Professional Documents
Culture Documents
Jan Timman
First published in 1997 by Gloucester Publishers pic, (formerly Everyman Publishers
pic), Northburgh House, 10 Northburgh Street, London, EC1V OAT
Reprinted 2003
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system or transmitted in any form or by any means, electronic, electtostatic, magnetic
tape, photocopying, recording or otherwise, without prior permission of the J:mblisher.
Distributed in North America by The Globe Pequot Press, P.O Box 480,
246 Goose Lane, Guilford, CT 06437-0480.
All other sales enquiries should be directed to Gloucester Publishers pic, Northburgh
House, 10 Northburgh Street, London, EC1V OAT
tel: 020 7253 7887 fax: 020 7379 4060
email: info@everymanchess.com
website: www .everymanchess.com
Everyman is the registered trade mark of Random House Inc. and is used in this
work under license from Random House Inc.
Editor's Foreword 5
Preface 6
This renowned work by Jan Timman was first published in 1 980 by RHM, but
never reprinted in English due to the subsequent collapse of the publisher.
A second, updated edition was published in 1 993, but this appeared only in
Dutch. The current edition includes all the new analysis which appeared in the
1 993 edition, together with some further minor corrections.
The Art of Chess Analysis remains one of the best examples of painstaking
analytical work ever written. It is very unusual for a leading player such as Jan
Timman to take on the difficult task of giving detailed annotations to the games of
other players. He has succeeded magnificently. The book is particularly instruc
tive in that Timman not only gives detailed analysis when required, but also cov
ers the plans and counterplans available to both sides, illuminating many of the
general principles governing chess strategy.
Petra Nunn
Chertsey, June 1 997
Preface
We must learn to live with Karpov as World Champion. His games are gener
ally less absorbing than Fischer's, but on the other hand, the title has had a bene
ficial effect on Karpov. He has continued to play, and his games have even begun
to show a little more colour. Games 17 and 1 8 are good examples.
My own play underwent a ripening process, and the analysis of games 1 4- 1 8
(game 1 3 was done later) flowed from my pen with great ease. There is a definite
difference between the analysis of games 1 -6 and that of games 14- 1 8.
During the last two years I have published little detailed analysis. But on those
few occasions I received more response than ever before- a happy phenomenon.
Four of the last five games in this book are brand new. I gave them a great deal of
attention and tried for the greatest possible precision. Obviously, imagination
takes its own course, like time and tide. Not only play itself but also analysis must
be fed by inspiration. My hope is that you will feel free to make critical com
ments.
J. H.1imman
Amsterdam, August 1 9, 1 979
Game One
Portisch - Smyslov
Can didates play-off Match (3), Portoroi 1971
Dutch Defence
In chess it is customary to play off ties for reserve places in the candidates
matches. Portisch and Smyslov, who had shared seventh place in the Interzonal
tournament in Palma de Majorca 1 970, played a six-game match which ended in
a tie, 3-3. Since none of the eight candidates withdrew to make room for a reserve,
the match stopped at that point instead of continuing to a decision.
Though it was in effect only a practice match, Portisch-Smyslov left us with
some interesting games, particularly this one. Smyslov played the Leningrad
Variation of the Dutch Defence very unconventionally. Portisch, who is known
for his methodical opening play and rather dogmatic handling of the middlegame,
was evidently thrown into confusion. Instead of striving for a small positional ad
vantage, he entered complications which Smyslov seems to have evaluated better.
1 d4 f5
This in itself is already a surprise.
As far as I know, Smyslov has never B
14 dxc4 16 lL'lc4
15 bxc4 li'lb6! (D) 17 .tel l:.ad8
18 l:tb 1 (D)
After 1 8 exf5 gxf5 1 9 1i'xf5 l:txd4
Black has too many squares; e.g., 20
w l:r.b1 e6 21 1i'c2li'ld5.
1 c4 c6
2 lL:!f3 dS
3 e3 lL:!f6 B
4 lL:!c3 e6
More active is 4 ...lL:!bd7 to answer 5
b3 with 5 . . . e5 . After 4 . . . lDbd7 5 d4 e6
the game follows Slav paths, but with
5 cxd5 cxd5 6 d4 White can try a sort
of Exchange Variation where the black
knight does not stand very well on d7 .
5 b3 lDbd7
6 i.b2 i.d6
7 d4 (D) The correct way to equality. Black
The point of the white set-up. In the wants to answer 9 0-0 with 9 . e5 ,
..
normal S l av opening the fianchetto when 1 0 cxd5 does not work because
of the white queen's bishop is hardly of 10 . . e4. Therefore White's follow
.
cause of Black's defeat. The fact is, with the moves b3-b4 and f2-f4), the
however, that it violates the general position would be won because White
positional rule against placing one's could then play e3-e4 at the right mo
pawns on the same colour squares as ment.
one's bishop. S till, Black would have great draw
19 l:lacl g4 ing chances by playing 28 . . . l0e6 here;
Now this is only an innocuous dem- e . g . , 29 xe6 (otherwise White can
onstration. not make progress : 29 f5 is an
20 l0r3 'ii'b6 swered by 29 . . . 5) 29 . . . fxe6 30 f4
21 l:lxc8 l:lxc8 fl 3 1 f2 h6.
22 l:lcl
White can ignore the ' threat'
22 . . . xe3 because after 23 l:lxc8+
i.xc8 either 24 We 1 or 24 1We5 wins. B
22 l0r6
23 l:lxc8+ i.xc8
24 'ii'c3 .td7
25 4 lDe8
26 a4!
White is going to fix the black
queenside pawns on squares the same
colour as Black's bishop, the result of
Black's eighteenth move.
26
'ii'c7 28
... ?
Black can prevent the fixing of his 29 n
queenside pawns with 26 . . . a5, but the The immediate 29 e4 rJ/;e7 ! prom
cure seems worse than the disease be ised little.
cause after 27 i.b5 i.xb5 28 xb5 29
e7
both White pieces have optimal possi Again, Black should try 29 . . . h6 fol
bilities. lowed by 30 . . . e6.
27 'ii'xc7 xc7 30 e2 g6
28 aS (D) Another pawn on the wrong colour,
Black now has not only the weak but this was difficult to avoid because
ness on d5 but also a more serious if 30 . . . 6 3 1 f5+.
weakness on b7 . If the knights were 31 d2 lDe6 (D)
not on the board and White's king 32 e6?!
could reach d4 (which should be pos I think a better idea is 32 c3 c5
sible because White could keep the 33 i.e2 ! (not 33 f3 d3 34 xd3 d6
black king off the squares e5 and c5 35 b4 g5 ! and the winning chances are
Lev Polugaevsky - Henrique Mecking 17
1 4 ltd 1 1Wc4 1 S 'ti'dS 'ti'xdS 1 6 ltxdS 1 8 ltb2 ! with some advantage for
b6 with advantage to Black (Mititelu White (Taimanov-Fischer, Vancouver
Hort, Luhaovice 1 97 1 ). 1 97 1 ).
2b) 6 ... lLlc6 ! ? 7 cxdS lLlxc3 8 bxc3 8... .i.g7
'ti'xdS 9 'ti'f3 'ti'xf3 10 lLlxf3 cxd4 1 1 9 lLlf3
cxd4 e6 and Black gets the same sort Very interesting positions can arise
of play as in variation 1 . In this case 7 after 9 ltb1 .i.dS and now:
lLlf3 seems the correct method. I) 1 0 f3 fS l l lLlh3 h6 1 2 l0f4 gS
2c) 6 . . . 1i'aS 7 tli'b3 cxd4 (stronger 13 lLlxdS "it'xdS 14 i.. f2 or 1 4 'ti'a4+
than 7 . . . lLlc6 8 l0f3 cxd4 9 exd4 lLlxc3 'ti'd7 1 5 'ti'xd7+ lLlxd7 1 6 .i.g3 0-0-0
10 bxc3 .i.e6 1 1 .i.e2 .i.g7 1 2 0-0 0-0 17 i.. xc4 eS ( 1 8 i.. e6 lthe8 !) and
1 3 cS with the better chances for Black does not stand badly in either
White, Taimanov-Filip, Wijk aan Zee case.
1 970) 8 exd4 .i.h6 ! 9 ltd 1 (9 lLlf3? gS ! 2) 1 0 .i.f3 c5 ! ? ( 1 0 . . . 0-0 1 1 e4
1 0 .i.g3 g4) 9 . . . 0-0 1 0 cxdS . So far i.. c 6 1 2 lLle2 is good for White) 1 1
Donner-G. Garcia, Cienfuegos 1 973, 'ti'a4+ .i.c6 1 2 .i.xc6+ lLlxc6 1 3 l::tx b7
and now, according to Donner, 10 ... lLld7 'ti'c8 and now 14 ltxe7 f8 1S lte4
1 1 .i.d3 lLlxc3 1 2 bxc3 lLlb6 1 3 lLle2 leads to intricate complications which
'ti'xdS leads to an equal game. This ap are probably not unfavourable for
proach by Black is probably the main Black.
reason many players with White now 9. 0-0
play 4 l0f3 .i.g7 before continuing 10 0-0 c5
with S .i.gS . Perhaps too sharp. If Black wants to
6 bxc3 dxc4 keep the pawn he can also try 10 ...c6 1 1
7 e3 i.e6 lOgS bS . White would then have to de
8 .i.e2 cide whether to capture on e6 or to
Two games with Fischer as Black bring his knight later to cS via e4.
continued with 8 ltb1 b6: 11 lOgS! (D)
1 ) 9 l0f3 .i.g7 10 lLld2 0-0 1 1 This leads forcibly to advantageous
lLlxc4 i.. d 5 1 2 'ti'd2 1i'd7 1 3 l0a3 cS play for White.
1 4 f3 'ti'a4 1 S lLlb5 with equal chances 11 i..d5
(Mecking-Fischer, Buenos Aires 12 e4 i..c6
1 970). The alternative is 1 2 . . h6. In the
.
22 h4 :ads
The players strengthen their posi
tions in the prescribed manner. w
23 hS
Probably an inaccuracy, which, how
ever, is not taken advantage of. With
23 a5 ! White can prevent Black's pos
sible knight manoeuvre and thus also
his counterplay.
23.. gxhS
The tendency to take such pawns
has been seen in Portisch's games be
fore. This time he misses the chance, White can win here in an intricate
prepared by his last two moves, to fur manner with 27 i.c7 ! . Here is the
ther undermine the configuration e5- analysis:
d6 with the strong 23 . . . ltlb6 ! 24 hxg6 1) 27 . . . i.xf3 28 exf6 'ft'xc4 (or
ltld5 ! (not 24 . . . hxg6 25 ltlh4 ltld5 26 28 . . . i.xd 1 29 i.xe6+) 29 'ft'xf3 and
Wg3 g5 27 ltlg6 ! ) 25 gxh7+ h8 26 then continuing as in the game.
We 1 (White must keep c3 defended) 2) 27 . . . ltlxe5 28 :xd8 ltlxc4 29
26 . . . :xd6 ! 27 exd6 f5 and Black has 'it'xe6+, etc.
good play in the centre for the ex 3) 27 . . . 'ft'xc4 28 :xd7 ' with a de
change. cisive attack' , I wrote in the frrst edi
24 :d2 tion of this book. S ome further
Now White has some control over explanation is called for. After the
the game again because 24 . . . ltlb6 is continuation 28 . . . :xd7 29 :xd7 'it'e4
bad on account of 25 exf6 ltld5 26 White has the crushing 30 :xg7+ !
Wg5 . xg7 3 1 exf6+ and the black king has
24 ... b6 no good escape.
25 :adl i. b7 27 ... 1Wxc4
One c an understand that Black is 28 i.e7 .i..xf3?
losing patience. Whatever he does, his Black will finally come out of the
position will worsen. The pressure on complications an exchange behind
e5 must not be eased, so his knight and without compensation. 28 . . . 'it'g4 !
must remain on d7. As a result, the keeps the struggle alive; e.g., 29 i.xd8
squares e7 and c7 must remain pro and now:
tected . I cannot give a reasonable al 1) 29 . . . ltlxf6 30 ltle1 Wg6 3 1 i.xf6
ternative to the text-move. 'it'xf6 with rough equality.
26 .i..xc4 ..,xa4 (D) 2) 29 . . . .i.. x f3 . This move, which I
27 exf6 gave in my original notes, loses on the
Svetozar Gligoric - Lajos Portisch 25
6 'iWg4, at first with success against White does not have to defend the
Uhlmann, then with catastrophic con pawn: 1 0 0-0 c4 1 1 i.e2 \i'xc3 1 2
sequences against Kovaevic . Larsen i.d2 'iWb2 1 3 l:lb1 followed by 1 4
Bobby Fischer - Bent lArsen 27
l:xb7 and White has a dangerous in Again Fischer has the chance to
itiative. capture on f6, but he does not concern
10 0-0 c4 himself with this possibility.
In combination with the next move, 13 ..t a3
this plan carries great danger. The al According to Byrne, 1 3 g3 would
ternative is 1 0 . . . h6 followed by cas also give White a smal l advantage.
tling short and only then, perhaps, to But in that case Black could accept the
aim for . . . f7-f6, recapturing on f6 with pawn offer without too many prob
the rook. lems: 1 3 .. .fxe5 14 dxe5 xeS 15 ltlxe5
1 1 .tel f6 (D) 'ii'xe5 1 6 ..tg4 'ii'x c3 1 7 ..txe6 0-0-0
Obviously, Larsen had studied his and the position is far from clear.
opponent's rare losses well. Fischer 13 fxeS
lost a long game against Mednis in the 14 dxeS es
1 9 60/6 1 U . S . Championship after 1 2 15 eS xeS
..ta3 0-0 ! 1 3 l:e1 l:f7 1 4 exf6 gxf6 1 5 Capturing with the queen is hardly
..t n l:e8 1 6 h4 g6 1 7 '6'h5 l:g7 to be considered: 15 . . . 'ii' xe5 16 ..txc4
1 8 g3 'ii'a5 19 ..tb2 d8 20 l:e3 f7 'ii' x c3 17 ..txd5 0-0-0 1 8 ..tb3 and
2 1 h 1 d6 and B lack already had Black has no compensation at all for
the initiative. the pair of bishops.
