to
(24 ... tOe3 25 l:te7+ (25 l:tff7? tOf5) call it a defence. So far I have not paid
25 ... ~d8 26 l:ff7 wins.) 25 .i.c5 l:tg5 attention to where the champions were
Black still resists. born. In this case strangely enough it
24 tOe2 .i.e5 25 l:tff7 adds to the interest. I was born in Baku,
See diagram on page 239. no other champion was born there. The
closest was Petrosian who was born in
25 ...':xh3? nearby Tbilisi. Emanuel Lasker was
I became too casual because of the born a long way from Baku in
Lasker examples. After 25 ... .i.d6 Berlinchen, which of course sounds so
White can keep up the pressure with similar to the German capital. My last
26 .i.c5!! Then 26 ... ':e4 (26 ... .i.xc5? world title match took place in London,
27 Afc7) 27 ':g7 .i.e5 28 l:txg6 ~d8 yet thinking of the German capital
29 l:ta7 and White has an edge. generates rather painful memories.
246
Kramnik did not play the Russian Now here are positions I had
defence (known in the West as the against Kramnik in my ill-fated
Petroff) in honour of his homeland. World Championship match in
He kept playing the Berlin defence London 2000.
(or wall). Lasker could not have
anticipated the existence of the G.Kasparov - V.Kramnik
Berlin wall. More importantly he
did well against the Berlin defence
and exceptionally well against the
exchange Ruy Lopez where Black
has a rather similar pawn structure a
half a pawn down. Here is the
position from which Lasker went on
to win.
E.Lasker - HerzILewittIKeidanski G.Kasparov - V.Kramnik
E.Lasker - HerzILewittlKeid'anski 8 'ii'xd8+ 'it>xd8 9 l:ldl + 'it>e8

Consultation game, Berlin 1896 10 ci:)c3 h6 11 h3
1 e4 e5 2 ci:)o ci:)c6 3 iLb5 ci:)f6 4 0-0

ci:)xe4 5 d4 ci:)d6 6 iLxc6 dxc6 7 dxe5
ci:)f5
The first game with this position in
the database is from Leipzig 1879. Bier
was White against Flechsig. But the
second game was indeed played in
Berlin in 1880.
247
11 ...~e7 14 ... l:tg8 15l:td2 ~c8 16l:tadl l::txd2
The next time this position occurred 17liJxd2
was in 1990 in the game Yudasin- He prepares to advance his pawns.
Rogers, Manila 1990. 17 •.. g5 18 g4 liJg7
12 ltJe2 Black's knight has moved five times
Lasker plays the move which is still to get to g7 from g8. Chess is weird
popular. sometimes, but it didn't confuse
12... ~d7 Lasker!
Later I played a different move in this 19 ltJe4 liJe6
position with Black and lost to Judit
Polgar. Here are the moves. 12 ... ltJh4
13 ltJxh4 ~xh4 14 ~e3 .i.f5 15 ltJd4
.i.h7 16 g4 ~e7 17 Citi>g2 h5 18 ltJf5
.i.f8 19 Citi>f3 .i.g6 20 l::td2 hxg4+
21 hxg4 l:th3+ 22 Citi>g2 l::th7 23 Citi>g3 f6
24 ~f4 ~xf5 25 gxf5 fxe5 26l::te 1 ~d6
27 ~xe5 Citi>d7 28 c4 c5 29 .i.xd6 cxd6
30 l::te6 l::tah8 31 l::texd6+ ct>c8 32 l%2d5
l::th3+ 33 Citi>g2 l::th2+ 34 Citi>f3 l:t2h3+
35 Citi>e4 b6 36 l:tc6+ ~b8 20.i.cl!
Showing intelligent flexibility. The
bishop is no longer useful on the a I-h8
diagonal.
20 .. .llg6
Black plays confusing moves just
like Kramnik did later against me.
21 .i.e3 c5 22ltJ2g3 b6 23 liJh5liJg7
24 ltJhf6+ ~f8 25 ltJh 7+ '11;>g8
26 liJef6+ Citi>h8
37 l::td7 and again I did not mind my

opponent invading on the seventh.
37 ... l:th2 38 ~e3 :f8 39 l::tcc7 I even
allow the second rook to join the first
on the seventh. 39 ... l::txf5 40 l::tb7+ Citi>c8
41 l::tdc7+ ~d8 42 l::txg7 ~c8 and I was
so upset I just rushed off home. 1-0
J.Polgar-Kasparov, Moscow rapid,
2002.
13 b3 l::td8 14 .i.b2 27ltJf8
With some remarkable jumping
See diagram on page 247. around here and there Emanuel obtains
248
an advantage. This forces exchanges, S 'iYxdS+ <i!fxdS 9 tDc3 i..d7 10 b3

which allow an invasion. h6 11 i..b2 <i!fcs 12 h3 b6
27 ... i..xfS 2S :tdS
White invades.
2S ...tDe6 29 lbcs 'iitg7 30 tDdS hS
31 gxhS :th6
13 nad1
At the same time I wanted to copy
and improve on Lasker's play.
My finesse was to use the ai-rook on
32 tDf6! the d-file. In exchange, I allowed
Flexible thinking again. He returns to Kramnik's king to go to the queenside.
the queenside and wins. This is an idea Romanishin introduced.
32 ... l::thS 33 <i!fg2 <i!fh6 34 h4 i..g7 Incidentally, the grandmaster from
35 hxg5+ tDxg5 36 :txhS+ i..xhS L vov defeated me a number of times,
although I beat him too.
37 tDe4 1-0
13 ... ttJe7!?
This is a manoeuvre by Zoltan
Here are my first two games against
Almasi. The Hungarian grandmaster
the Berlin. In the first game I tried to
played it just a few weeks before our
copy Lasker's play. It contributed a lot
match so I had no time to analyse it.
to the loss of my title. I had a reputation 14 tDe2
for being fonnidable in the opening but
against the Berlin the sharpness of my See diagram on page 247
sword was lost for a good while.
All goes according to the Lasker
plan.
G.Kasparov - V.Kramnik
14 ... tDg6
London 2000
1 e4 eS 2 ttJO tDc6 3 .i.bS tDf6 4 0-0

ttJxe4 S d4 ttJd6 6 .i.xc6 dxc6 7 dxeS
tDfS
This line carne as a surprise. I had
had limited experience with it. I
switched to 1 e4 as a main weapon only
in the early 1990s.
249
IS ltJel G.Kasparov - V.Kramnik

