The Instructor

A Battle of Opposites
The following tense encounter, offered for your perusal, is a most instructive one
for several reasons.

1) It featured two great players. One of them, Mikhail Tal, was at the peak of his
chess career. The game was played as part of the World Championship Match in
which Tal secured his title.

2) We can contrast the points of view of both players, and compare their
assessment of this battle’s ebb and flow, which is always both entertaining and
The very useful. It’s a shame that we so seldomly see such “face-to-face” annotations.
One of the few pleasant exceptions was the book of the 1969 Petrosian-Spassky
World Championship Match, consisting of notes to the games written by the
Instructor players’ trainers: Isaak Boleslavsky and Igor Bondarevsky. Studying this book
gave me a great deal of pleasure.
Mark Dvoretsky
Mikhail Botvinnik’s commentaries are taken from his four-volume selected
games – these will appear as bold text; Tal’s comments (and both sets have been
edited) will appear in italics; these are from his monograph on the first Botvinnik-
Tal match. This is a wonderful book; in my view, one of the best books in all of
chess literature. Tal’s annotations are quite genuine, and very detailed: each
game receives several pages of entertaining text. Tal was an outstanding writer,
with a lively, picturesque style. His comments never devolve – as so often
happens these days, now that computers have gotten involved in analysis – into
an endless rehash of variations. Quite the contrary: at every point in the game,
Tal offers us his view of what is happening on the board – a positional
assessment – and not a formalized one, either, but a lively, dynamic one. The
most valuable characteristic of this book is the way it overflows with
psychological observations and considerations. Psychology is a vital element of
the chess struggle; yet it is portrayed in the pages of chess literature in either too
primitive, or too formalistic and unconvincing a fashion. But here we can
observe a believable psychological picture of a great match and each game of
that match in particular, described by one of its main participants. An additional
important element, and also rarely seen, is that the times after each move are
noted. In short: Tal’s book can be recommended without hesitation to any
chessplayer, whether he seeks to achieve further progress in his game, or simply
wants a pleasurable read.

3) The Botvinnik-Tal match saw a clash of antipodes: players with diametrically

opposed approaches, both to the chess struggle and to life. I have always been
amused at ideological attempts to assign even such wildly differing chessplayers

file:///C|/cafe/Dvoretsky/dvoretsky.htm (1 of 15) [2/4/2005 1:06:56 PM]

The Instructor

to a single “Soviet School of Chess.” Of course, this is partly a matter of

definition: if by “school” you mean a common country and language, and
consequently all books and articles published and read in the Russian language;
if you mean that most of one’s contacts occur precisely with chessplayers from
that country; and if you’re also referring to the necessity of subordinating
oneself, to a greater or lesser extent, to the same set of rules and restrictions, of
working with the same set of chess and non-chess related functionaries, etc. –
why, I suppose that would be true. But to find a creative mesh, or matching ideas
– now, that’s hard. That great chessplayers influence one another is indisputable;
but such influence is hardly limited by national boundaries; and it’s hard to see
why Tal would belong to the same “school” as Botvinnik, to a greater degree
than, say, Fischer or Timman.

The game we shall be examining shows especially clearly not only the
differences in the two players’ styles, but also the positive and negative aspects
of each style. Well, this would be better discussed later: both while we are
analyzing this game, and after we have finished it.

4) Tal managed to find himself in a strategically lost position (playing White, and
very quickly, too!); but then, playing with rare energy and resourcefulness, he
confused his powerful opponent and saved himself. A useful lesson to be learned
from this example is that absolutely hopeless positions are almost nonexistent,
once you learn the methods of active defense.

Tal – Botvinnik
Moscow 1960, Match Game 3

1. e2-e4 (0.00) c7-c6 (0.00)

2. Nb1-c3 (0.01) d7-d5 (0.01)
3. Ng1-f3 (0.01) Bc8-g4 (0.02)
4. h2-h3 (0.01) Bg4xf3 (0.02)
5. g2xf3?! (0.01)

In this, as in other games of the match, Tal

used this sort of play to create original
opening situations. His first concern was
not whether his conceptions were correct,
but to force his opponent to use up more of
the time he would find so necessary in time-
trouble.

My opponent, like many others, considered

my main weapon to be my preparations for
the event. But he did not consider another
and perhaps still more important
component of my practical strength: positional understanding. It was this
that allowed me to find the proper plans in unfamiliar situations.