16 '6'd4 (D)
w
B
12 l:e1 !
An important improvement. If The queen keeps an eye on both
Black now castles short, White can flanks from here. Larsen, in his notes
capture on f6 and answer the pawn re to this game, points out an interesting
capture with ..th6. Black must there opening finesse. If this position had
fore play to win the e5-pawn. come about in a slightly different way
12 ... g6 (the way that has repeatedly occurred
28 The Art of Chess Analysis
24 1i'xd7 l:.ad8
Centralising the heavy artillery to
w the utmost.
25 'ii'xb7 (D)
28 ... llfl+
29 'iPgl llxg2+
Making the best of it.
30 'iPxg2 'ifd2+
31 'iPhl l:.xc6
32 i.. xc6 (D)
1 d4 c!Df6
2 c4 cS
3 dS g6 B
4 c!Dc3 d6
s e4 i. g7
6 c!Df3 0-0
7 i.e2 e6
8 0-0 exdS
9 cxdS a6
The alternative 9 . . l:te8 is more cur
.
he does this with 21 Afc 1 , to simulta Also dubious: the queen becomes
neously free fl for his knight, there tied to the protection of the knight.
follows, for instance, 2 1 . . .ltlf8 22 ltlfl One gets the impression that Stein was
ltlfg6 23 ltle3 ltlf4 24 ltlf5 Ad7 and in time trouble. Still good is 26 . . . 'il'e7
Black stands satisfactorily. The imme 27 Ab6 Ad8 , after which there is little
diate f2-f4 may also be considered, wrong with the black position.
however. 27 Ab6 c4 ?!
21 liJr8 (D) 28 'il'a4!
Threatens to take on d6. The black
pieces are in one another's way.
28 :rs
w 29 'il'a3 AcS
30 .tfl
30 Axb7 ltlxg2 (30 . . . Axa5 3 1
.txh4) 3 1 xg2 Axa5 gives Black
counterplay.
30 AbS
31 Axd6 'it'e7
3 1 . . .1lg5 is no better, because of 32
.te3 1fe7 3 3 .txh6. Now, however,
the knight on h4 still hangs.
22 f4! 32 Ab6 'ii'xa3
Now or never. 33 bxa3 ltlxgl (D)
22 gxf4
23 Axf4 ltlfg6
24 :n
The alternative idea 24 Af5 , so as W
possibly to play the rook to h5 later,
springs to mind. A double-edged busi-
ness.
24 'iVd8!
A strong move, the prelude to a
regrouping.
25 A a3 Ac7?!
More accurate is 25 . . . Ad7 to over
protect d6, followed by 26 . . . 'il'e7 and
27 . . . Ac8 . The reason becomes clear in 34 Axb5?
the next moves. A mistake which hands the initia
26 Ab3 ltlh4 tive over to Black. After 34 xg2
36 The A rt of Chess Analysis
B
w
i s not sufficient) and now, besides 20 4b) 1 7 lLle4 (D). Black then has
1i'xc2 l:xc4 2 1 dxe6, which leads to three possibilities:
better play for Black after 2 l . . .'ili'xc3 !
22 Wxc3 l:xc3 ! 23 exf7 l:f8, White
can offer the queen with 20 dxe6.
Langeweg judges that Black runs no
risk with 20 . . . l:xd3 2 1 l:xd3 'ili'xa3,
but I cannot agree, because after 22
exf7 , 22 . . . l:xc4 fails to 23 l:g 1 , and
otherwise a white rook gains control
of the g-file; for example, 22 . . . .!Od4 23
l:g 1 l:xg 1 + 24 'it>xg 1 lLlf5 (to stop
l:g3) 25 e6+.
4a3) 1 8 . . . g4 ! . Although not men
tioned by anybody, this move is very
strong. The main variation runs: 1 9 4b 1 ) 1 7 ... l:tg6 18 'ti'xa5 lLlxa5 1 9
'it>xg2 f3+ 20 'it>h3 (otherwise he dxe6 lLlxc4 20 exf7 :.f8 2 1 l:d5 b6 and
loses the rook with a lost position) White stands a little better (Korchnoi
20 . . . 1i'c7 ! 2 1 l:xd4 (mate in two was and Furman).
threatened) 2 I . . .Wd7+ and now: 4b2) 17 . . . g4 18 'ii'xa5 (if 1 8 fl
4a3 1 ) 22 h4 'ili'f5 ! ! , threatening f3 ! ) 1 8 . . . lLlxa5 19 e2 xe2 20
23 . . . Wh5+ and 24 . . . l:g8 and mate. xe2 l:g6 21 lLlg3 e4 22 h4 with
B arendregt drew my attention to 23 chances for both sides.
'ili'h6 which, it is true, does avert the 4b3) 17 . . . 1i'b6 ! 1 8 1i'e3 ( 1 8 'ii'c 3
mate, but after 23 . . . exd4 24 l:g 1 dxc3 f5 19 fl xe4 20 'ii'h 3+ c;tb8 2 1
the black attack continues despite the xg2 xc2 and Black wins) and now
restored material equality. Black can get a decisive endgame ad-
4a32) 22 g3 xh 1 23 l:g4 f5 24 vantage with 1 8 . . . 1i'xe3+ 1 9 fxe3 g4
'ti'c5+ b8 25 l:g7 (25 l:g5 f6) followed by 20 . . . f3. Also 1 8 . . . .!0d4
25 .. Jlc8 and White cannot hold on to 1 9 fl g4 20 xg2 f3 + 2 1 'ti'xf3
his extra material; e.g., 26 1i'b4 f4+ 27 lLlxf3 22 xf3 f5 is good because
h4 1i'd8+ 28 h3 1i'f6 29 l:g4 Korchnoi and Furman 's suggested
1i'h6+ 30 l:h4 1i'g6 3 1 l:g4 11t'h5+ 32 23 tLld2 is met by 23 . . . 1i'd4, while the
l:th4 11t'f3 mate. more natural 23 lLlg3 is strongly an
White must therefore play 1 9 e2 swered by 23 . . . 11t'g6.
or lDe2, but in either case 19 . . . lLlf5 is I hope I have shown with these
strong, perhaps too strong. lines that Black could have obtained a
My conclusion is that 17 11t'e3 must winning position by 1 6 . . . l:xg2, which
be rejected and that White must play: was indeed quietly assumed by others.
42 The Art of Chess Analysis
I
The move played by Petrosian is 20 ... 1i'xa3
not bad either. 20 . . . l:lc8 is an important alternative.
16
.i.rs White gets the advantage after 2 1 lL!e4
17 .i.d3 .i.xd3 1Wxd5 22 c3 f5 23 .!Llg3 f4 24 cxd4
17 . . . l0d4 leads to the same position fxg3 25 fxg3. O' Kelly, however, points
after 1 8 .i.xf5 .!Llxf5 1 9 1i'd3 ( 1 9 0-0 out the exchange sacrifice 22 . . . l:lxc3
.!Llh4) 19 . . . l0d4, but Black can also try 23 1Wxc3 when Black has just enough
1 9 . . . .!Lld6. It is strange that Petrosian compensation.
limits his choices like this. 21 f4
1 7 ... e4 is a whole chapter in itself. In Schaakbulletin 47, 2 1 l'Oe4 is
White can react as follows : correctly given as better. However, the
1 ) 1 8 .i.xe4? .i.xe4 1 9 .!Llxe4 .l:ge8 possibility 2 1 . . .11ha2 is not men
20 0-0 l:txe4 21 'ikd3 f5 and Black tioned: 22 lLlxf6 Ag6 23 'ife4.
wins. 21
.. .l:c8
2) 1 8 .!Llxe4 .i.xe4 19 dxc6 fleS ! . 22 .!Lle4 11Vxd3
3 ) 1 8 .i.e2 ! l:lxg2 1 9 'it'e3 .!Lle5 20 22 . . . "ii'x a2 leads to nice continu
fl . White stands better, according to ations such as 23 .!Llxf6 l:lxg2 ! Then
the Deutsche Schachzeitung, because 24 .!Lld7+ a7 25 fxe5 l:lcxc2 26
2 1 .!Llxe4 is possible when the black 'it'xd4+ leads to the beautiful king ma
rook withdraws. However, Black can noeuvre 26 . . . a8 27 .!Llb6+ b8 28
play 20 . . . .!Llg4 2 1 .i.xg4 l:lxg4 22 h3 .!Lld7+ c8 29 .!Llb6+ d8 30 Wh4+
l:lh4 with a difficult position for White e8 and Black wins. White must
(23 'it'g3 l:th5 or 23 l:ld4 'it'c5 !). therefore play 23 l:ld2 ! ( Korchnoi
18 'it'xd3 .!Lld4 and Furman), after which Black has a
19 0-0 b8 draw with 23 . . . l:lxc2 24 l:lxc2 Wxc2
20 h1 (D) (24 . . . .!Llxc2 25 We2) 25 Wxc2 .!Llxc2
26 .!Llxf6 .!Lle3 ! 27 l:le 1 (27 l:lf3 l:lc8 ! )
2 7 . . . .!Llxg2 28 l:lg 1 l:lg6 29 .!Lld7+ c7
30 .!Llxe5 .!Llxf4 3 1 .!Llxg6 fxg6, as
shown by Kholmov. Perhaps 26 fxe5
is worth trying as a winning attempt.
23 cxd3 (D)
This has not been commented on
anywhere. Yet 23 l:lxd3, with the idea
of attacking the knight's position, is
interesting. A drawn position arises af
ter the long, practically forced con
tinuation 23 . . . l:lxc2 24 g3 l:lxa2 ! 25
.!Llxf6 l:lc8 26 fxe5 l:lcc2 27 .!Llg4 h5 28
Bobby Fischer - Tigran Petros ian 43
lbfl ! (28 lbe3 llxh2+ 29 ci>gl llh3) is happening: 29 d4 llxd5 and though
28 ... llxf2 29 llxfl llxfl 30 llxd4 llf5 ! the black kingside pawns are indeed
(3 1 d6 c;t.>cs 32 llc4+ ci>d8). weak, if White attacks them Black will
23
llc2 get the c-file. Furthermore, Black has
24 lld2 the advantage on the queenside. White
Now 24 g3 is bad due to 24 . . . llxa2 should occupy d6 with his knight by
25 lbxf6 llc8, etc. playing 30 llc5 lidS 3 1 lbc4.
24
:Xd2 28
lbd4
2s lbxd2 rs 28 . . . lbb4 was suggested here, with
S harply seen. Examine also these the idea 29 lbc4 lbxd3 30 e6 fxe6 3 1
lines : lbd6 lle7 32 llxf5? llf7. Better seems
1 ) 25 . . . lle8? 26 f5, etc. 30 g3 and White stands a little better
2) 25 . . . lld8 26 fxe5 fxe5 27 llxf7 . (30 . . . b5 3 1 lbaS).
3) 25 . . . llc8 ! ? 26 fxe5 llc2 (or 29 lle3 lbc2
26 . . . fxe5 27 llxf7 llc2 28 lbfl llxa2 30 llh3
29 g4 ! ) 27 lbe4 fxe5 28 g4 ! with Naturally.
slightly better play for White. 30
llxeS
4) 25 . . . exf4 26 llxf4 lle8 27 llxf6 31 lbr3 llxdS
with advantage to White. Disapproved of by Kholmov. He
26 fxeS lle8 gives 3 1 . . .lle2 as correct, with the
27 llel lbc2 (D) variation 32 llxh7 lbd4 33 llxf7 llxa2
28 lle2 34 h4 lbxf3 35 gxf3 lla4 ! , drawing .
28 llc 1 is recommended by Panov. But 33 llxf7 is ridiculous and must be
The luminous point is 28 . . . llxe5 ? 29 replaced with the immediate 33 h4 .
lbf3 lle2 30 d6 ! ci>c8 3 1 ci>g l ! ci>d7 32 Then the black f-pawns only make it
ci>fl winning a piece. After 28 . . . lbb4 more difficult to stop the h-pawn.
(28 . . . lbd4? 29 lbc4 b5 30 lbb6) little 32 llxh7 (D)
44 The Art of Chess Analysis
practical point of view: i n many lines gives him the advantage in the end
White has the possibility of sacrificing game.
on f7, and the knight also has the pos 24 h3
sibility of returning to the defence via Although not as bad as was gener
e4. On the other hand, the strongpoint ally thought, this move is clearly an
on d6 can become shaky, as will be example of superficial calculation.
come apparent. Other moves:
22 i.a8 1) 24 .l:.d3 . This is refuted simply
23 i.c4 with 24 . . . .ixg2+ 25 ...xg2 1Wc l +,
Fischer must have played this winning the exchange.
strong defensive move, which protects 2) 24 ltle4 1Wxe5 25 ltlxc5 1Wxc5
the queen on e2 and in general brings 26 h3 (D).
the bishop back into the game, purely
by intuition. Attempts to re-exert pres
sure against f1 fail: 23 .l:.fl h4 24 lilxf7
h3 ! (even stronger than 24 . . . l:txt7 25 8
practical point of view: i n many lines gives him the advantage in the end
White has the possibility of sacrificing game.
on f7, and the knight also has the pos 24 h3
sibility of returning to the defence via Although not as bad as was gener
e4. On the other hand, the strongpoint ally thought, this move is clearly an
on d6 can become shaky, as will be example of superficial calculation.
come apparent. Other moves:
22 i.a8 1) 24 .l:.d3 . This is refuted simply
23 i.c4 with 24 . . . .ixg2+ 25 ...xg2 1Wc l +,
Fischer must have played this winning the exchange.
strong defensive move, which protects 2) 24 ltle4 1Wxe5 25 ltlxc5 1Wxc5
the queen on e2 and in general brings 26 h3 (D).
the bishop back into the game, purely
by intuition. Attempts to re-exert pres
sure against f1 fail: 23 .l:.fl h4 24 lilxf7
h3 ! (even stronger than 24 . . . l:txt7 25 8
together.