Lasker moved his knight to d2, so I Game 3, World Championship,
also paved the way for my kingside London 2000
pawns.
1 e4 e5 2 ltJo ltJc6 3 ~b5ltJf6 4 0-0
IS ... hS! 16 ltJd3 cS
ltJxe4 S d4 ltJd6 6 ~xc6 dxc6 7 dxeS
Not 16 ... ~b7? 17 ltJc5+!.
ltJf5 8 'ii'xd8+ ~xd8 9 ltJc3 ~d7
17 c4
In the third Berlin game of the match
17 l:td2!? was the right way
Kramnik did not repeat his previous
according to Lasker.
play and deviated here with 9 ... h6.
17 ... aS 18 a4
Then 10 l::tdl+ (10 h3 was my choice in
I wanted to keep the a-file closed.
the fourth and last Berlin of the match.
18 ... h4
I also employed it in our seventh and
This is confusing. Kramnik's last
final Berlin as well. 10 ... <jre8 He drew
three moves were c5, a5 and h4. What
with ease in the last game of the match
is he playing for?
with 10 ... ~d7. Still he changed for this
19 ltJc3 ~e6! 20 ltJdS <jrb7 21 ltJe3
one. I show you an abbreviated version
of Dokhoian's analysis. 11 b3 ~e8
12 ~b2 .tt.d8 13 %:tadl ltJe7 14 l1fel
ltJg6 15 ltJe4 ltJf4 16 e6 ltJxe6
21. ..l::thS
By means of unorthodox play Black
has equalised.
22 ~c3 17 ltJd4?! (17 ~e5!! This magical
If 22 f4 ltJe7. move would have given Kramnik a
22 .. .l::te8 23 l:td2 ~c8 severe headache. 17 ... %:tc8 [17 ... ~c8
Going back with the king. 18 ltJf6+ ~e7 19 ltJh4 g6 20 ltJd7!]
24 f4 ltJe7 18 ltJh4! White follows up with f4; and
Now he even goes back with the has a clear advantage.) 17 ... c5?!
knight and he has a reasonable and safe (17 ... ~c8 18 ltJf6+ gxf6 19 ltJxe6
position. Everything goes against logic. l:txdl 20 lLlg7+ fitd7 21 lhdl+ and
2S ltJa ltJrs Ill-Ill White is better. But 17 ... l:tg8! should
250
have been played when Black can 4.Jxb3 29 l%xb7 4.Jcl 30 4.Jxcl nxc3
almost equalise.) 18 4.Jf5 ~c8 19 4.Jxh6 Ih-lh Kasparov-Kramnik Game 9,
lbd 1 20 :xd 1 l%h8 21 ~f5 f6) 18 4.Jf5 World Championship, London 2000.
l:th7 19 ~f6 lIc8 20 .i.xg7 (20 f4!)
20 ... ~xg7 21 ~xg7+ :xg7 22 ~f6+ In our first Berlin after the match
cl;e7 23 ~xd7 lId8 24 4.Je5 :xd 1 Vladimir played 9 ... ~e8, a move he did
25 ltxdl ~f4 (25 ... 4.Jd4!) 26 cl;hl! ttg5 not use against me earlier. I got very
27 4.Jg4 :d5 28 lie 1+! cl;f8 29 ~xh6 close to beating him, but he escaped.
l:td2 10 h3 ~e7 11 ~g5 ~xg5 12 4.Jxg5 h6
13 4.Jge4 b6 14 :fd 1 4.Je7 15 f4 4.Jg6
16 l:tfl h5 17 :ael ~f5 18 ~g3 4.Je7
19 ~xf5 ~xf5 20 ~f2 ~d4 (20 ... h4
21 lid I ~e7 22 lId3 with a slight
advantage) 21 ':cl ':d8 22 :fdl cl;e7
23 lLJe4 h4 24 b4 nh5? This is what
happened in my game against Kramnik
at Wijk aan Zee 2001, but better would
have been 24 ... ~f5!? Here I missed
30 ':e5! ':xf2? This is a mistake, but the opportunity to gain an almost
he was already in time trouble. 31 l%f5 winning advantage.
~g7 32 ~g4 ':xg2 33 l%xf4 l:1xc2
34':£2 ':c3 35 ~g2 b5 36 h4 c4 37 h5
cxb3 38 axb3 nc5 39 h6+ ~f8 40 ~f6
l%g5+ 41 ~hl Finally I brought down
the Berlin wall. It was a last round
game; I had to win to grab the first
place from Kramnik. It eased my
feelings about the Berlin Defence.
1-0 Kasparov-Kramnik, Astana 2001)
II ~e4 c5 12 c3 b6 13 l:tel .i.e6 14 g4 25 g4! We both overlooked this
Ih-lh Kasparov-Kramnik, Game 13, simple move. 25 ... nhh8 (25 ... hxg3+
London 2000) IO ... ~e8 II h3 a5 26 lLJxg3 l:hh3 27 l:txd4!) 26 f5 and
12 ~f4 ~e6 13 g4 lLJe7 14 ~d4 ~d5 Black is in big trouble.
15 ~ce2 ~c5!? (New) 16 lLJxe6 fxe6 10 b3
17 c4 lLJb6! 18 b3 a4 19 ~d2 ~f7 In 2001 I had already played 10 l:tdl
20 .i.c3 l:hd8 21 nxd8 ':xd8 22 c;i?g2 against Krarnnik, but I still could not
ltd3 23 ':c I g5 24 l:tc2 axb3 25 axb3 get it right. I 0 ... ~c8 II lLJg5 ~e8
lLJd7 26 ':a2 ~e7 27 l:ta7 ~c5 28 f3 12 ~ge4 b6 13 h3 ~b7 14 g4 lLJe7
251
15 ~f4 h5 16 f3 c5 17 ~f2 lDc6 doesn't mean the disappearance of

18 lDd5 lDd4 19 c3 lDe6 20 ~g3 ~c6 Black's difficulties.
21 l%d2 hxg4 22 hxg4 c4 23 <itg2 l%d8 23 exf6 lDc6 24 %:td3 lUS 25 l:te4
~c8
24 %:tad 1 .lta4 25 l:[e 1 .ltc6 26 l:ted 1
.lta4 27 :el .ltc6 lh-th Kasparov-
Kramnik, Zurich rapid 2001.
10 ... h6 11 .ltb2 ~c8 12 :lad 1 b6
13lDe2!?
See diagram on page 247
I still trust Lasker's plan, but use an