In this case, it seems to me, in spite of the purely psychological plusses (an
absolutely new position complete devoid of any possibility that Black might have

file:///C|/cafe/Dvoretsky/dvoretsky.htm (2 of 15) [2/4/2005 1:06:56 PM]

The Instructor

done any home “grinding”), the move 5. gf has some positional basis. first of all
it strengthens White’s center, and second of all it opens the g-file along which he
might be able to create pressure in the future. If Black immediately tries to refute
this move and plays the straightforward 5...e5, then the following factor comes
into effect: in the open game, the strength of the pair of bishops (especially the
light-squared one (not having an opponent) sharply increases. Here, for
example, is one of the training games played before the match (against Alexander
Koblents, Tal’s trainer).

1. e4 c6 2. Nc3 d5 3. Nf3 Bg4 4. h3 Bxf3 5. gf e5 6. f4 de After 6...ef 7. ed cd 8.

d4, we have something like a King’s Gambit, in a version favorable to White.

That’s debatable: Black could go into an unclear endgame with 8...Qe7+! 9. Qe2
f3.

7. fe Qd4 8. Qe2 Qxe5 9. d4 9. Qxe4 was stronger, with a much superior

endgame. But could any of you have withstood the temptation of playing such a
tempting sacrifice, especially in a training game?

9...Qxd4 10. Nxe4 Be7 11. Bf4 Qxb2 Not 11...Nf6 12. Nd6+ Kf8 13. Qxe7+!
Kxe7 14. Nf5+ and 15. Nxd4.

12. Rd1Nf6 13. Nd6+ Kf8

14. Qxe7+!? Kxe7 15. Nf5+ Ke8! Black can’t

play either 15...Kf8 16. Rd8+ Ne8 17. Bd6+,
or 15...Ke6 16. Nxg7+ Ke7 17. Bd6+ Kd8 18.
Ba3+.

16. Nxg7+ Kf8 17. Bd6+! (17. Rd8+ Ke7)

17...Kxg7 After 17...Kg8 18. Rg1, Black is
defenseless, for example: 18...Qc3+ 19. Rd2
Qxd2+ 20. Kxd2 Ne4+ 21. Ke3 Nxd6 22. Ne6
mate.

But if Black plays 19...Qa1+ (instead of

19...Qxd2+?), then White has to take the draw (20. Rd1 Qc3+), since 20. Ke2?!
allows the only, yet sufficient, defense: 20...h5!

18. Rg1+ Ng4! The only move. Considerably worse would be 18...Kh6 19. Bf4+
Kh5 20. Be2+ Kh4 (20...Ng4!? 21. Rxg4 Qb4+ 22. c3 Qxf4 23. Rxf4+ Kg5 24.
Rxf7 would be hopeless as well) 21. Bg3+ (White has a mate in 4 by 21. Bg5+!
Kxh3 22. Kf1!) 21...Kxh3 (21...Kg5 lasts longer, although his position after 22.
Be5+ Kf5 23. Bxb2 is very bad) 22. Bf1+ (there are two ways of giving mate in
three here: 22. Kf1! and 22. Rh1+! Kg2 23. Rh2+ Kg1 24. Kd2 mate) 22...Kg4
23. Be5+ Kf5 24. Bxb2, with a powerful attack.

19. Rxg4+ Kf6 20. Rf4+ Kg7 Drawn by perpetual check. 20...Kg5 is too risky:
21. Rg4+ Kh5?! 22. Be2 Qxc2 23. Bf4! Qxe2+! 24. Kxe2.

file:///C|/cafe/Dvoretsky/dvoretsky.htm (3 of 15) [2/4/2005 1:06:56 PM]

The Instructor

At the very end of 2004 I came across Igor Zaitsev’s book Ataka v silnom punkte
in which he continues the analysis: After 24…f6 25.Rd6 Rf8 he looks at two
interesting lines, each of which deserves consideration:

(A) 26. Kf1 b6! 27. Re6 Na6 28. Re7 Rh8! 29. Reg7 Raf8! [29…Nc5? 30. Bg5!!
fg 31. R7xg5+ Kh6 32. f4! and mate is unavoidable] 30. Bg3 Kh6! 31. Bf4+ Kh5
with a draw. However, in the final position White need not be satisfied with a
draw, since after 32. Bd6! Kh6 [a mate that is already familiar to us arises after
32…Rd8 33. Bg3 Kh6 34. Bf4+ Kh5 35. Bg5!!] 33. Bxf8 Rxf8 34 Rxa7 leads to
a won ending.

(B) 26. Be3 b6 27. f4 Na6 28. f5 Rae8 29. Rg7 h6 30. Rg6 Rxe3+ 31. Kxe3. Now
by continuing 31…Nb4!, Black can successfully resist. Therefore stronger is 27.
Bd4! (instead of 27. f4) 27…Na6 [27…f5? 28. Be3 f4 29. Rxf4] 28. Bxf6.

Botvinnik correctly evaluated the position, and decided to keep it closed. In the
normal course of play, this would lead to a complex and roughly even game.