26 'ii'xh4 (D)
For the moment, White holds his
extra pawn. 26 ltlxt7 is again incorrect
since after 26 . . . xf7 27 .txe6+ f6 !
(not 27 . . .'ti'xe6 28 l:.d7+) 28 l:.fl +
e7 White has nothing for the piece.
26 gS
With gain of tempo, this frees a
square for the king so that the rook The bishop cannot be taken, but af
may be used for an attack along the ter 28 h2 1i'xc2 29 1i'h8+ (the white
half-open h-file. On the other hand, knight is in the way ! ) 29 . . . xf7 30
the position of the black king is weak l:.d7+ e8 3 1 l:.d8+ xd8 32 1i'xf8+
ened, which White can exploit. Also, c7 White has no perpetual check and
Fischer - Spassky 53
the black king escapes to the queen White has the opportunity to play the
side. knight back to c4 and to further neu
What better move does White have tralise the position with .if3. As Black
after 26 . . . 'ifxb2? The attacking at would have pawns on only one wing,
tempt 27 .td3 is easily brushed aside his winning chances would be limited.
with 27 . . . .th6. The best is 27 .tb3 !, 28 g7
defending c2, maintaining the threat Threatening the decisive . . . l:.f8-h8-
28 lbxf7, and at the same time allow h4 . The knight must return to the de
ing the knight to spring to c4 . How fence.
ever, B lack has regained his pawn and 29 lbd4 .:th8 (D)
can keep a solid positional advantage
with 27 . . . g5 .
27 'ifg4 .tcS
Black correctly saw that after w
27 . . . .:d8 White could force a draw
with 28 lbxf7 . After 28 . . . Axd 1 + 29
1kxd 1 Black can try:
1 ) 28 . . . xf7 29 1kd7+ with an im
mediate draw by perpetual check.
2) 28 ... 1kg3?? 29 lbh6+. All(!) com
mentators thought that White had a
perpetual check here - all except Don
ner, who after initially making the
same mistake, discovered that Black The first reports and analyses from
gets mated after 29 . . . g7 30 'ifd7+. Reykjavik all mentioned that Spassky
3) 28 ... 'ffe4 ! 29 .tn xfl 30 'ifd7+ had missed a win here. Not with
f6 3 1 'ifd8+ e5 32 'it'e7, and de 29 . . . .id6, when 30 lbf5+ draws at
spite everything, Black cannot avoid once. The important alternative is
a draw. 29 . . . .:td8. Now 30 lbxe6 fxe6 3 1 .:txd8
28 lbb5 1ke 1 + and 30 lbf5+ f6 ! don ' t work,
Fischer again takes his knight out so the knight must be defended. After
of play and thereby lands in a hope 30 c3 Black has these possibilities:
less situation. Instead, he can use the 1) 30 . . . .:th8 (with the thought that
weakened position of the black king now White can't exchange queens on
by showing that the knight was not re c3, as he can in the game) and now:
ally threatened after all : 28 b4 ! . After 1 a) 3 1 .id3 . The intention is to
28 . . . axb4 (the point is that 28 . . . .txd6 meet 3 1 . . . .:th4 by the constantly re
29 .:txd6 'it'xd6 30 'it'xg5+ leads to curring 32 lbf5+. However, Black sim
perpetual check) 29 axb4 .txb4 30 .te2 ply plays 3 1 . . .g8 (Nei's suggestion
54 The Art of Chess Analysis
of 3 l . . .i.b6 is also good, but 3 1 . . . f8 37 'iff4+ 'it'g5 38 1Vf3 with good win
is weaker because of 32 llfl with ning chances for Black.
counterattack) and the threat 32 . . . l:th4 2) 30 . . . i.d6 3 1 g 1 1i'e3+ 32 fl
is even stronger than before. (after 32 h 1 :th8 33 llJf3 i. f4 34
1 b) 3 1 llfl . First given by Smys :d7 g6 the threat 35 . . . f5 is deadly)
lov in 64. The idea is the same as in 32 ... i.g3 33 1i'e2 1Vf4+ 34 g1 i.h2+
variation 1 a but the execution is more 35 h 1 :h8 and now not 36 :n
refined. It is nevertheless hardly suffi :xh3 37 :xf4 i.xf4+ 38 g1 i.e3+
cient after 3 1 . . .l:th4 32 llJf5+ 'ilxf5 33 and wins, but 36 llJxe6 fxe6 37 :d7+.
l:[xf5 l:txg4 34 :xeS :xg2 35 :xaS Here the weakened position of the
(the showy 35 i.d5 leads to a lost black king again plays a role.
pawn endgame after 35 . . . i.xd5 36 3) 30 . . . 1i'e3 . Given by Donner.
.z:r.xd5 exd5 37 xg2 a4 !) 35 . . . i.f3 36 Black keeps the more direct attempts
b3 (Byrne's 36 i.fl :xb2+ is cer in reserve and maintains his grip on
tainly not better) 36 . . . g6 and pene the position. It is difficult to find a de
tration by the black king cannot be fence to the threatened . . . :d8-h8-h4.
stopped (37 :a7 f6 !). White can just survive, however, by
1c) 3 1 i.xe6 ! i.xd4 3 2 cxd4 (this playing 3 1 i.e2 with the neutralising
is where the advantage of c2-c3 shows) threat 3 2 i.f3 . After 3 l . . .i.xd4 32
32 . . . 1i'xe6 33 1i'xg5+ 'ii'g 6 34 'iie 5+ :xd4 f5 ! 33 'ii'h 5 :xd4 34 cxd4 g4
(D) and although three pawns are in White is again obliged to offer a piece:
sufficient compensation for a piece in 35 hxg4 'ii'xe2 36 'it'g5+ f7 37 gxf5 .
this position, an eventual win for His drawing chances are better here
Black is problematical. than in variation 1c.
All things considered, there is no
reason to fault Spassky's continuation.
His mistake comes two moves later.
8 30 llJf3 i.xf3
30 . . . 1Vf4 is an interesting try. After
3 1 1i'xf4 gxf4 32 i.e2 Black has more
than sufficient compensation for the
pawn, and 3 1 1i'xg5+ 'ii'xg5 32 llJxg5
seems no more attractive because after
driving the knight back Black can cap
ture on h3 with great force. However,
White still has a way out:
1 ) 32 . . . :th5 33 llJxe6+ fxe6 34
Olafsson now gives 34 . . . h7 . Nei i.xe6 and White has four pawns for
gives 34 . . . f6 35 'ifc7+ h6 36 d5 :tg8 the piece while h3 is defended.
Fischer - Spassky 55
1 e4 eS
For the first time in the match - and
not unexpectedly - Spassky plays an B
' open' defence. Just as he usually re-
plies to 1 d4 with an orthodox defence,
so he often answers 1 e4 with l . . .e5 .
2 lLlf3 lLlc6
3 i.bS a6
4 i.a4
I n game sixteen of this match,
Fischer reverted to the Exchange Vari
ation with which he had beaten Por
tisch, Jimenez, and Gligori quickly 7 d6
and convincingly in the 1 966 Havana In S anta Monica 1 966, Spassky
Olympiad. prepared the way for the Marshall At
4 lLlf6 tack against Fischer with 7 . . . 0-0. The
5 0-0 i.e7 result was a draw after 35 moves: 8 c3
6 l:tel bS d5 9 exd5 lLlxd5 10 lLlxe5 lLlxe5 1 1
7 i.b3 (D) l:txe5 c6 1 2 g3 lLlf6 1 3 d4 i.d6 1 4 l:tel
Fischer - Spassky 57
reason for trying something else this leads to difficulties after 14 lBb3 axb4
time. The then World Champion at 15 cxb4 l:ab8 16 lBaS c6 (or 16 . . ..i.a8
tained comfortable equality with Black 17 d5 ) 17 lBxb7 l:xb7 18 .i.b3 and
in that game: 1 l . . .exd4 12 cxd4 lBb6 White has gained the pair of bishops at
1 3 lBd2 cS 14 .i.c2 cxd4 15 lBhf3 l:e8 no cost (Suetin-Tringov, Titovo Uzice
1 6 lBxd4 .i.f8 17 b3 .i.b7, etc. 1966).
Fischer also once tried the other di
rect method against the Breyer, 1 1 c4,
versus Portisch in S anta Monica 1 966.
It was a gripping game in all its
phases, though it finally ended in a
draw. After 1 1 c4 c6 1 2 cS 1i'c7 1 3
cxd6 .i.xd6 1 4 .i.g5 exd4 1 5 .i.xf6
gxf6 1 6 1i'xd4 lBe5 1 7 lBbd2 :d8 1 8
1i'e3 3 ! 1 9 'ikh6 .i.f4 20 1fxf6 dif
ficult complications arose.
11 .i.b7
12 .i. c2 (D)
12 l:e8
13 b4 14 a4
58 The Art of Chess Analysis
After this game, this immediate at l:e4 24 'ii'f3+ lLlf6 25 j,xf6 l:r.el + and
tack became fashionable. Formerly, Black just managed to save himself
White always aimed for c3-c4. (Geller-Portisch 1 973).
14 lLlb6 15 aS lLlbd7
And this is currently the most com 16 j,b2
mon reply. Black tempts the a-pawn to In Savon-Vogt, Skopje 1 972, White
advance farther and will later aim for let his opening advantage slip away
. . . c7-c5 . Alternatives are: entirely with 16 l:r.bl d5 ! . White must
1 ) 1 4 . . . c5 1 5 bxc5 exd4 1 6 cxd4 pay attention first of all to the centre.
dxc5 and White stands better, though 16
... 'it'b8
it is not entirely clear (Balashov-Pod Black prepares . . . c7-c5 by indi
gaets, Moscow 1 966). rectly protecting the e-pawn. This
2) 14 . . . d5 . In principle Black's does not work badly here, but later ex
pieces are ideally placed to justify this perience brought Spassky around to
advance. After 1 5 dxe5 lLlxe4 1 6 lLlxe4 playing 16 . . . l:r.b8 . A sharp struggle de
dxe4 1 7 j,g5 ! f6 1 8 j,xe4 j,xe4 1 9 veloped against Planinc (Amsterdam
l:r.xe4 lLlxe5 2 0 l:r.d4 lLlxf3+ 2 1 'ifxf3 1 973) after some preparatory ma
'ii'c 8 White has not achieved much noeuvring: 17 l:r.bl j,a8 1 8 j,at g6 1 9
(Vasiukov-Zuidema, Wijk aan Zee c4 exd4 2 0 cxb5 axb5 2 1 lLlxd4 d 5 22
1 973). More enterprising is 1 5 lLlxe5 lLl4f3 dxe4 23 lLlg5 e3 ! 24 i.b3 and
dxe4 1 6 f4 as in Vasiukov- Kholmov, now, instead of 24 . . . j,d5, Black could
Dubna 1 97 3 . White stood a tiny bit have played 24 . . . exf2+ 25 xf2 j,dS,
better after 16 . . . exf3 1 7 lLldxf3 lLlxe5 as given by Keene.
1 8 lLlxe5 j,d6 1 9 j,f4. This variation does not seem unfa
3) 1 4 . . . a5 . This other way of form vourable for Black. Nevertheless, Smej
ing a square of pawns seems strange at kal varied on the 1 9th move against
first sight. It was popular for a while Browne at Wijk aan Zee 1 976 with
until it was discovered that 15 bxa5 19 . . . bxc4 . White had some advantage
l:r.xa5 1 6 l:r.b 1 leads to advantage for after 20 dxe5 lLlxe5 2 1 lLlxe5 dxe5 22
White in all variations; e.g . : i.c3 j,c6 23 'ife2 j,b5 24 lLlxc4 c5 25
3a) 1 6 . . . j,a6 1 7 d 5 'ifa8 1 8 j,a3 c5 bxc5 j, xc5 26 i.b3. Six months later,
1 9 dxc6 'ifxc6 20 j,b4 (Belyavsky-A. Karpov reached full equality against
Petrosian 1 973). Browne, in Amsterdam 1 976, with
3b) 1 6 ... 'ii'a 8 1 7 axb5 exd4 1 8 25 . . . l:r.c8 ! - a theoretical novelty on
cxd4 lLlxe4 1 9 lLlxe4 j,xe4 20 l:r.xe4 ! the 25th move - with the idea of cap
l:r.xe4 2 1 lLlg5 l:r.h4 and now 22 g3 ! is turing on c5 with the rook.
the correct reply to maintain the ad 17 l:r.b1 (D)
vantage, according to Geller. Instead, A very logical move . White covers
White played 22 lLlxf7 xf7 23 i.g5 b4 and at the same time places his
Fischer - Spassky 59
20 lDxeS
In 64, Polugaevsky recommends 20
B c4, which would achieve excellent re-
sults after 20 . . . i.d6 2 1 lDh4 . Better,
however, is 20 . . . lDxf3+ 2 1 1Wxf3 l:r.e6
and Black does not stand badly.
20 'ilxeS
21 c4 'iff4
22 i.xf6 (D)
2 6 'ii'f l an d Black does not get enough here. Black has great influence over
c ompensation for the exchange) 25 the whole board because of his control
fxg7 (or 25 .te4 'ii' x e l + 26 l:r.xe l of the dark squares .
.i. xe4 27 l:xe4 l:d2 28 .i.e5 b4 ! and 23 :ec1s
Black can hold the position thanks to It was clear which rook had to be
the finesse 29 fxg7 l:r.d 1 + 30 <t>h2 moved, because after 23 . . . l:ad8 24
.i.d6) 25 . . . 'ifxe l + (with queens on the 'ifc l 'ii'c 3 25 bxa6 .txa6 White has
board, the black king's shattered posi the riposte 26 .i.a4 ! with great advan
tion would be a factor) 26 l:r.xel .i.xg7 tage.