improved version and hold back h3. In
our last Berlin, when I finally stopped 26 f4?!
the slight embarrassment caused by this More testing was 26 h4!
defence, I developed the knight on e4. 26 ... gxf4 27 %:txf4? l%e8 28 il.c3?!
But even in my second game with the %:te2 29 l%f2 l:te4
line I followed Lasker. The initiative is gradually drifting
13 ... c5 14 c4 ~c6 15 lDf4 ~b7 away from me.
16lDd5lDe7 30 l:[h3 as! 31 l:[h5 a4 32 bxa4!?
16 ... %:te8 17 :ld3. lbc4 33 .ltd2 Iba4 34 lbh6 %:tg8?
Black is somewhat better after
17 l:tfel 1Ig8 18 lDf4! g5 19 lDh5
34 ... :xh6! 35 .ltxh6 c4 36 g4 c3.
:g6
35 l:[h7 lba2 36 l:txf7 lDe5 37 l:[g7
.J:f8 38 h3!?
After 38 h4 comes 34 ... lDd3!
20 lDf6
The knight IS Jumping around just
like it did in the Lasker game. 38 ... c4
20 ... .ltg7 21 :ld3! .ltxf3?! 22 :lxf3 Kramnik had little time left for the
.ltxf6 next few moves. Best was 38 ... lDd3!
My knight will not become a hero 39 n <ifi1d7! 40 l:[g8 :a 1+ 41 'it>h2 lDxf2
like Lasker's, but its disappearance 42 lhf8 rl;e7 and Black is safe.
252
39l%e7 47 h4?!
According to Kramnik 39 ~c3 Allowing an easy draw. Better was
would have given an edge. 4 7 ~h2! na6! which however also
39 ....!Dd3 40 f7 .!DIfl 41 l:te8+ ~d7 draws.
42 l%xfS ~e7 43 l:te8 ~xf7 44 l:txe7+
47 ...lta6! 48 .i.d4 l:ta4 49 .i.xe3
~e6 45 ~e3 .!Ddl 46 ~xb6 e3
.!Dxe3 50 Ibc3 nIh4 51 nf3 Ih-Ih
Probably, of all the past world

champions, Lasker was the least
'professional' in his approach to chess
- he had no need to be so focused -
maybe my own attitude represents the
modern approach of being highly
professional in all aspects.
253
Wilhelm Steinitz the 1st
Steinitz declared himself world 126 years earlier than me. In a way
champion after his 12 1/2-7 1/2 victory this is true, yet I was a good pupil
over Zukertort. Before Steinitz, who was taught to respect all world
Morphy was the best player, but they champions.
never met. Before Morphy, Anderssen
My junior trainers Oleg Privorotsky
was the world's best player. Steinitz
and Alexander Shakharov also showed
met Anderssen in a match in 1866 and
me Steinitz's games and I remembered
beat him 8-6 with no draws.
his ideas and employed them. So let me
Steinitz contributed a lot to the show you how his games affected me.
foundation of modem chess and also They may not bear such a strong
had many sacrificial games. resemblance as those in the previous
One might think his chess was too chapters, yet Steinitz's very strong
distant from mine, as he was born spiritual effect can still be traced back.
Here, Steinitz had a very nice W.Steinitz - P.Meitner

riposte when his opponent tried to
play on his first rank.
In the next two examples from

my career, my opponents played
on my back rank. I'm sure there
must have been an answer to all
this, but I failed to find it.
G.Kasparov - E.Magerramov A.Karpov - G.Kasparov
254
J.Lautier - G.Kasparov
Once Steinitz misled me because

I thought my opponent must have a
defence against any ploy on the
back rank. That is why I did not try
to exploit White's boxed in king
standing there all alone.
W.Steinitz - Ph.Meitner 18 'it>g2 ltJh5 19 l:lfl ltJc6

Vienna 1882 Black's knights control White's weak
spots most effectively.
1 e4 e5 2 f4 ~c5 3 ltJo d6 4 ~c4 20 b3 l:lhfH 21 iLe3 l:tbd8 22 'iVd2
~g4? ltJf4+ 23 'it>h 1 1;[(6 24 ]:tac1 ltJd4
Of course my opponents never made 25 'ii'a5 'ifh3 26 iLxf4 exf4 27 l:lfl
mistakes like this. ltJxo
5 !xe5 dxe5
28 e5
White retains the advantage.
6 ..txfi+! ~fH 7 ~b3 ltJc6 8 ltJc3 g6
28 ...l:lc6 29 e6 l1xd3
9 d3 'it>g7 10 ltJa4 .i.b4+ 11 c3 b5
12 cxb4 bxa4 13 ~xa4ltJxb4 14 ~b5? See diagram on p.254
14 0-0 was almost winning.
14 ... .i.xO 15 gxO l1b8 16 .i.c4 Black tries to take advantage of
-.h4+ White's unprotected rook on the back
Black has managed to get some rank.
compensation for the pawn. 30 e7! 'iVe6
17 ~flltJf6 Black seems able to hold the passed
Better was 17 .. :iVh3+! pawn.
255
Wilhelm Steinitz the lSI
31 ltell:hc4
21. .. aS
Magerramov sacri:0ces a pawn. He
was not able to hang on the pawn with
321be6!!
21...axb5 because then 22 ltxa8 l:txa8
White still takes the queen despite his 23 ~xf5 wins.
visibly very weak back rank. 22 'it'f3 lbc8 23 b3 l:tcs 24 bxc4
32 ... l:tc1+ 33 ~el!! dxc42S1Udl
A beautiful defensive move. I felt there was no need to waste time
33 ... ltJxel 34 ~eS+ defending the pawn. 25 lbb I was also
strong.
This leads to checkmate.
2S .. J:txbS 26 l:[d6 ir'e7 27 ir'c6
1-0 White is about to win.
27 ...1:tb2
G.Kasparov - E.Magerramov
Moscow 1976
1 e4 eS 2 ltJf3 ltJc6 3 ~bS a6 4 ~a4

ltJf6 S 0-0 ltJxe4 6 d4 bS 7 ~b3 dS
8 dxeS ~e6 9 c3 ~e7 10 ltJbd2 0-0
11 ~c2 fS 12 ltJb3 'ilfd7 13 ltJbd4
ltJxd4 14 ltJxd4 cS IS ltJxe6 'ii'xe6
16 f3 ltJgS 17 ~xgS
28 ~xfS?
This is not the main line. Spassky This is a dreadful mistake. White
played like this against Chekhov in should just divert the queen first with
USSR 1972, and I was following that 28 l:[d7! and then would win after the
game. decisive 28 .....e8 29 ~xf5! ':xf5
17 ... ~xgS 18 f41l.d8 30 l:txg7+.
28 ...l:[xfS 29 l:te6?
In the above-mentioned game Black
29 ltd7 was still good enough to stay
went back to e7. in the game. Then 29 ... 'iffB is met by
19 ~hl ~b6 20 a4 c4 21 axbS 30 'ii'e6+:n 31 l:tb7.
256
Wilhelm Steinitz the }st
29 .....a3!! 32 'ii'e6+ <i!if8 33 '6'c8+ ~g7