5. ... e7-e6 (0.17)

6. d2-d4 (0.02)

Worth considering was 6. d3 followed by f2-f4.

6. ... Nb8-d7 (0.19)

The most comfortable setup for his pieces. For now, Black does not define the
position of his king’s knight, so as not to give White the opportunity to play e4-e5
with tempo.

7. Bc1-f4? (0.10)

In our home preparations, we had looked at 7. Be3, intending to continue Qd2

and 0-0-0, in order to play actively in the center. In this case, 7...Qb6
accomplishes nothing, because of the simple 8. a3, when Black can’t continue
8...Qxb2 9. Na4. After 7. Be3, White would keep control of the vital f4 square; on
7...Bb4, he would have the simple reply 8. a3 at his disposal. To his cost, White
decided on a further “strengthening” of this variation, and found a most dubious
move over-the-board.

Carelessly played; now the maneuver Ng8-f6-h5 will come with an extra
tempo for Black.

7. ... Bf8-b4 (0.31)

This makes the above-mentioned maneuver stronger; besides, the bishop is

safe here, because the continuation a2-a3 and b2-b4 would make it
impossible for White to castle queenside.

A good positional move. Now Black responds to 8. a3 with 8...Ba5 followed by

file:///C|/cafe/Dvoretsky/dvoretsky.htm (4 of 15) [2/4/2005 1:06:56 PM]

The Instructor

Bc7, forcing White to trade the dark-squared bishops – not a good thing with this
pawn structure – or lose time retreating.

At the same time, Black avoids a trap. Had he played 7...Qb6, so as to meet 8. a3
with a more favorable 8...c5, White could have replied 8. Qd2, and on 8...Qxb2
9. Rb1 Qa3 10. ed ed (10...cd 11. Nb5 – the consequences of 11...Qxa2 or
11...Qxf3 are not clear, so 11. Rb3 Qa5 12. Nb5 Qxd2+ 13. Kxd2 deserved
consideration, with good compensation for the pawn) 11. Qe3+, when Black does
poorly with either 11...Be7 12. Rxb7 Ngf6 13. Rb3 Qa5 14. Bd6, or with 11...Kd8
12. Rb3 Qa5 13. Rxb7, when he can’t play 13...Bb4, because of the threatened
14. Bc7+. In all these variations, the white bishop stands very well indeed at f4.
However, as we have already noted, the strength of the two bishops is evident
only in open games, and Botvinnik is not going to play for this.

8. h3-h4? (0.24)

But this is just a mistake, based on a miscalculation. The strategic aim of the
move is to meet 8...Ne7 with 9. h5; the tactical point is that in some lines, White
can bring his rook out via h3. But the lesser evil was still 8. a3 Ba5 9. b4 Bc7 10.
Be3.

8. ... Ng8-f6 (0.36)

9. e4-e5? (0.47)

After this, White’s position is strategically

lost, since all hope of opening the game
disappears, and with it any compensation
whatever for his multiple weaknesses. When I
played 8. h4, I had intended the gambit line 9.
a3 Bxc3+ 10. bc de 11. fe Nxe4 12. Qf3 Qa5
13. Rh3 Ndf6. White had further calculated
the following variation: 14. h5 0-0-0
(14...h6!?) 15. h6 gh 16. Be5 Ng5 17. Qxf6
Nxh3 18. Kd2 Ng5 19. Bd3. I was examining
this position, and considering that White had a lot of initiative for his exchange
sacrifice – when suddenly I saw the “cold shower” – 19...Qxe5!, when the
outcome is a quite ordinary endgame... Typical Tal: a very long, and completely
unnecessary, calculation. The final position is easily won for Black, even without
that spectacular stroke on e5 (19...Rhg8!?). And earlier on, Black could also win
with continuations like 15...g5!, or 18...Nxf2!

Since the e4-pawn could only be defended by un-esthetic moves like 9. Qd3 or 9.
Qe2, White decided, “with pain in his heart” on yet another anti-positional
move.

Unexpectedly, White changes his plans. If he had intended to close the

center, this could have been done without the move 7. Bf4. He should have
decided on 9. Qd3.

file:///C|/cafe/Dvoretsky/dvoretsky.htm (5 of 15) [2/4/2005 1:06:56 PM]

The Instructor

9. ... Nf6-h5 (0.40)

10. Bf4-g5 (0.48) Qd8-a5 (1.00)

The means of realizing an advantage, once achieved, is a matter of each player’s

taste. Black could have played 10...f6 11. f4 g6. Of course, in that event, White
would have had some sort of murky hopes of creating play against the e6-pawn.
Above all else, Botvinnik is trying to deprive White of counterplay.