27 .i. xg7 <J;xg7 28 l:e7 l:d2 ! and Polugaevsky, however, thinks that
Black can just hold the balance. 23 . . . axb5 is Black's best. After 24
2) 22 . . . l:ed8 23 l:r.e3 ! (naturally, 23 llxb5 .i.a6 White has two rook moves:
exf6 is now pointless) 23 . . . lDe8 24 1) 25 l:b6 'ii'c 3 and White is in big
'ife2 and White has a great advantage. trouble. On 26 lL!b3 there can follow
3) 22 . . . lLld7 and now both 23 .i.e4 Polugaevsky's recommended 26 . . . c4
.i. xe4 24 l:xe4 'iff5 25 'iWe2 and 23 or the strong 26 . . . lled8 27 'ii'c l c4 28
lLlf3 llad8 24 .t e l 1Wxc4 25 .i.b3 are l:r.e3 'ii'e 5 . The attempt to keep the po
favourable for White. sition in balance by 26 l:r.b3 'ii'xa5 27
It is clear that Black must look in .id3 fails to 27 . . . lled8 (28 'ii'c 2 c4 or
variation 1 for any chance to maintain 28 'ife2 l:txd3 29 l:txd3 l:td8).
the balance. It is much easier for him 2) 25 l:tb3 . Surprisingly enough,
after the text-move. this move holds the white position to
22 'ifxf6 gether. Black does indeed win a pawn
23 cxbS (D) after 25 . . . c4 26 l:tf3 1i'd8 (or 26 . . . 'ilt'd4)
27 lL!fl , but his advantage is not great.
24 'ii'c l
Of course, the queen must get out
B of the pin.
24 'ii'c3
The beginning of an ambitious
plan. Other possibilities:
1) 24 . . . 'ii'f4. Aiming to exchange
queens . After 25 lL!f3 'ii' x c 1 26 l:texc l
axb5 27 l:txb5 .i.a6 28 l:tb6 .i.e2
White has achieved nothing. There
fore 25 lL!c4 !, with the intention of
sacrificing the exchange, comes into
White is a pawn ahead for the mo consideration: 25 . . . \i'xc l 26 l:texc l
ment but that is of minor importance axb5 27 l:txb5 .ta6 28 .ta4 .txb5 29
62 The Art of Chess Analysis
\
Game N ine
Spassky - Fischer
World Championship Match (19),
Reykjavik 1972
Alekh i ne Defen ce
Undoubtedly the most dramatic game of the match was the thirteenth. Fischer
chose the Alekhine Defence and Spassky, after treating it rather inaccurately,
soon found he was forced to offer a pawn for vague attacking chances. Fischer, in
tum, played superficially. Spassky's attacking chances became very real and for a
long time it was unclear who had matters best in hand. It developed into a very un
usual ending which Fischer finally decided in his favour.
Both players seemed to have been affected by that far from faultless game. The
next two games, both finally drawn, were full of serious tactical and strategical
mistakes. Then the weight of all that tension seemed to lift, and both combatants
played freely again. The series of five draws that preceded the last decisive game
contained chess of the highest level. Fischer continued to experiment in the open
ing, with Black as well as with White: a Najdorf in the fifteenth game without the
capture on b2, a Pirc Defence in the seventeenth, and again an Alekhine Defence
in the nineteenth.
The nineteenth game is discussed here. It is a textbook example of attack and
defence balancing each other; as in the tenth game, one can identify mistakes only
after deep analysis. It was also Spassky's final, mighty attempt to keep the world
title from Fischer - the title which would bring Fischer to a state of total inertia.
1 e4 l0f6 7 h3
2 eS lOds Whether or not this move is played
3 d4 d6 will prove to be important later.
4 3 .tg4 7 .thS (D)
An older continuation than the Black cannot very well capture, as
4 . . . g6 played in the thirteenth game of shown by the game Vasiukov-Torre,
the match. It was popular for a while, Manila 1 97 4 (via a different move or
until the latest experiences showed der): 7 . . . .t xf3 8 .txf3 l0c6 9 c4 l0b6
that White has several ways to get the 1 0 J.xc6 bxc6 1 1 b3 0-0 1 2 l0c3 a5 ? !
advantage. 1 3 .te3 l0d7 1 4 'it'h5 ! with great posi
S .tel e6 tional advantage.
6 0-0 .te7 8 c4 l0b6
70 The Art of Chess Analysis
9 lbc3
The capture on d6 used to be played
here automatically. With the text
move, White intends to wait for Black
to play . . . lC!b8-c6.
9
. 0-0
9 . . . dxe5 10 lC!xe5 .i.xe2 1 1 1Wxe2 ! 11 .. .i.xf3
1Wxd4 1 2 l:ld 1 1Wc5 1 3 b4 1Wxb4 1 4 12 .i.xf3
lC!b5 is too dangerous. In Gaprindashvili-Kushnir 1 969,
10 .i.e3 dS the game that originated this s stem
The point of White's avoidance of (but without White's h3 and Black's
exd6 on his ninth move is that this .i.h5), White recaptured with the
push would be more favourable for pawn, forcing the black knight to re
Black if there were no pawns on c7 treat. The white doubled pawn signi
and e5. 10 . . .li)c6 is followed by 1 1 exd6 fies no disadvantage, as appears from,
cxd6 1 2 d5 and White keeps an endur among other games, Pokojowczyk
ing advantage after either 1 2 . . . .i.xf3 S chmidt, Poland 1 976 (again, without
1 3 .i.xf3 li)e5 14 dxe6 fxe6 15 .i.g4 or h3 and .i.h5): 1 1 gxf3 li)c8 12 f4 .i.h4
1 2 . . . exd5 1 3 lC!xd5 lC!xd5 14 1i'xd5 . 1 3 .i.d3 g6 ( 1 3 . . . lC!e7 is probably bet
1 1 cS (D) ter. Then Enklaar's 14 .i.xh7+ is not at
This leads to a great advantage in all convincing since White retains the
space. Although experience has shown bad bishop and Black is able to block
that White can certainly expect an ad ade the position. Better is 14 'ii'h5 lC!f5
vantage, a different move to gain the 1 5 .i.xf5 g6 1 6 'ii'g4 exf5 1 7 1i'f3) 1 4
upper hand has been tried more re f5 ! exf5 1 5 'ii'f3 c 6 1 6 h 1 h8 1 7
cently; e . g . , Sznapik-Schmidt, Polish l:lg 1 lC!e7 1 8 1i'h3 lC!g8 1 9 .i.xf5 ! with
Championship 1 977, went 1 1 cxd5 advantage. After 1 9 . . . gxf5 20 'ii'g 2
Spassky - Fischer 71
Black must return the piece to prevent 1 7 i.xc6 bxc6 1 8 'ii'g4 f5 and Black
mate. certainly does not stand worse. His
12 l0c4 doubled pawn is compensated for by
13 b3 his majority on the kingside.
A fter the match, the system used 2) 15 J.g4 and now:
by Fischer understandably became 2a) 1 5 . . . 1i'd7 1 6 exf6 (This is best
popular. Geller particularly, one of now that the black queen is on a less
Spassky's seconds in Reykjavik, made favourable square - as in the above
grateful use of improvements found variation, e7 is better. 16 e4 dxe4 1 7
during the match (see also the sixth exf6 gxf6 ! i s not good) 1 6 . . . i.xf6 1 7
match game); he introduced 13 i.f4 b4 with freer play for White.
against Hecht in Budapest 1 97 3 , and 2b) 15 .. .f5 1 6 i.e2 and White moves
achieved quick success after 1 3 . . . tlJc6 the bishop to d3. Again he stands a lit
14 b3 tlJ4a5 1 5 'ii'd 2 b6 1 6 l:lac 1 bxc5 tle better.
17 dxc5 i.xc5 ? I 8 lll xd5 i.d4 19 b4 Petrosian draws attention to the
exd5 20 bxa5 'ii'd7 2 1 l:lxc6 'fi'xc6 22 idea 14 . . . lllc6, a typical Petrosian wait
'ikxd4 :ads 23 l:lc 1 'ii' b7 24 i.g5, 1 -0. ing move. White's best is 15 l:lb1 fol
The latest example from practice lowed by 1 6 b4 (the immediate 1 5 b4
is equally discouraging : 13 i.f4 lllc6 is premature because of 1 5 . . . tlJxb4 1 6
14 b3 lll4a5 15 l:lc 1 b6 1 6 llla4 ! (after l:l b 1 tlJc6 1 7 l:lxb7 tlJa5); e.g. , 1 5 . . . b6
1 6 'ikd2 the improvement 1 6 . . . bxc5 17 16 b4 and 17 11fa4 with advantage to
dxc5 l:lb8 ! 18 i.xd5 exd5 1 9 lllx d5 White.
l:lb5 ! 20 b4 lll xb4 2 1 lll xb4 'ii'x d2 22
i. xd2 l:.xc5 is possible, with roughly
equal play, as in Geller-Timman, Wijk
aan Zee 1 975) 1 6 . . . i.g5 17 i.xg5 1i'xg5 w
1 8 'ii'd 3 l:lab8 1 9 i.g4 'ii'f4 20 l:lfd 1 f5
2 1 exf6 l:lxf6 22 'ii'e 3 with advantage
to White (Geller-Timman, Teesside
1 975).
13 lllxe3
14 rxe3 b6 (D)
This manner of attacking the pawn
chain is dubious. On the other method,
1 4 . . . f6, White has two reactions (the
direct solution with 1 5 exf6 i.xf6 1 6 IS e4
i.g4 11fe7 gains nothing): A strong answer. 1 5 b4 promises
1 ) 1 5 e4 dxe4 ! 16 i.xe4 (or 1 6 nothing after 15 . . . a5 16 a3 ( 1 6 11fa4
lll x e4 fxe5 1 7 dxe5 tlJc6) 1 6 . . .lllc 6 ! lll d 7) 1 6 . . . axb4 1 7 axb4 lllc 6 ! . White
72 The Art of Chess Analysis
17 bxcS 1i'as (D) Black not only protects the queen with
The only possible follow-up to the this move but also opens an escape
previous move. If White were able to route for his queen's rook. White,
play 1 8 1Wa4 he would have the posi however, should not be discouraged;
tion well under control; for example, now he can sacrifice the bishop on d5 :
17 ... 7 1 8 11fa4 1Wc7 19 l:ab1 :ab8 20 exd5 cxd5 2 1 .txd5 ! exd5 22 xd5
20 1Wa6 with enduring pressure. 1Wb7 23 1We4 l:a7 24 :bt and White
Spassky - Fischer 73
1i'c3) 20 . . . lDa6 2 1 lDe3. In some cases the black king's shelter: 2 1 .txf7+
Black can sacrifice his knight on c5 :.xn 22 :.xf7. xf7 23 1i'h5+ g8 24
for three pawns. 1i'xg5 . At the moment, White has
19 J.bS ! three pawns for the piece and Black
Spassky 's decision to play for at cannot satisfactorily oppose them with
tack is fully justified, considering the his badly placed knight; e.g., 24 . . . 1i'c3
previous variations, and was probably 25 :.d 1 exd5 26 e6 lDc7 27 1i'e5 and
made when he played his last move. White controls the board.
White has many chances and Black 2) 20 . . . exd5 2 1 J.xf7+ l:.xf7 22
must defend carefully, as we shall see. l:lxf7 and now:
19 cxdS (D) 2a) 22 ... lDc6 23 1i'f3 ! (not 23 1i'h5
lDxd4 24 ltafl ? .th6 ! - but not
24 . . . lDe6? 25 ltxg7+ lDxg7 26 'it'f7+
and mate) 23 . . . 'ifb4 24 :n 1i'xd4+ 25
w h l 1i'xc5 26 l:.c7 ! and White is win
ning.
2b) 22 . . .li'd2 23 1i'g4 (D) (the dif
ference from the actual game is that
here the d4-pawn is protected; 23 l:.c7
achieves nothing because of23 . . . lDa6
24 ltb7 J.e3+ 25 h l 1Wxd4 26 1i'xd4
J.xd4 27 l:r.d l lDxc5 28 l:r.c7 lDa6 !)
the black king would be unsafe after advantage: 22 :n ! . The white rooks
26 . . . e7 27 'ii'f5 . Black must there are connected so that the white queen
fore return the piece with 26 . . . g8 . can threaten to penetrate the black po
But after 27 1i'e6+ h8 28 Wxc6 l:tg8 sition destructively from either side:
29 Wxd5 White retains the better pros via a4, g4, or h5 . After 22 . . . lL!c6 23
pects as he can disarm the attacking 'iWg4 ! 'ii' x d4+ 24 h l 'ii'xe5 25 exd5
try 29 . . . .i.f4 with 30 1i'f3 . Black has no satisfactory way to re
If 28 . . . l:td8 (instead of 28 . . . l:tg8) 29 capture: if 25 . . . exd5 26 :ae 1 .te3 27
1i'b7 ! (much stronger than 29 1i'xd5 'iff3 or 25 . . . 'ii'xd5 26 l:tad 1 1i'e5 27
g6) and the white c-pawn is very dan l:ld6 (27 'ii'f3 'ii'x c5 and Black's f8-
gerous; for example, 29 . . . ti'xd4 30 c6 square is covered twice) and now:
'ii' b 6 (the same reply would follow 3a) 27 . . . l:tf8 28 'ii'x e6+ (surpris
30 . . . 1i'c4) 3 1 l:tb1 'ii'xb7 32 l:txb7 ! and ingly, 28 l:txf8+ xf8 29 'ii'f 3+ e7
wins. 30 :xc6 1i'e 1 + 3 1 h2 .i.e3 ! gives
Nevertheless, the text-move might Black dangerous threats, so White
also have led to a significant advan must take the perpetual check with 3 2
tage, as deep research has shown me. l:tc7+ d8 3 3 1i'f8+ xc7 34 'it'd6+,
20 ... :xn etc.) 28 . . . 'it'xe6 29 J:[xf8+ xf8 30
21 .:.xn 'ii'dll l:txe6 and the endgame is advanta
This brilliant defensive move geous for White;
forces a drawable endgame. Nearly all 3b) 27 . . . lL!e7 28 l:txe6 'it'd5 (or
other moves lose quickly: 28 . . . 1Wxc5 29 h4 and White wins) 29
1) 2 1 . . . .te3+ 22 h2 xf7 23 l:[fe 1 (threatening 30 J:[6e5) 29 . . .1Wf5
'ii' h 5+ e7 24 .:.n lL!d7 25 1i'f7+ (29 . . . h6 30 h4 wins) 30 'it'c4 h8
d8 26 c6 and wins. (30 . . .'ilt'd5 3 1 'ilt'xd5 lL!xd5 32 J:[6e5
2) 2 1 . . .lL!c6 22 'ii'g 4 (simpler than wins material) 3 1 l:txe7 .i.xe7 32 :xe7
Olafsson's 22 1i'h5 ) 22 . . . xf7 23 and White should win.
l:tfl + g8 24 'ii'x e6+ and White will With the queens on the board,
have no less than four pawns for the White has attacking chances because
piece. the black pieces are always hanging.