34 exf6+ lbf6 35 '6'g4+ ~h8
See diagram on page 254 I have run out of play.
0-1
I thought I had a riposte, just like the
very first world champion. A.Karpov - G.Kasparov
30 lIdl Game 17, World Championship,
My plan was to play on his back rank Lyon/New York 1990
as well by 30 l:r.e8+. But, after 30 ... ltfS
31 "xc4+ <iii>h8 1 d4 l"Llf6 2 c4 g6 3 l"Llc3 d5 4 cxd5
l"Llxd5 5 e4 l"Llxc3 6 bxc3 ~g7 7 i.e3
Kramnik started with 7 l"Llf3 in the
second game of our world title match in
London. I just gave him a pawn and
lost. Okay, he played well, but still.
7 ... c5 8 i.e3 '6'a5 9 '6'd2 i.g4 10 l:lbl
a6 11 lIxb7 ~xf3 12 gxf3 ttJc6 13 i.c4
0-0 14 0-0 cxd4 15 cxd4 i.xd4 16 i.d5
i.c3 17 '6'cl ttJd4 18 i.xd4 i.xd4
19ltxe7 lla7 20 lha7 ~xa7 21 f4 "'d8
32 'ili'a4, missed 32 ... ~c5 which 22 'ii'c3 i.b8 23 'ii'f3 ~4 24 e5 g5
defends two pieces with one move. 25 l:te 1 '6'xf4 26 'ifxf4 gxf4 27 e6 fre6
Then 33 "xa3 i.xa3 34 l:r.xfS+ ~xfS 28 ':xe6 <iii>g7 29 ~ha6 :f5 30 i.e4
35 g3 l:r.b5. l:r.e5 31 f3 l:r.e7 32 a4 l:la7 33 :b6 i.e5
30 ... l:txf4 34 nb4 ltd7 35 ~g2 :d2+ 36 ~h3 h5
Black can even afford to take this 37 l:[b5 ~f6 38 a5 :a2 39 l:tb6+ ~e7
pawn. 40 i.d5 1-0 Kramnik-Kasparov,
London 2000.
7 ... c5 8 'ii'd2 0-09 ttJO i.g4 10 ttJg5
cxd4 11 cxd4 l"Llc6 12 h3 i.d7 13 lib 1
':c8
311%f6
Neat but ineffective.
31. .. gxf6!
White not only has no mating attack,
he does not even have a perpetual. 14 l"Llo
257
Karpov did not take the pawn with

14l:txb7? I thought Kramnik would not
either. But Vladimir did not investigate
Karpov's play. 14 ... liJxd4 15 ~xd4
~xd4 16 'ir'xd4 l:tc1+ (l6 ... 'i!Va5+?
17 ~4) 17 ~dl
25 ~xd2!
One can easily can. miss the fact that
the bishop can take back the knight.
25 .. Jle8

And now Black wins the queen with
the lovely finesse 17 ... ltd 1+! 18 'ifi1xd 1 26l:te6!
~a4+. Oh no! White is able to gain
14 ... liJII5 15 ~d3 ~e6 16 0-0 ~e4 domination over the e-file.
17 l:tfdl b5 26 ... ~e5
Maybe this is a bit optimistic. 26 ... llxc6? 27 dxc6 1Ir'xc6 (27 .. :ikc7
18 ~g5! a6 19 l:lbel ~xd3 20 l:xe8 28 'ifd7) 28 'iVd8+ Damn, unlike
'iVxe8 21 'it'xd3 Steinitz I have no riposte against the
back rank mate as after 28 ... ~f8
29 ~h6 wins.
27 ~e3 ~b8
27 ... ~xc3 28 "xc3 llxc6 29 dxc6
(29 'ifxc6 "ili'a7 30 e5 'ii'd4) 29 .. .'iVc7 is
no fun for Black at all.
28 'ifd4 f6 29 ~a5 ~d6
Black still can't take the pawn.
21..Jle8?
Black should defend the e7-pawn
with the queen from d7 or b7.
22 :lel 'ifb7 23 d5 liJe4
After 23 ... h6 24 Ji.f4liJc4 25 liJdl g5
26 ~g3 liJxdl 27 lle7! White is better
as Mikhail Gurevich pointed out.
24 liJd2 liJxd2
258
30 1fc3 :e8
Of course with such domination
White must be winning.
31 a3 ~g7 32 g3 .i.e5 33 'ii'c5 h5
34 .i.c7 ~al 35 .i.f4 'ii'd7 36 :c7
This is more or less the end.
36 ... 'ii'd8 37 d6 g5 38 d7 l:tf8
39 .i.d2 .i.e5 40 l:tb7 1-0
14...~e8
J.Lautier - G.Kasparov I knew how to neutralise White's
Tilburg 1997 witty pawn sacrifice.
15 ~xc6 bxc6 16 1%a4 f6 17 l:tfal
1 c4 c5 2 ~f3 ~f6 3 ~c3 d5 4 cxd5 'ittf7 18 :xa7+ l:txa7 191ba7+ ~e7
tiJxd5 5 e4 tiJb4 6 ~b5+ ~8c6 7 d4
cxd4
20 :'c7?? Ih-Yz
With this bad mistake Lautier offered

8 a3 a draw. I accepted. I thought that had I
This is the start of the so-called played to exploit the back rank, the
Dream Variation. This move occurred Frenchman would have a riposte just
first in a dream of Hungarian 1M like Steinitz. But 20 ... c5! would win a
Navarovszlcy. His friend Csom played pawn as 21 bxc5? nb8 wins.
it first and beat Stean with it in Las
As well as Botvinnik, Steinitz also
Palmas 1978. However for me this
contributed to my understanding of the
game proves to be not a dream but a
concept of attacking on the h-file
nightmare.
when the opponent has play on tht
8 ... dxc3 9 1fxd8+ ~xd8 10 axb4
g-file.
cxb2 11 .i.xb2 e6 12 0-0 ~d7 13 .i.xc6
~xc6 14 ~e5 Here is the game I had in my mind:
259
Reiner - W.Steinitz 17 J:.g2