In this match, I strove for a strategy that Tal would find most unpleasant,
which was: to avoid opening the game, when his calculating abilities would
give him a great advantage. This was exactly why I rejected 10...f6.

However, I was not consistent enough in my employment of this technique;

additionally, poor sporting form interfered with my implementation of this
form of struggle against Tal.

Note here that Botvinnik is talking just about aiming for closed, strategic
positions, and not at all about passivity, or avoiding favorable complications.
Passive opposition plays right into the hands of aggressive and resourceful
players such as Tal. Battling them requires firm counteraction, on a strict
positional basis – this is precisely how one exploits an opponent’s positional
liberties.

11. Bf4-d2 (0.51) Qa5-b6 (1.03)

12. a2-a3 (0.53) Bb4-e7 (1.04)
13. Bd2-e3 (0.58)

Now on 13. Bg5 f6 is considerably stronger.

13. ... g7-g6! (1.13)

Of course not 13...Qxb2, because of 14.

Na4.

White’s weaknesses are fixed; now Black can

proceed to lay siege to them. His knight plans
to relocate to the hugely profitable f-square.

From this moment, Tal begins a skillful

process of complicating the game, attempting
to lure his opponent into winning a pawn in
one way or another.

14. Nc3-a4! (1.13)

Now if Black replies 14...Qa5+, White can choose between 15. Qd2 Qxa4 (Black
has a favorable endgame with 15...Qxd2+) 16. b3 Bb4 17. c3 (17. Qxb4!? Qxb4+
18. ab isn’t bad, either) 17...Qxb3 18. cb, with compensation for the pawn; or 15.
c3 b5 16. Nxc5 Bxc5 17. b4! Both lines lead to an improvement in White’s
position. Botvinnik does not deviate from his previously-laid plan, and

file:///C|/cafe/Dvoretsky/dvoretsky.htm (6 of 15) [2/4/2005 1:06:56 PM]

The Instructor

concentrates his firepower on the h4-pawn.

14. ... Qb6-d8 (1.18)

15. Qd1-d2 (1.10)

Tempting Black into the line 15...Bxh4 16. Rxh4 Qxh4 17. Bg5 Qh2 18. Qb4.
When you are playing Tal, looking at such lines is just a waste of time. Even
if objectively poor, these lines would favor him subjectively. Black therefore
chooses the prosaic transfer of the knight to a strong position at f5.

The World Champion’s logic is understandable, but not indisputable. Such

players as Viktor Korchnoi and Lev Polugaevsky – outstanding calculators
themselves – did not believe Tal’s calculations, tested them, and sometimes
found mistakes, which they then successfully exploited. Not accidentally, both
players had terrific plus scores against him.

15. ... Nh5-g7 (1.21)

Ostentatiously avoiding any sort of complications which in this instance would

have been “life-giving balm” to White. Curious variations might arise after
15...Bxh4. White could then play either 16. Bh6, threatening Bf1-h3-g4, with a
more pleasant position than the one he gets in the game (one of his weaknesses
has disappeared, in any case); or he could decide upon the very sharp 16. Rxh4
Qxh4 17. Bg5 Qh2 (on 17...Nxe5?, Tal gives 18. Bxh4 Nxf3+ 19. Ke2 Nxd2 20.
Kxd2, with active White pieces; but instead White just wins by 18. de! Qxa4 19.
b3) 18. Qb4 f6 19. Qxb7 Rb8 (19...0-0 20. Bh6) 20. Qxc6 fg, and now not 21.
Qxe6+ Kd8 22. Nc5 Nhf6!, as given by Tigran Petrosian, but the immediate 21.
Nc5! 0-0 22. Nxd7, with a very sharp game, perhaps not unfavorable to White.

Objectively, the whole line still favors Black: he obtains a clear advantage by
continuing 19...0-0! 20. Bh6 (20. Qxd7 fg 21. Qxe6+ Kh8-/+) 20...Rfd8 21.
Qxc6, and now either 21...Kh8 22. Qxe6 fe-/+, or 21...Ng7 22. Bxg7 (22. ef Rac8
23. Qa6 Qxh6-+) 22...Kxg7-/+.

Had Black played 15...b5 here, White would naturally have tried for an open
game, regardless of material loss, by continuing: 16. Nc5 Bxc5 17. dc Nxe5 18. 0-
0-0, etc.

Let’s check out this “etc.”, by continuing the variation: 18...Nxf3 19. Qb4!? a5
20. Qg4 Qf6 21. Be2 Ne5 22. Qd4 Nd7 23. Bxh5 gh – Black’s advantage is
obvious. No better would be 18. Be2 (intending 19. f4) 18...Nc4 19. Qc3 Qf6-/+.
So objectively, White would have done better to avoid the pawn sacrifice by
playing 16. Nc3.