3) 2 1 . . .1i'c3 is the most reasonable 22 'ii'xd2
alternative. Black ties the white queen There is nothing better; 22 l:tc7
to the protection of the queen's rook. lL!a6 23 l:tc6 lL!b4 24 l:txe6 dxe4 leads
Nei now gives 22 exd5 exd5 23 J:[bl to a sharp position in which Black
lL!c6 24 l:tbb7 .th6 25 'ii'g4 1i'xd4+ 26 would have the better chances.
'ifxd4 lL!xd4 27 l:txa7 with a compli 22
.. .txdl
cated ending; as I see it, White has the 23 .:an
better chances. Olafsson gives a far The rooks are finally connected. It
more convincing way to maintain the seems like a whole game has been
76 The Art of Chess Analysis
played, but we are only just past move exd5 exd5 26 l%d7, since the black
twenty ! knight would stand too passively.
23 6 (D) However, as Olafsson points out, 24
l:c7 fails to an unexpected combina
tion, namely 24 . . . dxe4 ! 25 l:xc6 e3,
and all at once the e-pawn becomes in
credibly dangerous. On 26 l:txe6 there
follows 26 . . . e2 27 l:bl l:f8 ! with the
terrible threat 28 . . . :n + 29 l:xfl
i.e3+ and wins. Initially I thought
that White had a stronger riposte in 26
l:[f4, but subsequent investigation
showed me that White still does not
come out of it well after 26 . . . e2 27
l:e4 e l 'it'+ 28 J:[xe l .t xe l 29 l:txe6
l:d8 ! .
24 exd5 White's passed pawns are not
Spassky must have seen the forced strong enough, because his king is too
draw position already. The play would far away. After 30 l:d6 l:xd6 3 1 exd6
have remained more complicated after 'l;f7 the block of pawns is going to be
24 l:c7, which Byrne and Nei con swept away by Black. Somewhat bet
sider more dangerous for Black. The ter is 30 c6, but then Black continues
point is that after 24 . . . li)xd4 25 :en cold-bloodedly with 30 . . . f7 3 1 c7
i.h6 26 exd5 exd5 27 l:xa7 White :C8 32 l:c6 .ta5 and the c-pawn falls.
reaches the same position as in the So Spassky had seen that he had to
frrst line of variation 3 above (after avoid this variation.
Black's 2 1 st move), a complicated po 24 exdS
sition in which White nevertheless has 25 l:d7
the better chances. Byrne and Nei both The position still seems critical for
note, however, that the white attempt Black, but Fischer quickly dispels that
24 l:c7 would achieve nothing after illusion.
24 . . . lt)d8 25 l:e7 lt)c6 26 l:xe6 li)xd4 25 .te3+
27 l:e7 .te3+ 28 h l dxe4 29 l: l f7 26 hl .txd4
li)e6 ! and Black h as the advantage. Capturing with the bishop im
But this variation gives Black no more proves the co-ordination of his minor
than a draw if White decides to play pieces.
his rook back to c7 on the 26th move. 27 e6 .teS! (D)
Moreover, White can undertake a Fischer's defence is hair-fine. The
well-founded winning attempt with 25 squares d6 and c7 are taken away from
Spassky - Fischer 77
30 n
The safest solution. White would
be the only one with winning chances
after the continuation 30 . . l%xd6 3 1
.
and B agirov's book on the Alekhine better to play energetically for attack
Defence (revised edition, I 979), give with the (most often temporary) piece
I I j,b5+ j,d7 I2 'ii'e2 as the refutation, sacrifice 1 6 liJe5 ! j,xb5 17 j,g5 and
citing the game Ljubojevic-Moses, the complications are probably in
Dresden I 969, which went: I2 ... liJxd5 White's favour.
1 3 e6 fxe6 I4 'ii'x e6+ !De7 I 5 liJf3 Nevertheless, the question remains
'ii' f6 I 6 1!i'e2 !Dc6 I 7 liJe4 'ii'e6 I 8 whether l i j,b5+ is a clear refutation
j,c4 with great advantage for White. of 9 . . . 1i'h4+. White has a much more
Adams points out in the British Chess solid approach in 1 I j,f4. The ending
Magazine, however, that all the theory after I l . . . g5 I2 j,g5 'ii'xe5+ I3 'ile2 is
books have forgotten an old game be clearly better for White.
tween Drogomiretsky and Kaiev from Now we return to the position after
the semi-final of the All-Russian Cham 9 . . . c4 (D).
pionship, S verdlovsk I 934, where
Kaiev improved Black's play with
I 5 . . . 1fb4 (D) (instead of I5 . . . 1ff6).
w
10 ll)fJ
Once again, sound development
above all. Estrin and Panov recom
Black gets a decisive attack after I 6 mend 10 a3 to keep Black's king's
j,xd7 + liJxd7 I 7 liJg5 0-0-0 I 8 ll)f7 bishop out of b4, but Black stands well
!Dc6 I9 !Dxh8? ( 19 !Dxd8 is a little after the simple IO . . . j,c5 . B agirov's
better, although Black still has many recommendation, I 0 d6, is again pre
chances after 1 9 . . c!Oxd8) 19 . . . lbd4 ! 20
. mature, because White w i ll have diffi
'ii'e4 liJf6 2 1 'ifg2 :e8+ 22 d l 1i'c4 culty defending the e-pawn after
- he does not even bother taking the 1 0 . . . !Dc6; e.g., 1 1 !Df3 j,g4 or 1 1 j,f4
time to capture the knight on h8 ! In g5 . Finally, 1 0 1i'd4 also promises lit
stead of helping his opponent to com tle after I O . . . !Dc6 1 1 1i'e4 liJb4 and
plete his development, White does now 1 2 a3 liJ4xd5 1 3 liJxd5 1i'xd5 1 4
80 The Art of Chess A111llysis
17
.i.xgl Wxf6 29 lle l with good compensation
Black takes the rook, a much criti for the pawn.
cised decision. Hort gives this move a l c) 1 9 0-0-0. Zaitsev and Shashin
question mark in the Encyclopedia of now give a line which superficially
Chess Openings yes, we are still
- seems to give Black equal play ; i.e.,
very much in the opening - as do Ko 1 9 .. ..i.xgl 20 l:txgl 'ii'c 5 2 1 lle l l:tae8
tov, Blackstock, and Wade in their 22 .i.e? llxe7 . But Marovic goes fur
joint book World Championship lnter ther and concludes that White clearly
zonals 1973. Zaitsev and Shashin were has the better chances after 23 dxe7
the first to criticise the capture of the Wxe7 24 e6! fxe6 25 llxe6 h8 26
rook. They give two variations in 64: llxe7 llxf4 27 l0<15 llxf3 28 xb6
1 ) 1 7 . . . 8d7 1 8 d6 Wc6 and now: xb6 29 llxb7 . Let us underline this
I a) 1 9 llg4? xeS 20 Wxe5 l:tae8 judgement: White is in fact winning.
2 1 .i.e? .i.xd6 and Black wins. 27 . . . lld4 (instead of 27 . . . llxf3) would
1 b) 1 9 l0e4. This move was sug be tougher, but 28 a3 would best illus
gested by Ree during an informal trate the helplessness of Black's posi
analysis session with me when we first tion. While White quietly prepares to
saw the game. The point is that Black further strengthen his position, Black
runs out of moves after 1 9 . . . .i.xgl 20 cannot free himself.
.i. f6 ! xf6 2 1 exf6 llfe8 22 Wh6 2) 1 7 . . . 1le8. How treacherous the
llxe4+ 23 fl 1i'b5+ 24 xg l . Soon position is can be seen in the variation
after this variation was published in 1 8 d6 .i.xd6 1 9 b5 .i.xe5 20 xc7
Schaakbulletin, Marovic found a hole .i.xc7 + 2 1 'ii'e4 with the better ending
in it. In the Yugoslav magazine Sa for White, a variation originally given
hovski Glasnik he showed that Black by Ree and myself and unquestion
has the much stronger 23rd move ingly adopted by Kotov, Blackstock,
23 . . . lle l + ! He continues with 24 g2 and Wade in their book. However, it
lle2+ 25 xgl Wc5+ 26 h l llxh2+ ! fails to consider an important finesse
27 Wxh2 Wf5 with unclear play. I for Black. Adams gives 1 8 d6? .i. xd6
wonder whether he saw that 28 19 b5 (D) and now:
.i.xf7+ was possible - the bishop can 1 9 . . . 1lxe5 + ! ! 20 Wxe5 ..tb4+ and
be captured only on pain of mate. Black wins.
However, even after 28 . . . h8 White Therefore, after 1 7 . . . lle8 White
has achieved little. The bishop will be must resort to 1 8 ..tf6, the move ana
forced to retreat when Black captures lysed by Zaitsev and Shashin. After
on f6, and thus 28 .i.xf7+ seems to 1 8 . . . 8d7 there can follow:
be merely a wasted tempo for the win 2a) 1 9 d6 (Zaitsev and Shashin).
of a relatively unimportant pawn. However, Black comes out of it well
White's best seems to be 28 Wg3 after 19 . . . xe5 (not 19 . . . Wxd6? 20
Bronstein - l.Jubojevit 83
probably remembered the line incor 2 1 . . .'iVxh2 22 'ii'g 4. Black then loses
rectly and Ljubojevic, the eternal op his bishop, and White has sufficient
timist, probably wanted to try the compensation for the exchange after,
system once more. e.g., 22 . . . h5 23 'iVxg l 'iVxg 1 + 24
18 ... 'ireS xg 1 8d7. I think the reason Ljubo
It was difficult to foresee that this is jevic repeated the variation against
the wrong square for the queen. Be Gheorghiu was that he had found
sides the text-move, I analyse the fol 19 . . . 'ii'd4 ! and felt that the resulting
lowing possibilities: possibilities were sharp enough and
1) 18 . . . 'iVc6. A suggestion of the not onesidedly in White's favour. A
Danish analyst Bo Richter Larsen. Af correct assumption, as later became
ter 1 9 .i f6 8d7 20 'iVh6 xf6 2 1 apparent. Against Grii nfeld at the Riga
exf6 .l:fe8+ White has nothing better Interzonal 1 979, after 20 .l:d 1 Ljubo
than 22 e4, after which the play is jevic did indeed play 20 . . . 'iVxb2 and
similar to that in variation 1 b above won relatively quickly after 2 1 e6
(after Black's 1 7th move). 1 9 0-0-0 is 8d7 22 e7 'iVxh2 23 exf8'iV+ .l:xf8
much stronger. White wins at least a 24 'ti'xh2 .ixh2 25 f6+ g7 26
whole rook back while maintaining a xd7 fud7 27 .ie7 .l:b8 28 cit>f2 .ie5
great positional advantage by 1 9 . . . .ic5 29 .l:c l c5 30 ltd 1 .if6 3 1 .ixf6+
20 e6 8d7 2 1 e7 ! . xf6 32 .ic4 .l:.d8 33 cit>g3 a6 34 .in
2) 1 8 . . . 'ii'c 5 . The commentators cit>e5 35 f4+ cit>e6 36 .ic4+ cit>f6 37
either did not mention this move or cit>f3 b5 and White resigned.
dismissed it with 1 9 e4 'ii'e 3+ 20 Meanwhile this variation has again
'ikxe3 .ixe3+ 2 1 .ixe3 with satisfac been developed further. In order to
tory play for White; if 1 9 . . . 'iVb4+ 20 pursue this, I again consulted the
cit>fl and White wins. But why should 'Alekhine bible ' . Instead of 2 1 e6
Black panic and give check? After the White should play 2 1 f6+ cit>h8 22
centralising 19 . . . 'iVd4, it seems White .l:d2. More than four pages, full of
generally does not have such danger analysis, then follow in the book. The
ous threats against the black king. If best move seems to me to be 22 ...1i'al +.
20 f6+ cit>h8, and 20 .l:d 1 'iVxb2 2 1 In a game Griinfeld-Wiemer, Tecklen
f6+ cit>h8 22 'iVh4 'iVxh2 also leads burg 1 984 (so Griinfeld tried it again !)
nowhere. this was followed by 23 cit>e2 c6 ! 24
After 1 8 . . . 'iVc5 19 e4 'iVd4, the in 'ii'h 4 h5 25 e6? t0d4+ 26 .:txd4 1i'xd4
dicated move is 20 cit>fl so that after and Black had a won game. The Griin
the virtually forced 20 . . . 'iVxb2 White feld variation has not brought very
has a choice of places to put his rook. much success. Siebenhaar et al give
The best is 2 1 .l:el with the idea of in 25 lbg4 as better, after which the wild
directly protecting the e-pawn after suggestion 25 . . .f5 is probably good.