Game 4, match, Vienna 1860
1 e4 e5 2 lOf3 lOc6 3 d4 exd4 4 .i.c4

.i.c5 5 0-0 d6 6 c3 .i.g4 7 'iWb3 .i.xf3
8 .i.xn+ <itfS 9 .i.xg8 l:txg8 10 gxf3 g5
11 'it'e6 lOe5 12 'iff5+ ~g7 13 ~hl
rlr>h8 14 l:.gl g4 15 f4 lOf3 16 ':xg4
17 .. .'ii'xh2+!!
The sacrifice on the h-file provides a
cute mate on the g-file. Chess is
confusing isn't it? But marvellous for
sure. By the way I also used the motif
of having a knight on f3 and a g-file
16...'iih4!! rook to beat Sunye in Graz 1981 in a
What a nice way to show the sacrificial game.
superiority of the h-file attack over that 18 llxh2 %igl mate
on the g-file! Checkmate and what a neat one!
Steinitz's openings are no longer played in high-class tournaments, though

he has left his mark on this phase of the game. Also I had a completely
different repertoire. Steinitz's legacy was that the king can take a walk to the
centre and several times I used this technique myself. Here are two of
Steinitz's games where his king successfully took an active role.
W.Steinitz - J .Zukertort W.Steinitz - G.Neumann
260
In my games, I did not mind taking G.Kasparov - V.Kramnik

risks and played with my king in
the centre too. You saw my 1993
World Championship loss to Short. I
selected it for the Botvinnik chapter
where I pushed my g- and h- pawns.
But I have plenty more examples of
king play in the centre - in fact you
can see this even in the very last
regular game of my career.
G.Kasparov - V.K.ramnik V.Topalov - G.Kasparov
W.Steinitz - J.Zukertort S <it>e2

London 1872 This is Steinitz's variation. White has
occupied the centre and in return for the
1 e4 e5 2 ·lLIe3 lLIe6 3 f4 exf4 4 d4
loss of castling rights his king will seek
'ir'h4+
shelter behind his central pawns. Of
course if Black can demolish these
pawns, White's king will be exposed.
Between 1900, when Chi gorin played,
and 1963 when Averbakh employed the
opening as White there was only one
game with Steinitz's line.
S... dS 6 exdS Ji.g4+ 7 llJf3 O~O-O
8 dxe6 Ji.eS
This is quite a wild line.
261
9 cxb7+ <it>b8 10 lOb5lOf6 11 ~d3!?
Steinitz continues to walk with his

king in the centre.
11 .. :ifh5
After 11...~f5+ 12 ~c3 lOe4+
13 'It>b3 'iWf6 Lasker lost to Shipley,
USA 1893.
21 ~a3
The king continues on his journey...
21. .. g5 22 b4 'iib6 23 'ilfd4
Zukertort keeps on playing despite
having to swap queens.
23 .....xd4 24 cxd4 lOb6 25 ~b2
lOc4+ 26 ~b3
The king has to move because of the
check but Steinitz liked to move his
king voluntarily as well.
12 ~c3 ~xd4+?
26 ... lOxb2 27 ~xb2 l:txd4 28 <itc3
Zukertort misses his chance to play
12 ... a6! %1hd8 29 %tad 1 l:4d6 30 lbd6 lbd6
13 lObxd4 'iWc5+ 31 .l:[dl %1f6 32 ~c2 <iPxb7 33 ~xh7
~b6 34 h3
14 ~b3 'iib6+ 15 ~b5

White has enough extra material to
give back some to cover his king. 34 ... fJ
15 ... ~xf3 16 'ii'xfJ lbd4 Black can exchange some pawns but
Recovering one piece but he is still in not all of them.
arrears. 35 gxfJlbfJ+ 36 Ad3 .1:[12 37 a4 as
17 ~c6 'iWaS 18 c3 l::td6 19 'ilVc4 a6 38 bxa5+ <itxa5 39 :d5+ ~b6 40 a5+
20 ~a4lOd5 ~a7 41 ~d3 %tfJ 421bg5
262
Just like the previous game he moves

away from the centre.
1I ... dS 12 ~g4+ ~b8 13 eS
White can keep the position closed.
13 ...... g6
42 .. Jhb3
White still has one extra pawn, and
that is enough.
43 ':fS lib7 44 ltcS fS 4S lixfS lie7
46 l:tgS lid7 47 lieS l:tg7 48 ':e8 ':gl
14 <it>f2
49 .i.e4 ltcl + SO <it>b4 cS+ SI ~bS
Returning the king to the centre.
1-0
14 ... hS IS .i.b3 f6 16 exf6 'iixf6+
It is symbolic that Black resigns in
reply to a king move.
17 "'0 'ii'xO+
Exchanging queens takes the
pressure off the king.
W.Steinitz - G.Neumann 18 gxO g6 19 ttJe2 ttJfS 20 .i.xfS
Dundee 1867 gxfS
Black's pawns are vulnerable.
1 e4 eS 2 ttJc3 ttJc6 3 f4 exf4 4 d4 21 c3 .i.d6 22 .i.f4 <it>c8 23 .J:[hgl
-.h4+ S <it>e2 d6 6 ttJO .i.g4 7 .i.xf4 ~d7 24 llg7+ ttJe7 2S ':agl
~xO+ 8 ~xO
S••• ttJge7 9 ~e2 0-0-0 10 ~e3 "f6+
2S ... <it>e6
The king helps the pawns but Black's
king becomes a target in the centre.
26 .i.xd6 .J:[xd6 27 ttJf4+ ~f6
11 ~g3 28 ttJd3 .J:[b6 29 b3 ':h6
263
9 ... 36 10 e4 c5 11 d5 c4 12 i.c2 ~c7

13 ltJd4 ltJc5 14 b4 cxb3 15 axb3 b4
16 ltJa4ltJcxe4 17 i.xe4ltJxe4 18 dxe6
30 ltJe5
White traps the rook in a remarkable
way.
30 .. Jtb5 31 34 ':35 32 b4 ':36 18 ...i.d6
33 ltJd7+ lti'e6 34 ltJc5+ 1-0 Black's pieces look scary, but if they
are not able to make any tangible
Using the king in the centre was a threats he can have problems.
technique I employed right up to my 19 exf7+ "ihf7
very last game. Here is the first loss Kramnik prefers to retain the right to
from the mid-1990s. castle. There is nothing wrong with
19 ... ~xn initiating a march to the
G.Kasparo,7 - V.Kramnik centre and meeting 20 "ifh5+ with
Dos Hennanas 1996 20 ... g6.
20 f3
1 d4 d5 2 c4 c6 3 ltJc3 0jf, 4 ltJf3 e6 This is one of my specialities, I like
5 e3 ltJbd7 6 i.d3 dxc4 7 i.xc4 b5 to block the b7-bishop. In my career I
8 i.d3 i.b7 beat Karpov five times in the main Ruy
Lopez and in each of those games I
blocked his b7-bishop. I did it by
playing d5 or D. Perhaps my win in the
second game of our 1990 world title
match in New York, where my f2-D
was a theoretical novelty, virtually
refuted Karpov's opening. But I
also scored nice victories against
the Hedgehog where I reinforced e4
against a b7-bishop.
20 ... ~h5 21 g3
9 0-0 The immediate king excursion
Unlike Steinitz, I castled here but as resulting from 21 fxe4 would be fatal.
you will see I did not mind returning to 21...'ii'xh2+ 22 ~f2 0-0+ 23ltJD i.g3+
the centre. 24 ~e3 "it'xg2 wins as White's king is
264
Wilhelm Steinitz the JJI
too exposed. However an alternative is