16. Be3-g5 (1.19)

file:///C|/cafe/Dvoretsky/dvoretsky.htm (7 of 15) [2/4/2005 1:06:56 PM]

The Instructor

White can catch his breath. The scout on h4

is shielded now, but its position is strangely
reminiscent of the “Trishkin Caftan;” soon,
the pawn at f3 will start “wobbling”, too.

16. ... h7-h6! (1.28)

A fine positional continuation. Black frees his

opponent of his weak h-pawn, but at the cost
of exchanges favoring Black. On 16...Nf5 17.
Bxe7 Qxe7 18. h5, White’s position would be
much improved.

The unavoidable opening of the h-file will be good for Black, since the f-
pawns’ weakness will become concrete.

17. Bg5xh6 (1.28)

Taking matters into the endgame by 17. Bxe7 Qxe7 18. Qb4 b6 would have
turned the game into a process of realization of Black’s overwhelming advantage
in position. Quite naturally, White tries to maintain the tension as much as
possible.

But trading queens is not forced. It’s not clear whether Black has any great
advantage after 18. Bd3. Consequently, one can also doubt the efficacy of
Botvinnik’s “academic” approach to the position (15...Ng7, instead of the more
straightforward 15...Bxh4 or 15...b5).

17. ... Ng7-f5 (1.29)

18. Bh6-f4 (1.30) Rh8xh4 (1.31)

If Black takes the pawn with a minor piece, it will be pinned on the h-file.

There was no reason to weaken his position by 18...g5 19. Be3 Nxh4 20. 0-0-0 –
the exchange of rooks favors Black.

19. Rh1xh4 (1.35) Nf5xh4 (1.31)

20. 0-0-0! (1.35)

Defending the f3-pawn is a hopeless task.

Thus, 20. Qe3 could be met by 20...Bg5 21.
Bxg5 Qxg5, etc. White doesn’t have to take
on g5 with the bishop: 21. 0-0-0! is better.
And instead of the spectacular, but
insufficiently effective, move 20...Bg5, Black
has the much stronger 20...Qa5+! 21. Nc3
c5!, with an overwhelming advantage.

file:///C|/cafe/Dvoretsky/dvoretsky.htm (8 of 15) [2/4/2005 1:06:56 PM]

The Instructor

After 20. Be2, White loses the two-bishop advantage and the right to castle at the
same time, after 20...Ng2+.

Now he expected, after 20...Nxf3 21. Qe3 Nh4 22. Bh3 Qa5 23. Bg5!, to disturb
Black’s king. This variation may also be continued: 23...Bxg5 24. Qxg5 Nf3 25.
Qf4 Qxa4 26. Qxf3 c5 27. Rh1! (threatening 28. Bxe6) 27...Qc6 28. Bf1 0-0-0
29. Qxf7 cd 30. f4, with chances for both sides.

20. ... b7-b5! (1.37)

At the appropriate moment – right here, after White has castled – this sortie
is completely justified. After 21. Nc3 Nb6 Black has a very strong attack.
(Tal didn’t like 21...a5, with the threat of 22...b4; he also (though
mistakenly) feared 21...Qa5, which he could have met by 22. Be2 Nb6 23.
Rh1.) The other possibility – taking the f3-pawn – would have been
inconsequent, as it would hand the initiative over to my opponent.

Tal employs a stratagem which is characteristic of him: at any cost, he tries

to open up the game, in order to make it harder for his opponent to calculate
the variations.

Passive play would be absolutely fatal to White; therefore, he makes one more
desperate try: at the cost of a pawn, he tries to take the initiative.

21. Na4-c5! (1.42) Nd7xc5 (1.40)

22. d4xc5 (1.42) Be7xc5 (1.43)

It would have made sense here to take another pawn by 22...g5! 23. Bg3 Nxf3,
for example: 24. Qe3?! (24. Qe2 g4 25. Bg2 was better, with chances for
equality) 24...g4 25. Kb1 Qa5 (threatening 26...b4) 26. Qf4 Qa4!, or 25. Be2 Bg5
26. Bf4 Bxf4 27. Qxf4 Qg5 28. Qxg5 Nxg5 29. Bxg Ne4 – in both cases, Black
has a great advantage. On the other hand, the text looked absolutely natural, and
so neither side probably even noticed there was an alternative possibility.

23. Bf1-e2 (1.48)

23. Bg5? fails against 23...Nxf3.

23. ... Bc5-e7 (1.48)

White has some sort of compensation for the sacrificed pawn. The position has
been opened somewhat, the White pieces are now more active, and his rook is
ready to occupy the open file. With accurate play, of course, this would not have
been enough for equality; but already, time-pressure was making its presence
felt.