Bronstein - Ljubojevit 85
1 e4 cS 8 f4 /0:6
2 lDt"3 d6 9 .te3 J.d7 (D)
3 d4 cxd4
4 ttixd4 lDf6
5 lDc3 e6
6 .tel
In this match, Karpov employed the
modest 6 .i.e2 for the first time in his
life. In five earlier games he tried the
sharp Keres Attack with success: he
won all five, including a famous one
against Hort in Moscow 1 97 1 .
6 J.e7
7 0-0 0-0
Karpov - Spassky 89
And this is typical of the way i.c4, and Black does not have enough
Spassky played against Fischer two compensation for the sacrificed ex
years earlier. In the eighth game of the change.
1 972 match, Spassky played an in After 1 8 . . . g7 (or 1 8 . . . h5) 1 9 lLlb l
comprehensible tactical blunder with (to regroup), then Black can reply
1 9 . . . ltif6-d7 (Fischer-Spassky : I c4 c5 1 9 . . . 'fi'e6.
2 ltic3 ltic6 3 ltif3 ltif6 4 g3 g6 5 .tg2 19 .tg4
i.g7 6 0-0 0-0 7 d4 cxd4 8 ltixd4 Tal writes that Furman predicted
ltixd4 9 Wxd4 d6 lO i.g5 i.e6 l l Karpov 's moves here and on the 24th
'fi'f4 'ila5 1 2 l:.ac l l:.ab8 1 3 b3 l:.fc8 move. Indeed, Karpov's old teacher
14 'fi'd2 a6 1 5 i.e3 b5 1 6 .ta7 bxc4 1 7 always joined in the analysis when his
i. x b 8 l:.xb8 1 8 bxc4 i.xc4 1 9 l:.fd l pupil stood well, and at such times it
ltid7). Now in the ninth game o f this was impossible to remove the smile
match, he commits an equally incom from his face.
prehensible positional blunder with The text-move is actually very easy
1 8 . . . lLlf6-d7 . to find: it is the only way to prevent
The exchange of White's bad light Black's positional threat 1 9 . . . i.c5 .
squared bishop for the knight leads 19 hS
to a strategically ruinous position. The 20 i.xd7 1i'xd7
correct plan for Black is to force the 21 'ii'c4
exchange of White's other bishop; A consequence of White's 1 9th
for example, 1 8 . . . g7 (the immediate move. Black would have nothing to
1 8 . . . h5 is not bad either) 1 9 l:.c l (In complain about if he could play his
tending to bring the bishop to c4. Per queen to e6.
haps 1 9 l:.d l is better, but if 1 9 1i'c4 21 .th4
'fi'a6 .) 1 9 . . . h5 20 i.e2 lLlh7 and now 22 l:d2 'ii'e1 (D)
2 1 .tc4 is answered by the advance
. . . f7-f5, and other moves are answered
by 2 1 .. .i.g5.
In 64, Tal gives the line 18 ... 'fi'e6 1 9 w
A c 1 l:.ad8 20 i.e2 l:.d4 for Black. He
must have done this analysis in a great
hurry, because we had both concluded
in the press room during the game that
this attempt to keep the bishop out of
c4 was inadequate on account of 2 1 b3
l0xe4 (what else?) 22 i.xd4 exd4 23
l0xe4 (better than 23 i.c4 ltlxf2 ! 24
.t xe6 fxe6) 23 ... 'fi'xe4 24 i.d3 or 24
Karpov - Spassky 91
M osc o w 1974
Queen's I n d ian Defence
The Karpov- Korchnoi match, the finals of the 1 974 candidates series - only later
did it become clear that it was actually for the World Championship - was greeted
by only lukewarm enthusiasm by the chess world. Karpov won two games with
White right in the opening; both times Korchnoi had deviated from his favourite
French Defence. Then Korchnoi won a game with White. Karpov won another
game when Korchnoi cooked his own goose in a horrible manner, and Karpov's
second loss came when he underestimated the dangers in a very clearly drawn po
sition. And it rained draws.
Yet there were no boring games, although the combinational possibilities only
rarely surfaced. One of these games is the eleventh.
The opening followed a traditional path and both players then undertook tradi
tional manoeuvres. Analysis shows that much hidden beauty did not come to the
fore; the game seemed to follow a gradual progression to an ending which, at frrst,
both players were trying to win. Korchnoi finally turned out to be the only one
with chances. Karpov, slippery as an eel, managed to trade dow11 to a rook end
game which seemed to be only a draw despite Korchnoi's two extra pawns.
10 0-0 lLld7
1 1 i.f4
The next time Korchnoi had White,
he deviated with 1 1 l:d 1 . The idea was
probably to answer 1 1 . . . c5 with 1 2
dxc5 bxc5 and immediately fianchetto
the queen's bishop with 1 3 b3 and 1 4
i.b2. After 1 1 . . .l:e8 1 2 i.e3 i.d6 1 3
l:ac l a 5 1 4 'ii'c 2 c 6 Black certainly
did not stand worse.
11
cS
12 dxc5 bxcS
7 d5 Black would not have enough com
More or less forced because of the pensation for the positional disad
threatened e2-e4 . Later, Karpov also vantage of the isolated pawn after
took to answering 7 0-0 with 7 . . . d5 . 1 2 . . . lLlxc5 1 3 'ii'd l . After the text
As Ree points out in the Haagse Post, move he has a reasonable version of
this was Botvinnik's method of play the hanging centre.
ing for a win with Black in the queen 's 13 l:fd1 tLlf6
Indian. The former World Champion 14 'ii'c2 1i'b6
would then meet 8 lLle5 by 8 . . . ti'c8 - 15 lLld2
but White can gain the advantage with White reveals his plans. After an
9 cxd5 exd5 10 i.g5 . undefined build-up typical of his style,
The present World Champion ap Korchnoi is ready for 16 e4, which
proaches it differently; after 7 0-0 d5 8 Black opposes with his following
lLle5 he plays 8 . . . lLla6, a move first move.
used in Smejkal-B yrne, B iel 1 976. It 15
. :res
i ntroduces an important area of open 16 1i'b3 (D)
ing theory, since the position can also This move received unanimous
be reached via the English or the Cata praise. Polugaevsky, in his notes in 64,
lan. called it Korchnoi 's best move of the
8 cxd5 lLlxdS first half of the match.
Seen in the light of the above com White indirectly increases the pres
ments, 8 . . . exd5 is a worthy alternative. sure on d5 ; for instance, 1 6 . . . l:ad8 is
White doesn' t have many choices prevented. This is a good tactic in this
other than 9 lLle5 , and then 9 . . . c5 can sort of position. The player opposing
be played, possibly followed by lLlb8- the hanging pawns must work mainly
a6-c7. with pin-pricks to entice an early . . . c5-
9 lLlxd5 exd5 c4 or . . . d5-d4 or to create disharmony
94 The Art of Chess Analysis
after the simple 23 . . . li)f6. Polugaev 25 . . . l:r.c8 available. This shows a cer
sky's variation continues with 23 li)f5 tain underestimation of the strength of
We6 24 .th3 Wc6, but now Black is White's position, since after 25 b4
squeezed in a bottleneck: the white .l:r.c8 26 .txe4 dxe4 27 li)d7 White
knight comes back with great force wins a pawn. If he were actually faced
(25 li)d4) and Black's prospects are with the problem of 25 b4, the future
gloomy. World Champion would probably
22 .tes lBe4 have solved it in another way, by ac
23 .tal :ads cepting the challenge with 25 . . . cxb4 .
24 tBes 'iVb6 (D) After 2 6 fie? Black has the following
This offers the possibility of a com possibilities:
plicated combinational twist, but one
which would not have turned out
badly for Black. His position is not en
viable, but Karpov is often at his very
best when he really stands badly and is
threatened by genuine danger.
back the pawn with no problems. pressure along the a-file: 35 lL!a5 . The
' Roughly equal chances,' according rook must then retreat since 35 . . . l:ta6
to Polugaevsky ; but, in my view, the fails to 36 i.c3, after which the c
better chances are more likely Black's, pawn walks through unhindered. After
since he has the safer king position. It 35 . . . llc7 36 lL!b3 White has consoli
is important that 35 'ii' b 5, with the dated his extra pawn, but without hav
point 35 . . . l:tc8 ? 36 i.xc5 'ti'xc5 37 ing real winning chances. It is true that
l:.d8+ and wins, achieves nothing be Black cannot win the pawn back di
cause of 35 . . . l:td8, and Black's posi rectly with 36 . . . i.xb3 37 axb3 i.xc5
tion has already been strengthened. because in the bishop ending after
33 lL!c4 the continuation 38 llc l i.b6 39 llxc7
White has finally reached what he i.xc7 40 i.xa7 i.d6 4 1 i.d4 i.b4 42
had in mind: a solid extra pawn. Black f3 f5 White really does have winning
cannot capture the c-pawn because of chances. Not, however, by walking his
the same twist as in the previous note: king over to the queenside, because
33 . . . i.xc5 34 i.xc5 'ii'x c5 35 l:txd5 then he can make no progress: 43 fl
'ii'xc4 36 l:td8+ , etc. f7 44 fxe4 fxe4 45 e2 g6 46 d I
Another try to hold the pawn is 33 e6 47 c2 d5 , and Black comes
i. b 2, which would lead t o a con just in time. If 48 i.c3, then 48 . . . i.d6
vincing result after 33 . . . 'ii'xc5 34 49 i.d2 .i.e5 follows. The b-pawn
.ta3 !, but Black can parry the attack cannot advance because Black's king
simply with 33 . . . 'ii'b7 . gets to c4. The winning attempt, there
33 'ii'c6 fore, is 43 g2 f7 44 fxe4 fxe4 45
34 'ii'xc6 l:txc6 (D) h3 ; White plans to penetrate the
35 ::tel kingside in order to attack e4 and thus
Polugaevsky here gives a method to divert the black bishop from b4; for
prevent Black's rook from exerting example, 45 . . . h5 46 h4 ! g6 47 g4
Korchnoi - Karpov 101
i.e7+ 4 8 '1t h 3 and White will gain against 4 1 lLlxf3 and the plan of trans
ground. ferring the knight to d3 . He gives the
There is no hurry to win the pawn variation 4 l . . .i.xf3 42 xf3 c6 43
back, of course. Black keeps the bal e4 i.xc5 44 f5 d6 (Enklaar ac
ance with the simple 36 . . . f7. and tually gives 44 . . . d5, but then White
White has nothing better than to posi has the extra possibility 45 i.xc5 .:r.xc5
tion himself so as not to lose the c 46 e4+ followed by the exchange of
pawn. rooks) 45 l:txc5 l:txc5 46 i.xc5+
35 n xc5 47 e6 c4 48 f7 b3 49
36 a3 .:r.a6 xg7 xa3 50 xf6, and after the re
37 .:r.cJ e6 spective pawn marches, a queen end
38 lLldl d7 game arises which Enklaar judges is
39 f3 (D) drawn. In my view, however, it is won
With both players short of time, for White because of the bad position
Korchnoi forces his opponent to make of the black king. Instead of this line,
a small fundamental decision. Black has a far more sober continu
ation : 4 l . . .<li>c6 42 lLle l .ie4 and now
he can capture the knight without risk
if it moves to d3, as he could also if the
8 rook were on a4.
An acceptable explanation is that
Black had one move to play before the
time control and that 40 . . . l:ta5 was
careless. After 40 . . . l%a4 White would
certainly not stand better any longer.
41 e4 i.c6
The sealed move. All interim re
ports said that Korchnoi would be a
pawn ahead with slight winning
39 exf3 chances, according to Flohr in Schach
This adds some dynamism to the Echo. But, he added, Karpov had said
position, but it gives up space Black after adjournment that he considered
could have held with 39 . . . f5 . With this the position favourable for Black and
choice Karpov shows that he is not in would be playing for a win. ' A.. little
tent on simply insuring a draw. fairy tale from the Moscow woods,'
40 f2 :aS concluded the commentator jokingly
Why not 40 . . . .:r.a4 to prevent e3- - a childish way to treat a pronounce
e4? Enklaar explains in Schaakbul ment by a man about to become the
letin that the text-move is directed World Champion.
1 02 The Art of Chess Analysis
,
I?
better is an art - or rather a technique -
- B - '
' !>.'
that many modern grandmasters have
mastered. But the power to hold a po
rl 8 . - sition that is slightly worse is pos
a o
!>.' !>.'
sessed only by the absolute greatest.
=
Among them, Fischer was a shining
!<:'
example. In a lost position against Ma
. - - tulovic in the 1 97 1 Interzonal tourna
ment, for instance, he avoided a draw
It is typical of Karpov's play, and by repetition of moves.
one of his strong trumps, that in posi The third player worthy of adding
tions where he can get the worst of it to this list is Korchnoi. Very often he
or already stands clearly worse, he has been willing to allow his position
continues indefatigably, neither per to become critical in order to keep an
mitting a further worsening of his po extra pawn. In this game, his position
sition nor losing sight of his possible at adjournment is not critical and his
winning chances. This is part of the extra pawn hardly matters. He would
profile of a real fighter at the highest not have doubted for a moment that he
level. A typical example was his game had good reason to play for a win. And
against Adorjan at Las Palmas 1 977. had anyone told him afterwards what
The Hungarian grandmaster had the Karpov had thought about the posi
advantage from the opening, but he tion, he would have reacted with little
could not consolidate his advantage more than a vague laugh. This is what
against the fast-moving World Cham makes their games against each other
pion and twice ran into time trouble. so hard, despite the often simplified
Just before the end, during his second positions.
bout of time trouble, he had a draw in 42 xf3 e6
hand but misplayed it, and Karpov fi In his book Anatoly Karpov, His
nally won. I watched the post-mortem. Road to the World Championship,
At one point, when he was still in dif Botvinnik writes that Karpov can head
ficulties, the World Champion could for a draw here with 42 . . . :a4 43 e3
have simplified to a drawn pawn g6 ! 44 i.xf6 :xa3 and adds:
down endgame. 'In this level position, Karpov plays
' Why didn' t you play it then?' for a win; dangerous tactics when a
asked Adorj an ' Didn ' t you see it?' pawn down.' Didn' t Botvinnik him
Karpov replied, 'I saw it, but why self use to do that in his days of glory?