21 h3 'ife5 22 f4 'iff6 23 ~b2 with a
complicated middlegame.
21 .•. 0-0
Nor can the black king take an early
stroll: after 21...llJc5? 22 l:le 1+ ~f7
23 llJf5! 'ifxf5 24 'ifxd6 White wins.
But after 21...llJxg3! 22 hxg3
(22 'ife 1+ llJe4 23 l:la2 0-0) 22 ... 0-0
23 l:la2! (23 ~g2? 'ii'g4) 23 ... ~xg3
27 ~e3 'ii'xa2 28 <it'xe4 and White is
24 1:Ig2 ~e5 25 llJc5 1:Iad8 26 ~e3
winning as Black can do hann to the
~c8! Black is slightly better according
king in the centre.
to Kramnik.
2J llJf3?!
22 fxe4 -.oJ! Bringing the a I-rook into the defence
On 22 ...1:Ixfl +?! I would have with 23 l:la2 was the right move.
undertaken a glorious king-march in 23 ... ~xe4 (23 ... ~xg3? 24 llJf5! l:lxf5
the centre. Had Steinitz seen it he 25 l:lxf5 ~xe4 26 l:lg5 wins.) 24 l:tel
would have loved it. 23 'ifxfl ~b7 (24 ... l:tae8 25 1:Ie3! .i.b7 26 l:lxe8
(Recapturing with the king simply l:lxe8 27 'iffl (27 l:te2 1.1£& 28 llJe6
loses) 23 ... .i.xg3 24 hxg3 ~xe4 ~xg3 29llJxfS .i.xh2+ 30 l::txh2 "g3+
25 l:la2! The most precise way of 31 <it'fl "xh2 32 "g4 itbl + White's
preparing the king march. (25 'ife2!? king is too exposed, and Black holds.)
'ifh 1+ 26 ~f2 'iVh2+ [26 ... :fS+? 27 .....g4 (27 ... 'iVh5 28 llJf5!) 28 llJf5!
27 ~f4] "e4 29 "c4+ and Black is active but
has only one pawn for the piece.)
25 "d3 (25 llJe6 ~xg3 26 l::te3 1:IO!)
25 ... :ae8 26 ':xe8 llxe8
27 ~e I! [27 <ite3 "xg3+ 28 ~d2

"g5+ 29 'ife3 "g2+ 30 'ii'e2 .g5+
with a perpetual is given by Kramnik.]
27 .....g1+ 28 "f1 "xd4 29 "c4+ and
Black has problems in the ending.)
... and though Black has only a pawn
for the piece his bishops provide him
25 ... 'ifh I + 26 <ittf2 'iih2+ with sufficient compensation.
265
23 ...~xg3 24 tOeS? 25 .. .lhfl+ 26 'ii'xfl "xfl+ 27 ~xfl

After 24 .e2! ':'xO 25 ':'xO ~xh2+ l:tc8 28 ~e3 ~f4.
26 ~f2 (26 ~hl ~xe4!) 26 ... 'ilfh4+ 25 ... 'ii'xh2+ 26 ~n
(26 ... ~g3+ 27 ':'xg3 l:t£8+ 28 ':0 I thought Black had no more than a
'ifh4+ 29 ~e3 'it'xe4+ perpetual.
30 <iPd2 After the king march in the 26 ... ~e6!

centre the position transposes to an This came as a cold shower.
equal endgame.) 27 ~fl ~e5 Black However, I still hoped that Steinitz's
has other playable moves as well method would work.
(27 ... ~d6 28 ~b2 ~xe4 29 'ilic4+ <iPh8 27 ~g5!?
30 ~xg7+ and White can force a draw After 27 'ii'd3 'ifhl+ 28 ~e2 'ifel
mate; 27 :ta2 ~b5+ 28 tOd3 ~xd3+ 29
by 30 ... <t>xg7 31 'ii'd4+ Wg8 32 'it'c4+
':'xd3 (29 'it'xd3 'ir'h 1+) 29 ....:£8+ also
~h8 33 'ifd4+) but White can stay in
leads to a checkmate.
the game. After 27 ... ~e5 Black can
If 27 l:ta5 ~c7!.
exploit the king in the centre. 28 'ifc4+
27 ... ~b5+ 28 tOd3 l:te8!
~h8 29 'iff7 'iWh 1+ (29 ... h6 30 'ifxb7 Kramnik conducts the attack
l:td8 31 ~e3=) 30 ~e2 'ilfg2+ effectively.
29 lI82
Other moves also don't offer much
resistance, e.g. 29 ':'el 1i'h 1+ 30 ~e2
l:lxe4+ or 29 :la5 ~xd3+! or29 ~e3
lhe4 30 l:l£8+ ~xf8 31 "0+ ':'f4!.
29 .....hl+
Kramnik was already in time trouble
and misses 29 ... ~xd3+! which is a
forced checkmate. 30 :lxd3 'i!ih 1+
31 ~e2 'ifg2+ 32 ~e3 :'xe4 mate.
30 <iPe2
31 We3 'ii'g5+ 32 ~f2 'ii'h4+ with a
perpetual check. See diagram on page 261.
24 ....:xO 25 lhO I hoped that my king would survive,
I did not want to enter a worse just like Steinitz's, especially in view of
endgame after 25 :ta2. Then comes the time trouble.
266
30 ...::t:xe4+ 31 <it>d2 13 ...'ifd5

If 31 .i.e3 'ifg2+ 32 :f2'ii'xf2 mate. I had already decided not to castle.
31.....g2+ 32 <it>c1 "xa2 331bg3 14 :dl 'ifb5 15 h3
Here 33 :fS+ <it>xfS 34 'iff3+ .i.f4+ If 15 e4 lLlg4 16 .i.b5+ <it>fS.
leads to a checkmate. 15 ...:d8t 16 ttJd4
If 16 e4 ttJxe4.
16.....d5
Exchanging queens was safe enough
but I wanted a sharper fight.
17 lLlo <it>e7
This is clearly Steinitz's influence. I

intentionally did not castle and now
advance with my king. I would never
33 .....al+ 34 <it>e2 'ir'e3+ 35 <ii?bl
have lost this game if I had played
l:Id40-1
safely.
I did not want to risk eventually
18 e4
being checkmated, therefore I resigned.
Kramnik sacrifices a pawn to get
some play.
18 ...lLlxe4 19 ~e3 ~xe3 20 'ihe3
Zurich rapid 2001
'ii'e5 21 'ife1 ttJf6
1 d4 lLlf6 2 e4 e6 I did not feel like playing 21 ... f5,
After losing my title in 2000, like though it was playable. Then 22 :ac I
Muhammad Ali I started to float like a 'i!fb6.
butterfly and broadened my opening 22 :tac1 'ii'b6 23 lLle5 :d4?
repertoire. One move leads to two different
3 lLlo d5 4 lLle3 dxe4 tactical motifs. I became too optimistic.
This was one of my new openings. I knew how many times Wilhelm won
5 e3 a6 6 .llxe4 b5 7 .i.d3 e5 8 a4 b4 with his march in the centre. I couldn't
9 lLle4 lLlbd7 10 ttJ:xf6+ lLl:d6 11 0-0 wait to score with his brand of king
~b7 12 dxe5 .i.:xe5 13 'ii'e2 play.
267
Wilhelm Steinitz the lsi
24 ~xa6!!? I thought he might have missed this.