And so, Black has an extra pawn in a quiet position; the outcome would
appear to be decided. Unfortunately, no – there are still 17 moves to be
made before the time-control, and there’s no longer so much time left to
think.

file:///C|/cafe/Dvoretsky/dvoretsky.htm (9 of 15) [2/4/2005 1:06:56 PM]

The Instructor

But not so little, either: 42 minutes.

24. Kc1-b1! (1.50)

A useful prophylactic move. White takes his king off the c1-h6 diagonal,
intending to continue c2-c4 under the right circumstances – such as after
24..Kd7. After 24...Kd7 25. c4?! bc 26. Bxc4, Black would have had the
powerful reply 26...g5!, followed by 27...Nxf3. White would therefore have done
better to play 25. Rh1.

24. ... Qd8-c7 (1.56)

Black prepares to castle long.

25. Rd1-h1 (1.53) 0-0-0 (1.58)

26. Bf4-g3 (1.54) Nh4-f5 (2.00)
27. Rh1-h7 (1.54)

Perhaps my first truly active move in the

game. Now Black has to resolve the problem
of how to defend the f7-pawn. Short of time,
Botvinnik chose a safe, but passive
continuation.

After 27...Nxg3 28. fg Qxe5 29. f4 (29. Rxf7

Bf6 30. c3 was also possible) 29...Qf6 30.
Qa5, Black’s weakened queenside deprives
him of winning chances. But with 27...Bc5!,
followed by Rd7, Black would retain a very
solid position; and after the necessary
prophylaxis, he could begin advancing his queenside pawns. This would, in the
end, apparently, lead to a win.

The white rook’s invasion of the seventh rank is not dangerous. It’s only one
threat, which could have been met most simply by 27...Bc5, so as to defend
the f7-pawn, if necessary, by rook to d7. The text is more passive, although it
still doesn’t spoil anything.

27. ... Rd8-f8 (2.04)

On this square, it turns out the rook is not so secure. The white bishops soon find
their voice.

28. Bg3-f4 (1.55)

White’s idea is quite simple: he intends 29. Bd3, followed by trading on f5; then,
after Bh6, he can start to “harvest” the seventh rank. Here again, Black had to
play 28...Bc5. Botvinnik counted on trading off the active rook, but meanwhile,

file:///C|/cafe/Dvoretsky/dvoretsky.htm (10 of 15) [2/4/2005 1:06:56 PM]

The Instructor

he leaves the residence of his king insufficiently guarded.

28. ... Qc7-d8 (2.12)

29. Be2-d3 (2.00) Rf8-h8 (2.19)

The further course of the game and the

analyses which followed confirmed that
this was the most logical continuation,
parrying the threatened 30. Bxf5 gf 31.
Bh6, and leading to simplification.

Whereas, in point of fact, neither the further

course of the game nor analysis supported
Botvinnik’s opinion.

The conclusion of the idea undertaken by

Botvinnik back on move 27. Black forces the
exchange of rooks, but fails to achieve the desired result. 29...g5 looked tempting
here, and only after 30. Bh2 (30. Bxf5!?) would Black continue 30...Rh8. White
had an interesting response to this: 31. Bxf5 ef 32. Qd3! Rxh7 (32...f4 is also
unclear after 33. Qf5+ Kb8 34. Rxf7! Rxh2 35. Qe6 Bc5 36. Qxc6 Qc8 37.
Qxb5+) 33. Qxf5+ Qd7 34. Qxh7 Qe6 35. f4, and the endgame turns out to be
favorable to White, since he has gotten rid of his doubled pawns, and his queen
rules quite effectively from the rear.

30. Rh7xh8 (2.12)

I really wanted to sacrifice my rook here by 30. Rxf7!? Qe8 31. Qa5 (after 31.
Rf6 Bxf6 32. ef e5 [or 31. Rxe7 Qxe7] the compensation for the exchange is
insufficient) 31...Qxf7 32. Qxa7, with the terrible threat of 33. a4. At first, it
seemed to me that White is assured of a perpetual check here; but then I found
the strong move 32...Rh7! for Black, when he manages to get his king to the other
side after 33. a4 Bd8 34. Qa8+ Kd7 35. ab Ke8! (not 35...cb 36. Bxb5+ Ke7 37.
Qa3+!) 36. bc Qa7, when Black wins.

But the enforced trade of rooks turns out to be not so unpleasant after all: White
gets the chance to begin an attack on the king, which soon brings about a
peaceful conclusion.