Korchnoi - Karpov 103
43 e3 48 gxfS
Karpov's remark after the game is Flohr recommends 48 . . . xf5, but
characteristic : that 43 o!Dc4 followed after 49 J_f2 Black would face the
by 44 o!Dd6 is the safest way to the same problems as in the game.
draw here. 49 J_f2 J.. g7
43 l:a4 Karpov, apparently realising that he
44 J:.b3 g6 had played too adventurously, offered
45 <ili>d3 a6 a draw here. Korchnoi; according to
Preventing 46 tDc4. Flohr, answered only with a shrug of
46 j_eJ J_bS+ his shoulders.
47 c2 rs (D) so .l:.e3+ <ili>d7
51 .l:.f3 e6
52 l:te3+ d7
53 l:[f3 <ili>e6
White's repetition of moves not
only wins time on the clock but also
demonstrates that, at the moment, he
is the only player who can choose
whether to play for a win or a draw.
54 b3
After this king move, Black's rook
finds itself frequently in a compro
mised position. All this is a conse
quence of Black's 47th move.
But this is much too optimistic. 54 hS
Black opens the position at a time 55 .l:.e3+ d7
when White is ready for action and 56 ttJf3
thus White gets chances to realise his White's winning attempt is now be
extra pawn. The position would re ginning to gain momentum.
main balanced after 47 .. ,J_g7 . 56 J_f6
It is striking that Botvinnik is the Karpov continues as if nothing
only commentator who more-or-less much was going on. It is high time to
approved of the text-move, on the aim for further pawn exchanges to free
grounds that Black must somehow put the black rook. Better is 56 . . . f4 and
his initiative to use. This necessity, then 57 . . . J.. f6 only if White's rook re
however, seems totally absent. treats with 57 l:e l . The difference is
48 e:xfS+ that now White will be able to move
White thus effortlessly rids himself his rook laterally after the knight
of a weak pawn. reaches a more active post than f3 .
1 04 The Art of Chess Analysis
1 e4 e6 3 t0c3 .i.b4
2 d4 dS 4 d2
1 08 The Art of Chess Analysis
and enter the endgame a solid pawn 2 1 lDe5 with a continuing White at
ahead. White cannot prevent this with tack, or gets mated after 1 7 . . .'fle7 1 8
1 1 h3 because Black then uses the 1i'c8+ 1i'd8 1 9 l:te 1 + lDe7 20 l:lxe7+
won tempo with 10 . . . fxg5 1 1 lDexg5 (the final sacrifice) 20 . . .xe7 2 1 1i'e6+
i.d6, again with the threat of ex f8 22 'ii'f7 mate.
changing queens. White is not without These variations give a good pic
chances after 1 1 i.d3 1i'g4 1 2 i.e3, ture of how strongly the white pieces
due to his centralised pieces and ad combine in the attack.
vantage in development; but Black's 13 i.c4 bS
compact pawn structure makes it diffi A necessary move. Black prepares
cult to find compensation. to develop his knight via a6 so that it
1 1 lDexgS 'iff6 may support the threatened e6-point
Now it is not possible to aim for the from c7 and may possibly go to d5 .
exchange of queens with 1 1 . . . 'ii'f5 be The immediate 1 3 . . . lDa6 fails because
cause White mates prettily with 1 2 the black position is too weakened af
i. d 3 1i'g4 1 3 i.g6+ f8 1 4 lDh7+ ter 1 4 i.xa6 bxa6. Gulko then in
l:txh7 1 5 l:ld8+ and mate next move. tended 15 'ii'c4 e5 16 l:ld2 ! with the
12 i.bS+! threat 17 lDxe5, and things remain ex
The correct way for White to make tremely difficult for Black after, say,
use of his chances. After 1 2 i.c4 lDc6 1 6 . . . lDh6 1 7 l:.e l with continuing
1 3 l:lhe 1 lDd8 Black is indeed some heavy pressure.
what cramped, but there are no really 14 i.xbS (D)
vulnerable points.
12
.. c6
This takes the c6-square from the
knight, but there is no decent alterna 8
1 d4 lilr6 8 .. .i.xn
2 c4 e6 9 'it>xn 'ii'xdS
3 lilc3 .i.b4 Capturing with the e-pawn would
4 e3 b6 give White the position he wants after
5 lile2 .i.a6 10 f3.
In a later round in this tournament I 10 'ii'd3 lilbd7
tried the controversial 5 . . . lile4 against The game mentioned in the intro
Donner. He answered with the surpris duction, Gligoric-Portisch, Torremo
ing 6 'ifc2 f5 7 g3 ! and after 7 . . . lilxc3 linos 1 96 1 , continued IO . . . lilc6 1 1 e4
8 lilxc3 .i.b7 9 d5 b5 ? he could have 1i'd7 1 2 .i.g5 lilg8 1 3 e2 f6 1 4 .i.c 1
gained a great advantage with 10 .i.d2 ! lilge7 with equal chances (Gligoric).
(instead of 10 .i.g2). Probably 6 . . . .i.b7 A possible improvement is 12 .i.a3 .
(instead of 6 . . . f5) is more accurate. With the text-move Black achieves
6 lilg3 .i.xc3+ a more harmonious development of
7 bxc3 dS his pieces.
8 cxdS 1 1 e4 'it'aS
An unusual but not unknown move The point of the previous move.
which probably does not offer much. 12 eS (D)
Gligorit - Portisch 1 15
w
w
8 w
entirely watertight, for White can try probably intended, then 33 :bn ! set
37 :b2 (instead of 37 a6), though in tles the matter. The white pieces are
view of the undefended position of the joined again !
white king this attempt to win has lit 32 ... gS (D)
tle chance of success.
It is also interesting to examine ex
actly how the perpetual check comes
about in the final position. w
(Haren) rightly points out that Black 36 "fkc7 l:txb6 37 11Vxb6 'Wxd5+ 38
in fact wins with 35 . . . ..te3 ! . The white l:lfg2 l:.a2 and Black wins.
queen cannot come to the rescue of the 2) 35 'it'd? "ii' x d7 36 o!Llxd7 'iiih 8 !
defence. 37 o!Llf6 l:.f8 ! (the only good square
2) 35 'ii' b 5 . The best move. After for the rook; for example, 37 . . .l:.b8 38
35 . . . l:.e2 36 l:.xe2 "ii'xe2 White gives a o!Llh5 ! ..te3 39 l:.xf7 ..txg1 40 .!Df6
knight back with 37 11t'b2 in order to with unstoppable mate) 38 l:.xf4 (the
get a strong passed a-pawn. only try to get something out of it)
33 l:lea8? 38 . . . gxf4 39 'iiig 2 (White can create
Time trouble is affecting both play mating threats only with the king's
ers . The imperturbable 33 . . . h4 ! is help) 39 ... l:ta2+ (Black makes no pro
necessary, instead of the somewhat gress with 39 . . . 'iitg 7 after 40 o!Lld7 fol
panicky text-move. Suddenly we see lowed by 4 1 f3+) 40 <iPf3 l:.xh2 4 1
the disadvantage of White's last move: l:r.g4 h3 42 xf4 l:r.f2+ 43 'iit g5 . White
his rook on g 1 takes an important finally threatens mate. However, Black
square away from the king, so Black has the sobering 43 . . . l:r.g2 and wins.
can calmly leave his rook on a6. If Ulf Andersson and I worked out this
White then still tries to justify his pre extremely adventurous variation to
vious move, he quickly ends up in a gether. We would very much have
hopeless position, as can be seen after liked to see the inventive white play
3 3 . . . h4 34 o!Llb6 l:lxaS (D) and now: bear fruit, but it was not to be.
The conclusion is that 34 o!Llb6
loses. After 33 . . . h4 White must try 34
l:tgfl (D). This is the correct square
for the white rook, as we also saw in
the notes to White's 32nd move. This
gives Black a free tempo for . . . h5-h4,
but the situation is not critical for
White. Again he threatens to capture
the rook on a6.
Now Black cannot continue his
counterattack, since 34 . . . 1i'h3 fails to
35 l:.g2 ! l:.e2 36 l:.xg5 + ! and White
decisively goes on the attack. The
1 ) 35 'Wc6 with the intention of variation continues 36 . . . !iii> f8 37 'i'b8+
meeting 35 . . . l:te3 with 36 'ii'x d6, with e7 38 'tli'c7+ <iPf6 39 'ii'd 8+ .l:te7 40
good prospects for White. Black has a l:tg2 and White wins.
better rook move, however: 35 . . Jlb8 Black is therefore forced to head
(threatening 3 6 . . . l:ta6 to win a piece) for an endgame with 34 . 'i'c8 . White
. .
Geller - Spassky 1 27
then maintains some winning chances Black's position would collapse very
because of his strong a-pawn which is quickly after 35 lLle4.
conveniently protected by the knight. 35 1i'xa8+ h7
But Black also has strong pawns, pro 36 c4
tected by the bishop, and an exchange Protects the d-pawn in a natural
sacrifice on a5 is a possibility. way.
34 lLlxd6 (D) 36 l%f6
Already the fatal blow. White not 37 'it'b7 1i'h3
only wins an important pawn, but the After 37 . . .'ii'x d3 the simplest win is
knight, which was posted soundly but 38 1i'e7 with the threat 39 l:lxf4. Black
not very actively on the queenside, sets a final trap with the text: he threat
now enters the skirmishes on the king ens 38 . . . .txh2 with at least a draw.
side with tempo. 38 'ii'b2
34 l%xd6 A simple preventive measure.
In desperation Black sacrifices the 38 . .tes
exchange without getting anything for 39 1i'e2 1-0
it. 34 . . .'it'f6 is only worse, however; (Black lost on time)
Game S ixteen
Lju bojevic - Andersson
Ho og oven To urnament, Wijk aan Zee 1976
Sici l ian Defence, Scheven ingen Variati o n
Kavalek once mentioned, with a mixture o f astonishment and admiration, how re
markable it is that Andersson and Ljubojevic can walk to the tournament hall to
gether laughing and joking and then, across the chess board, face each other in a
life-and-death struggle. That they do is fortunate, for otherwise the chess world
would be poorer by a lot of interesting games.
At Wijk aan Zee 1 976 Ljubojevic started tremendously with three wins in a
row. This is the third. A clash of styles is clearly apparent here. The actual course
of the game is extremely original, and the hidden possibilities that show up in the
analysis add extra refreshment. Seldom have I analysed a game with so much
pleasure and devotion.
1 e4 c5 9 f4 d6
2 lilf3 e6 10 'i!Vel 0-0
3 d4 cxd4 11 'i!Vg3 Jld7
4 lLlxd4 llX6 12 e5 (D)
5 lLlc3
Unusual for Ljubojevic. I have never
seen him use any system against the
Taimanov system other than the build
up of the Maroczy wall with 5 lLlb5 d6
6 c4, which he handles in his own
highly refined manner.
5 'i/c7
6 Jle2 a6
7 0-0 lLlf6
8 Jle3 Jle7
The game thus enters the paths of
the Scheveningen Variation. The char
acter of the Taimanov system can be A new idea which has had repeated
maintained with 8 . . Jl b4, but nearly
. success in recent years. The first time
all the lines arising from 9 l0xc6 bxc6 was in Geller-Timman, Hilversum
10 lLla4 are slightly better for White. 1973 (with the added moves l:tad l for
Ljubojevit - Andersson 1 29
member being very surprised, but I fxe5 ltlxe5 1 6 .if4 (Mensch's vari
quickly realised that Black faced a ation does not work now as the black
hopeless task. After the picturesque b-pawn is sufficiently protected after
continuation 1 3 e5 ltle8 1 4 ltle4 dxe5 1 6 .ih6 ltle8 1 7 .if4 .id6) 1 6 . . . i.d6
15 fxe5 ltlxe5 16 .if4 f6 17 .ig4 I 17 :ad 1 and the threat of 18 ltlb3 is
managed to survive to the 70th move very unpleasant.
by sacrificing the e6-pawn. The fourth instance of this pawn
The second time it was played was sacrifice that I know of is a game Ma
in Tal-Hartston, Hastings 1 973n4. tulovi6-Janoevi6, with the additional
The difference from the diagram posi moves a2-a3 and . . . b7-b5 . There it was
tion was the added moves g 1 - h 1 and a real pawn sacrifice - as it is here in
. . . b7-b5, which did not seem to be a Ljuboj evi6-Andersson - and Janoe
great improvement for B lack. As the vi6 was wrong not to accept it.
Hoogoven chess player J. Mensch In the present encounter, Andersson
showed in Schaakbulletin 8 1/82, ac manages to show that the white action
cepting the pawn offer is a precarious is premature . And a good thing, too.
business here too: 1 3 e5 dxe5 14 fxe5 My first reaction when I played over
ltlxe5 1 5 .ih6 ! ltle8 1 6 .if4 f6 17 .ig4 this game was, ' If this is good, then
or 16 . . . .id6 17 .ixb5 ! , and in either Black can't play the Sicilian anymore.'
case White wins the pawn back with The game of chess has not yet reached
advantage. Hartston did not accept the that point.
pawn sacrifice, and this soon proved 12 ... dxe5
fatal: 14 . . . ltlxd4 (instead of 14 . . . ltlxe5) 13 fxe5 ltlxe5
1 5 J.xd4 ltle8 1 6 .id3 .ic6 1 7 'ifh3 Naturally.
g6 1 8 :ae 1 l:td8 19 'ife3 l:td7 20 .ib6 14 .if4 .id6
1i'c8 2 1 ltle4 .ixe4 22 .ixe4 h5 23 15 :ad1
1Wh6 ltlg7 24 l:txf7 ! . A quiet but very pregnant move.