Kramnik sees a very deep 29lOxb4
exploitation of the risky placement of White has only one pawn for the
the black king. But he misses a simple piece, yet Black is struggling.
win with 24 lOc4!. Then 24 .. .lhc4 29 ... ~e2
(24 ... 'ii'c5 25 lOe3 'itb5 26 lOf5+ wins.) After 29 ... ~b7? 30 a5 ~e7 31 :b8
25 l%xc4 a5 26 'iVe5 and Black has ~d5 32 a6 l%f8 33 lOxd5+ exd5
nothing for the exchange. the lovely 34 l%c8! wins.
24 ... l%xdl Slightly better was 29 ... 1;e7
24 ... Ae4 would have resisted for 30 lOc6+ ~f6 31 b4 g5 32 b5 ~b7
much longer. 25 'ii'd2 l%xe5 26 ~xb7 33 4Ja5! ~d5 34 b6 ~e7 35 b7 (35 :b8
lOd5 (26 ... ':d8 27 ~xd8+ it'xd8 :f8) 35 ... ~xb7 36 lOxb7 and the
28 l%xd8 ~xd8 29 ':c4 b3 30 ltc3 knight ending will be tough for Black.
lte1 + 31 ~h2 l%e2 32 l%xb3 l%xf2 33 as Black's best was to divert the rook
wins according to Kramnik..) 27 ~xd5 from d8 by 29 ... ~c4! 30 ':c8! ~e2
(27 ~c6 l%c8 28 ~b5 and White is 31 f3 rJ;e7 (31.. .h5 32 !it>f2 ~d 1 33 a5)
somewhat better.) 27 .. .lhd5 28 'iff4 32 ~f2 ~d7 33 ':c3 and White wins
and with his king under pressure, Black back the piece, whilst retaining
has to fight to survive. promising chances.
2slbdl ~xa6
25 ... ~d5 is met by 26 ~b5.
Alternatively 25 ... 'ifxa6 26 'ii'xb4+
~e8 27 :d6 4Jd5 28 lha6 lOxb4
29 l%b6 wins.
30 f3!
Kramnik's great idea is to trap the
bishop. Black is paralysed because of
the pin.
26 'ifxb4+! 30 ... hS
Kramnik sees a great idea on the After 30 ... ~e7 31 lOc6+ ~f6 32 b4
horizon. g5 33 b5 r:l;;g7 34 ~f2 ~c4 35 b6 wins.
26 .. .'iixb4 27lOc6+ ~f8 31 b3! ':h6
I had Steinitz in mind and thought I If31...~e7 32 l%d2.
would be able to handle any back rank 32 ~f2 ':g6 33 ~xe2 l%xg2+
checkmate threat. The rook finally becomes active, but
28 ':d8+ lOe8 it is too late.
268
34 ~d3 ':g3 35 a5 lbf3+ 36 ~c4 15 ... lIfS

1-0 "" After 15 ... 'iVb6 16 e5 ttJg6 17 11f7+
I resigned because White's two 'it>e8 Black can do a lot with his king.
queens ide passed pawns are too strong. 16 ':'xfS 'it'xf8 17 ttJxd4 cxd4
18 ttJe2 "f6 19 c3
V.Topalov - G.Kasparov
Linares 2005
1 e4 c5 2 lLlf3 lLlc6 3 lLlc3 e5 4 ..i.c4

d6
I played this line a few times against
Leko, who repeatedly rejected castling
even though he nonnally plays safe.
5 d3 ..i.e7 6 0-0 lLlf6 7 lLlh4 lLld4
After 7 ... lLlxe4 8 dxe4 i.xh4 9 f4 19 ... l:tfS
White has very nice compensation. Now, in order to launch an attack
first, I sacrificed a pawn. 19 ... dxc3
8 g3 ..i.g4 9 f3 ..i.e6 10 i.g5 lLlg8
20 '1IVa4+ lLlc6 21 1:{0 'iVg5 22 ""'3
10 lLlb 1, as Karpov played against
lLld8 23 lLlxc3 1i'e3+ 24 ~g2 l:tc8 leads
Spassky, is a rather similar knight to a complicated position, as Topalov
move. pointed out.
11 ..i.xe7 lLlxe7 12 f4 exf4 13 ..i.xe6 20 lLlxd4 lLlc6 21 '6'0 'tlhO+
fxe6 14 ':xf4 'it>d7 After 21 ... lLlxd4 22 'ii'xf6 l:txf6
23 cxd4 ':0 24 1:{d 1 g5 25 ~g2 (25 g4?
See diagram on page 261. d5!) 25 ... g4 Black probably holds,
though once, in Game 32 of their world
title match in Buenos Aires 1927,
Even in the last game of my care~r I
Alekhine freed a passive rook against
follow the principles of a world Capablanca and managed to beat him.
champion. The king should be safe in So who knows? But who do you trust
the centre. when two world champions play
15 lLlf3 against each other?
269
Wilhelm Steinitz the l·ft
22 lhfl l:bn + 23 'itxfl 4Jxd4 used Steinitz's idea. It made me realise