30. ... Qd8xh8 (2.19)

31. Qd2-a5 (2.13)

file:///C|/cafe/Dvoretsky/dvoretsky.htm (11 of 15) [2/4/2005 1:06:56 PM]

The Instructor

The natural continuation. Black can’t get his

queen back in time, since the a7-pawn is
loose. On 31...Kb7, White sacrifices his
bishop by 32. Bxb5 cb 33. Qxb5+ Kc7 (or
33...Ka8 34. Qc6+ draws) 34. Bd2, and
White’s threats have suddenly become too
dangerous. Well, not too dangerous: after
34...Qxe5 (and a few other moves as well),
White still has to give perpetual check.

After 31...Kb8, the bishop sacrifice no longer

works, since Black can meet 32. Bxb5 with
the intermediate move 32...Bd8. 32. Bxf5, intending to answer 32...gf with 33.
Be3, doesn’t work either: Black refutes the attack by 32...Qh1+ 33. Ka2 Qxf3

34. Be3 d4! 35, Bxd4 Qd5+, maintaining

great winning chances.

However, White could first insert the moves

32. a4 b4, and now, 33. Bxf5 would be very
strong: on 33...Qh1+ 34. Ka2 Qxf3 35. Be3
d4 36. Bxd4, the d5 square is now controlled
by White’s queen. One may conclude that
White’s attack is quite sufficient to draw.

Instead of 32...b4, Tal also looked at the

queen check.

Now 32...Qh1+ 33. Ka2 Qf3 34. ab

34...Qxf4 35. bc (35. b6!? ab 36. Qxb6+ Ka8

37. Qxc6+ Kb8 38. Qb6+ Ka8 39. Bb5+¾) is
risky, for instance: 35...Qb4 36. c7+ Kc8 37.
Ba6+, or 35...Qxe5 36. Qb5+ Kc7 37. Qb7+
Kd6 38. Qb8+ (38. c7).

Unfortunately, this variation contains a

serious hole. In the position from the previous
diagram, instead of the losing bishop capture,
Black plays 34...Qxf2! (preventing 35. b6,
and preparing to bring the queen back) 35. bc
Qb6. He keeps the extra pawn; on the other
hand, after 36. Qa4 =/+, the game’s outcome remains unclear.

And now, a word from Botvinnik:

Here, Black would have won by 31...Kb8 32. Bxf5 (32. Bxb5 Bd8) 32...Qh1+
33. Ka2 Qxf3 34. Be3 d4 (35. Bxd4 Qd5+). Unfortunately, I missed the move
34...d4 in time-pressure (it was pointed out later by Petrosian).

file:///C|/cafe/Dvoretsky/dvoretsky.htm (12 of 15) [2/4/2005 1:06:56 PM]

The Instructor

True, Tal later tried to show that White still obtains a draw by playing 32.
a4, instead of 32. Bxf5. But, without trying to cast doubt on the multiplicity
of complex variations he presented, we note that Black can avoid all this
sophistry, and just force a favorable endgame by 32...Qd8! 33. Qxd8+ (33.
Qa6 Qb6) 33...Bxd8 34. ab Nd4 35. bc Nxf3, with threats of 36...Bc7 and
36...Nd4. He could also play 35...Bc7, for example: 36. c4!? dc! (36...Nxf3 is
inferior: 37. cd ed 38. Bxg6!, or 36...Nxc6 37. cd ed 38. Bb5) 37. Bxc4 (37. Be4?
f5!) 37...Nxc6, with advantage to Black.

Nevertheless, White did have a clear path to the draw, which went unnoticed by
the grandmasters. After 31...Kb8!? 32. Bxf5! Qh1+ 33. Ka2 Qxf3 (see the next-to-
last diagram), White must play, not 34. Be3?, but 34. Bd2!!, and when Black
takes the bishop with his pawn, 35. Bb4!, with perpetual check (and if he takes
the bishop with the queen, 35. Be3!).

31. ... Qh8-h1+ (2.25)

32. Kb1-a2 (2.13) Qh1xf3 (2.28)
33. Qa5-a6+ (2.14) Kc8-b8 (2.29)
34. Qa6xc6! (2.18) Qf3xf4 (2.29)
35. Bd3xb5!? (2.19)

White could have had the draw right here, by playing 35. Qxb5+ Kc7 36.
Qa5+, but he goes fishing in muddy waters instead.

Now White threatens 36. Ba6. Black has but one defense.

35. ... Qf4xe5 (2.29)

36. Qc6-e8+ (2.20) Kb8-b7 (2.29)

36...Kc7 would have been more cautious.

Now, pure chance saves Black.

37. Qe8-c6+ (2.27)

White used up almost all his time calculating

the variations after 37. Ba6+! (the bishop has
to be taken, else Black either gets mated or
loses his queen. For instance: 37...Kc7 38.
Qc8+ Kb6 39. Qb7+ Kc5 40. Qb4+ Kc6 41.
Qb5+ Kc7 42. Qb7+, and Black can choose
either 42...Kd8 43. Qc8 mate, or 42...Kd6 43.
Qb8+). So, 37...Kxa6 38. Qc6+ Kd5 39. c3 threatens mate in two moves, and
39...Bxa3 40. Kxa3 Qd6+ 41. b4+ is no salvation.

Black has only one chance: 39...Qe2, when White has to force perpetual check by
40. Qc7+ Kb5 41. Qb7+ Ka5 42. Qc7+. This was a definite practical chance;
however, White had stood so poorly for so much of the game, that he decided not
to “tempt fate”, and forced the draw.

37. ... Kb7-b8 (2.29) Draw.

file:///C|/cafe/Dvoretsky/dvoretsky.htm (13 of 15) [2/4/2005 1:06:56 PM]

The Instructor

I should like to conclude by showing my readers two more aspects of this battle
we just examined.

1) When analyzing an unfavorably concluded game, chessplayers have a habit of

discovering the last error which had an effect on the outcome, and seizing upon
precisely that error as the cause of their misfortune. So it was here: Mikhail
Botvinnik explained his failure to win this game by a single error he committed
in time-pressure on the 31st move. Such an explanation does not withstand
criticism. And it’s not even that his “error” did not exist: already, there was no
win. And even if there had been one, it’s still impossible to believe that a
grandmaster who had already missed – as we saw – more than one favorable
possibility, should now have solved this precise problem correctly, especially
considering how difficult it was.

When converting an advantage against a stubborn and resourcefully defending

opponent, there always comes a moment when it becomes necessary to switch
from a strictly positional or technical means of operation to accurate calculation
of variations so as to uncover the one correct means of capitalizing on the
advantages of your position. This problem is examined in detail in my lecture
Exploiting an Advantage, contained in the book Technique for the Tournament
Player, by M. Dvoretsky and A. Yusupov.

Botvinnik evidently had programmed himself to reject any and all complications.
Throughout the course of the entire game, he consciously avoided any paths –
even those favorable to him – requiring concrete calculation, and containing a
higher risk of error. As a result, his enormous positional advantage gradually
melted away and finally disappeared.

2) Isn’t it amazing how many purely tactical errors occur in the commentaries
written by Tal – a chess genius (this is not a cliche, but a “medical diagnosis”),
whose chief strength lay precisely in tactics? Now, this fact did not surprise me,
because I made a detailed study of Tal’s work – and shared some of my
observations in my book, School of Chess Excellence 2 – Tactical Play. Note
also, that all his errors were committed in his commentaries – in the game, such
tactical errors on his part almost never occurred. The reasons for his surprisingly
weak treatment of the position in the first half of the game lie in a different field
altogether (even if Tal did consider some of his strategic errors, such as 8. h4?,
the result of miscalculations).

Tal was a chessplayer with a clearly drawn intuitive bent to his thinking. In sharp
positions, he almost unerringly sensed the proper direction in which to search,
what prospects lay down this or the other path. In his head, a multitude of ideas
whirled; he saw lengthy variations in a split-second, with many unexpected,
spectacular points. He saw – but he did not accurately test them – they served
only as guideposts, and inspired his chess forays. When the time came to make a
final decision, and turn this or that previously noted idea into life, Tal would re-
examine them, and as a rule, he found mistakes (from his own annotations, it
follows that such episodes also occurred in the game we have just examined).
Then, he would correct his plans, choosing the optimal path to his goal (which he
usually found, since Tal’s intuition rarely betrayed him; as a rule, he generally

file:///C|/cafe/Dvoretsky/dvoretsky.htm (14 of 15) [2/4/2005 1:06:56 PM]

The Instructor

assessed the position and the overall direction of the game correctly).

Understandably, many of these variations never saw daylight, since his opponent
chose a different path. But they remained in his memory, and then were set down
in Tal’s annotations, without being further tested at the board, with all their
shortcomings. Well, to each his own style of play and commentary. Perhaps Tal’s
texts do contain a lot of mistakes – on the other hand, they give off the true aura
of chess struggle, describe the real feelings of a chess genius, and for that reason,
they remain most unusually engrossing and useful.

[ChessCafe Home Page] [Book Review] [Columnists]

[Endgame Study] [ Skittles Room] [Archives]
[Links] [Online Bookstore] [About ChessCafe] [Contact Us]

Copyright 2005 CyberCafes, LLC.All Rights Reserved.

"The Chess Cafe®" is a registered trademark of Russell Enterprises, Inc.

file:///C|/cafe/Dvoretsky/dvoretsky.htm (15 of 15) [2/4/2005 1:06:56 PM]