I knew the principle of having little that somehow they were able to use the
chance of survival in a pawn ending, champion's legacy better than myself.
but during the game I lost faith in all At this stage, writing the Great
champions and principles. Predecessors books was well under
24 cxd4 d5 way. After this game I made the
I played this instantly. Maybe playing dramatic announcement of my
for a fortress with 24 ... ~e7 was better. retirement from professional chess.
25 ~f2 rJ;e7 26 lPfJ ~f6
You may joke that I retired to become
an amateur in politics. But let me
remind you I gained the chess crown in
my second match against Karpov.
Maybe I will become president after
entering the race for a second time and
will then rule for 15 years just as I did
in chess?
Some may think I'd rather fight
Putin than the new generation of
27 h4?? chessplayers on the chessboard.
27 <ifi1g4 wins after 27 ... g6 (27 ... h6
But now I'll make a sensational
28 ~h5; 27 .. /.ti'g6 28 <ifi>f4 <iti'f6 29 h4
confession... I stopped playing chess
h5 30 g4 hxg4 31 ~xg4 rwtg6 32 <ifi>f4
not because of any particular interest in
'ltrf6 33 h5 wins.) 28 h3! (28 h4? h6=)
Russian politics, but because my last
28 ... h6 (28 ... h5+ 29 ~h4) 29 h4 as
game made me lose all faith in my
Krnic pointed out.
Predecessors.
27 ... g6??
A blunder in reply. 27 ... h6! draws,
since if 28 <ifi1g4 then 28 ... g6.
• • • •
28 b4 b5 29 <iti'f4 h6 30 ~g4 1-0 Now at last we take back the
I resigned here. Topalov was a point commentary from Garry and speak
behind me before this last round game, with our own voice!
therefore I had to share the first prize. Maybe these losses were hard to take,
Once I finished the game I somehow but they were a necessary part of
sensed that Veselin would win the next Kasparov's career. His losses and the
World Championship. champions' wins were instructive and
one can learn and profit considerably
I suddenly understood that I could
from studying them. Just as Kasparov
trust nobody - not even the champions.
himself did.
Then who could I follow? How was I to
carry on playing chess? My successors What did these losses help him to
Kramnik and Topalov beat me when I achieve? He completed a fabulous
270
career, the best a chessplayer has ever had a plus score against all his
produced. He was the dominating force close rivals, except Kramnik and
in world chess for approximately two against some of them very convincing
decades. He was world champion for plusses.
15 years. He was a world-class player
His contribution to the development
for 25 years.
of chess is immense, although we
He produced the greatest number of need still more time to fonn a
superb creations by anyone who ever comprehensive judgement.
played our game. Kasparov himself
He left the game in a different state
estimated 250 of his games were of top
from when he found it. Partly because
quality and we are inclined to agree
of computers, chess culture would have
with him. In fact if one counts the great
developed anyway, but his unique
games he lost or drew maybe he played
artistry has been a telling factor.
even more than three hundred
superlative games. Furthennore no This time we dared to joke around
other player faced such strong with his games, but we never trivialised
opposition as him. them, nor for a single moment forgot
that they will continue to bring joy to
Very few won as many individual
new generations of chess fans.
tournaments as him - and he must be
one of the players who gained the most Kasparov's impact on chess will be
material rewards from chess. He also felt for as long as it is played.
271

Karolyi T., Aplin N. - Kasparov. How His Predecessors Misled Him About Chess (2009)

Uploaded by

Copyright:

Available Formats

Karolyi T., Aplin N. - Kasparov. How His Predecessors Misled Him About Chess (2009)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Karolyi T., Aplin N. - Kasparov. How His Predecessors Misled Him About Chess (2009)

Uploaded by

Copyright:

Available Formats

First published in the United Kingdom in 2009 by

An imprint of Anova Books Company Ltd

Copyright © B T Batsford 2009

Text copyright © Tibor Karolyi, Nick Aplin

The moral right of the authors have been asserted.

All rights reserved. No part of this publication may be reproduced, stored

Reproduction by Spectrum Colour Ltd, Ipswich

Or try your local bookshop

Anatoly Karpov the 12th 9

Robert James Fischer the 11 th 54

Boris Spassky the 10th 81

Tigran Petrosian the 9th 94

Mikhail Tal the 8th 123

Vassily Smyslov the 7th 143

Mikhail Botvinnik the 6th 169

Max Euwe the 5th 192

Alexander Alekhine the 4th 202

Jose Raul Capablanca the 3rd 218

Emanuel Lasker the 2nd 235

Wilhelm Steinitz the 1st 254

A.Karpov - P.Markland A.Karpov - G.Kasparov

25 ...'ii'g7! 37 ... a4!

38 l::td2 l:tb6 39 iLf1 ~n 40 l:tee2 A.Karpov - P.Markland

would be able to get down to the

See diagram on page 9.

Again I push the rook's pawn just

1 d4 ltJf6 2 e4 g6 3 ltJe3 dS 4 iLf4

Karpov contributed to one of my losses in the final of a knockout

V.Korchnoi - A.Karpov 8 cxd5 xd5 9 xd5 exd5 10 • xe7

15 'ilVxd5 28 ... h5!

41..J:ta3 42 nd6+ ~g7 43 h4 nb3 Would you believe what happened?

Sadovsky - A.Karpov G.Kamsky - G.Kasparov

Z.RihU - A.Karpov Y.Nikolaevsky - G.Kasparov

J.Nunn - A.Karpov L.Zaid - G.Kasparov

1 c4 ttJf6 2 ttJc3 e5 3 g3 g6 4 ~g2

See diagram on page 19.

llllJd5 Karpov achieves symmetry with his

Karpov puts his knight on the edge

The next game is unusual, as I was

jVf7 49 'iWxc6 d4+ 50 ~bl d3 51 cxd3 17 g3 liJdf6 18 c5 ~d7 19 lIb3 ~h6

1 d4 ttJf6 2 c4 g6 3 ttJc3 ~g7 4 e4 d6

Black gets his knight to the h-file, but

24 ... lLlh3+!! 25 c;t>g2 l:hb5 26 ~g3

24.i.c4! Now back to my game with Kamsky:

2S ... 'VWh4 Y.Nikolaevsky - G.Kasparov

34 l:txf8+ <;t;xf8 35 l:Ixe5 'ii'g4+ 36 Wfl

The knight naturally goes to the edge. Malta 1980, or Bouaziz-Karpov,

32 lObI! lOg7 33 lOd2 lOe6 34 b4 2S ... lOb8!! 2610D lOd7 27 '1t>g31OcS

32 dxeS ':xb2+ 33 ~fl ltJxc3

Karpov sacrificed the e5-pawn in a sharp Sicilian against Ljubojevic.

A.Karpov - L.Ljubojevie G.Kasparov - A. Yermolinsky

1 e4 e5 2 1Of3 d6 3 d4 lOf6 4 lOc3

26 l:td8+!! This lovely shot decides

9.•. b5 42 ... ~d6 43 'iVb8+ ~xd5 44 'iVd8+

10 ... .i.xc6 11 .i.d3 .i.e7 12 e5 dxe5

23 ... 'iIi'd4 24 -.dl a5 25 It:Jh2 g4

I did not mind sacrificing the e5

See diagram on page 2 7.

Karpov sacrificed the b6-pawn in S.Garcia Martinez - A.Karpov

And below is another example