In h i s game a s White against Kar The simple threat is 1 6 ltlb3, after
pov at Nice 1 974, Hartston showed which the various pins would cost
that he had learned something from Black material . He therefore frees the
his encounter with Tal . For unclear c7-square for the bishop.
reasons he did not play Tal 's 13 e5, but 15 'ifb8
first 1 3 a3 and after 1 3 . . . :ab8? only The only move. Other methods to
then 14 e5 . Karpov did not take the solve the problem of the pins fail ; e.g.,
pawn and ended up, as I did against 1 5 . . . ltld5 1 6 ltlf5 ! exf5 1 7 ltlxd5 'ii'c 5+
Geller, with a dismal position. The more 18 .ie3 'fi'c6 19 ltlf6+ h8 20 ltlxd7
so with a rook on b8, taking the pawn with complete destruction, or the fan
would have led to the same reaction tastic 1 5 . . . ltlf3+ 16 :xf3 e5 17 .ih6
1 30 The Art of Chess Analysis
8 8
l:.xe6+ and mate next move) 25 In his column 'Game of the Month'
'ifxg7+ q;xg7 26 l:.g3+ with perpetual which appears in many chess maga
check. zines the world over, Gligori refers to
24 f6 an analysis by chess enthusiasts from
Alternatives were: Bosanski Shamac in Yugoslavia. In
1 ) 24 . . . f5 . This is also a move that the last variation they continue with
creates a chance not considered by 29 . . . l:.f6 ! (instead of 29 . . . l:.d 1 +) with
Ulf. Yet it is not easy to refute. After the intention, after the too hasty 30
25 l:.h4+ q;g8 26 1V g6, play branches 1i'e7, of giving mate with 30 . . . l:.d 1 +
once again: 3 1 q;f2 :n + 3 2 q;g3 'ii' b 8+ and mate.
1 a) 26 ... ltlxe5 27 'ifxe6+ l:.f7 (not The move 29 . . . l:.f6 has no further
27 . . . ltlf7 28 l:.ch3 and Black cannot threat, however, so White can take the
stop the mate even though he can take time to play a quiet move: 30 c4 ! . Be
the knight with check) 28 'ifxe5 (D) sides the threat to capture the bishop,
(now not 28 l:.ch3, however, because this move also threatens the now
then Black can indeed save himself: crushing 3 1 1i'e7. I see no satisfactory
28 . . . 'ii'x c5+ 29 q;h 1 ltlg6 30 'ii'x g6 defence for Black.
q;f8 3 1 1i'e6 ltlf6 and Black has won 1 b) 26 . . . l:.f6. An interesting at
the game of cat and mouse). tempt to defend which was suggested
Now B lack has hardly any moves. by the German player K. Werner. The
He can still try 28 . . . l:.d8, but the idea is that after 27 xf6 c!Clxf6 White
threatened penetration of the rook cannot follow up with 28 l:.ch3 be
need not bother White, as we see in cause of 28 . . . 'ifxc5+ 29 q;h 1 q;rs 30
the line 29 l:.ch3 l:.d 1 + 30 q;f2 l:.fl + l:.h8+ q;e7 3 1 'ifxg7+ q;d6 32 l:.xa8
3 1 q;g3 f4+ 32 q;g4 c!Clf6+ 33 q;g5 1i'xc2 and Black has the advantage.
and the white king escapes while his Stronger, therefore, is 28 a4 ! , in order
colleague faces mate in a few moves. on 28 . . . e2 to continue with 29 b4.
Ljubojevi{: - Andersson 1 33
The transfer of the other rook to the Threatening mate in three. The
kingside cannot then be avoided . Also white knight plays a useful role in the
28 ... l:[d8, in order to begin a counter attack even though it's still pinned.
attack after 29 axb5 l:[d l + 30 f2 26 ltld8 (D)
axb5, is not quite sufficient on account
of 3 1 g3 1t'b8+ 32 h3 l:[h l 33 g3
and the white king is safe. Inciden
tally, it is important that White ftrst w
Attempts to keep the attack going S pain, where Raymond Keene and I
demand great sacrifices, but White were playing. The day before I re
cannot afford to make them because ceived the Hoogoven tournament bul
his knight is still pinned; e.g., 30 i.xg7 letin that contained this game, Keene
lilxg7 3 1 l:th7 l:g8 32 h i l:ad8 and gave me his striking opinion of Ljubo
now White must try to find a perpetual jevic's play. He said he did not see Yu
check. But not 32 . . . lile5 ? 33 lile4 ! goslavia's top grandmaster as a deep
with a decisive attack. Henk Jonker strategist, and not even as a player
tried to find good play for Black with with any sort of healthy ideas about
the queen sacrifice 26 ... 'ifxc5+ 27 l:xc5 chess - but this was precisely what
lilxe5 . The three minor pieces would made his play so successful and so dif
certainly compensate for the queen if ficult to counter; that is, it consisted of
White did not have an immediate way a series of tricks. Very deep tricks in
to win material, also later given by deed, he added : they can be twenty
Jonker: 28 l:h8+ f7 29 l:xf8+ xf8 moves long.
30 'ifa3 !, threatening both a crushing As far as this game is concerned,
discovered check and 3 1 l:xb5 . Keene was right on target. On the 1 2th
27 i.d4 b6 move Ljubojevic started a series of un
The weakness that this creates is clear complications, and now, eight
not at all obvious, but Ljubojevic must een moves later, Black seems to be
have spotted it immediately. lost: while his rook is hanging on one
28 lilxe6 lilxe6 side of the board, mate in one is threat
29 'it'xe6+ "ikf7 ened on the other. But it's still only a
30 "ike4 (D) trick.
If Black had not been in serious
time trouble, or perhaps if he had not
fallen so deeply under the spell of
White's play, he would undoubtedly
have found the courageous path to the
draw.
30 . gS?
Correct is 30 . . . 'ikxa2 ! , and the black
pieces seem to be co-ordinated again.
A very dangerous check is threatened,
and if White takes the bait he will be
in big trouble. On 3 1 "ikxa8 Black frrst
plays the calm 3 1 . . . g5 (D); White does
This tournament was played at the not have a queen check on d5 and the
same time as a tournament in Orense, white rook cannot leave the fourth
Ljubojevic - Andersson 1 35
1 e4 cS
2 l0t'3 lDc6
3 d4 cxd4
4 xd4 f6 w
5 lDc3 d6
6 i.gS e6
7 ..d2 a6
8 0-0-0 i. d7
9 f4 i.e7
10 f3 bS
11 .ixf6 gxf6
12 <i>b1 .. b6
13 i.d3 b4
14 2 aS Fischer-Spassky match. I spent a long
15 fS a4 time thinking about whether there
1 6 f4 ..cS (D) was an alternative to 1 6 . . . 11t'c5 , but I
After a slight transposition of moves could see nothing else against the
we have entered the 1 8th game of the threatened 1 7 fxe6 fxe6 1 8 i.c4. For
Karpov - Timman 1 37
32 l:.xe4 37 l%c3+ d6
White must enter this four-rook 38 h3
ending as after 32 l0d5 l%ge8 B lack Typical Karpov. He has no real win
keeps his strong passed pawn on e4. ning chances and is not at all in a
32 .i.xf4 hurry.
33 l%xf4 l%xg2 (D) 38 l%e2
39 Acd3 ci?eS
40 a4 :.as
41 l%xb3 :Xa4
w 42 l%b8 ltd4
A nice finesse which makes things
a little less troublesome for Black.
43 l%e8+ d6 (D)
34 l%f7+ c6
35 l%xh7 ltbS
The draw can probably be held in
more than one way. With the text
move Black plans to hold his centre
pawn and exchange his b-pawn for
White's a-pawn.
36 l%h3 dS (D) The adjourned position. I expected
my opponent to seal 44 l%d8+ because
the king would then have to return to
e5 and thus White could reserve the
w choice of which rook to capture. It
matters little, though, because it is still
a draw.
44 l:txd4 AxeS
45 c2 lte3
The most active : Black is ready to
answer 46 l%d3 with 46 . . . l%e4. His
king is ideally placed in the centre.
46 l%h4 ci?cS
Karpov - Timman 141
47 llh8 lle2+ SO hS d3
48 b3 d4 51 h6 b6
49 h4 llh2 (D) I had calculated long before that the
king would arrive just in time.
52 c3 d2
53 c2 b7
w 54 b4 a7
SS bS b7
White cannot win the d-pawn un
less he can put Black in zugzwang. The
move h6-h7 is therefore necessary
sooner or later. After that, however,
Black doesn't need the pawn to draw.
56 h7 a7
57 dl b7
lf2.1fz
Ga me Eig hteen
Ti m man - Karpov
Euwe Tournament, Amsterdam 1976
N i mzo-ln d i a n Defen ce, Len i n g rad Va riatio n
In May 1 976 a four-player tournament in honour of Max Euwe was held for the
first time. The venue was uniquely situated in the Van Gogh Museum in Amster
dam, and the demonstration room.was no less unique: a tent of mirrors which for
a few months stood in the Museum Square. Professor Euwe turned 7 5 during the
tournament amid an appropriate setting. A phonograph record was even made,
with text added to music found in an old barrel-organ book from the time when
Euwe was World Champion: the 'Euwe March.' All together, a perfect 1 930's at
mosphere was created.
Karpov's participation was an extra attraction. The World Champion did not
have many tournaments on his schedule for the rest of the year, and many people
wondered whether he was taking too great a risk by playing in such a short tour
nament. After all, even a single 'accident' could seriously jeopardise his expected
first place. Karpov clearly felt this himself, for he played very cautiously. In the
first round he defeated Browne, who had actually achieved a drawn position but
went under in terrible time pressure. Four draws followed and in the last round a
win against Olafsson secured Karpov's first place.
He faced his most anxious moments in the fifth round, in the game given here.
After missing a win, I finally had to be satisfied with a draw.
advantage. But B lack's rook need not by Bohm. After the 'exchange' on b2,
move : 28 . . . c3 ! 29 &tJxc7 i.xa2 and, the white king moves back and forth
amazingly, White has only one way to between al and b2 so that he cannot
stop the c-pawn: 30 &tJe6+. Black must be forced into zugzwang.
take the knight to prevent it from get From all this it seems that White
ting to d3 or e2 via f4 : 30 . . . i.xe6. would have had to answer 27 . . . l:txd4
Now comes 3 1 l:lbl c2 32 llc l i.b3 with 28 exd4, which would have been
33 f2 b5 . With his king on d2 or d3, more or less the same as the actual
the best White could hope for would game c9ntinuation.
be to return the exchange on c2; there 2 a4
fore, 34 e4 b4 35 e3 a5 36 d5 a4 37 Not nice, fixing the pawn on the
d4 (D). B lack cannot prevent the wrong colour; but if White wants to
white king's occupation of c5 by put make any progress, it's virtually un
ting his own king on d6, for then f3-f4 avoidable.
and e4-e5 would follow. So it seems to 28 a6
be all over now, since on 37 . . . a3 38 29 &tJa3 l:txd4
c5 is decisive. Black, however, has a 30 exd4 l:tc6
sparkling finesse which the World 31 .!Oc2 e7
Champion was not able to work out 32 l:tb2
completely over the board. Threatening 33 &tJb4. The push 3 3
37 . . . .ta2 ! . After 38 l:txc2 b3 39 a5 bxa5 3 4 l:ta4 l:tb6 doesn ' t lead to
l:tc8+ e7 40 l:tc7+ d8 (but not anything.
40 . . . d6?? 4 1 l:tb7 followed by 42 32 ... aS
f4 and 43 e5+ with mate ! ) and Black 33 dS! (D)
wins, so White has no better than to The only reasonable winning at
head for a draw with 39 l:tb2 ! (instead tempt. 33 &tJe3 seems promising but
of 39 l:tc8+) 39 . a3 40 c3, as given
. . leads to nothing after 33 . . . d6; e.g.,
Timman - Karpov 147
36 . d7
Now everything is in order again.
37 h4 c7
33 i.xdS 38 l:te6 (D)
The fact that Karpov thought a few A little joke at the end of an ener
minutes before making this capture in vating game . The pawn ending is won
dicates that he had overlooked White's for White after 38 . . . i.xe6 39 xe6+
last move - or that he wanted to create d6 40 xeS xeS 4 1 e3 dS 42
the impression that he had overlooked f4 ! , etc. Karpov, of course, will have
it. nothing to do with it.
34 tDd4 l:r.cS
The best square for the rook.
35 l:r.xb6 hS?
Maybe a consequence of overlook
ing White's 33rd move. The steady re
action is 3S . . . d7, to be able to
defend his weak points in time.
36 12?
In slight time trouble, I reacted
automatically with the answer I had
planned for 3S . . .'1d7 . Things would
be much more difficult for Black after
36 l:r.a6. He would have to give up a
148 The Art of Chess Analysis
38 d7 42 Ab6 c7
39 l:b6 c7 43 Ae6 d7
40 Ae6 44 Ab6 c7
40 l:bl is also nothing because of 45 Ae6 1/2-112
40 . i.f7 and Black is again ready to
. . The draw was agreed here, before
besiege White's a-pawn. adjourning. Playing for a win with 45
40 d7 A bl would have been risky due to
41 Aa6 i.b7! 45 . . . i.c8 46 o!Db5+ d8 ! .
Game N ineteen
Spassky - Korch noi
Final Candidates Match (4), Belgrade 1977
French Defence, Wi nawer Variation
After Korchnoi sought asylum in the West in August 1 976, the Russians system
atically tried to make life difficult for him. First Pravda and 64 published a letter
signed by most Russian grandmasters condemning and criticising him. Then the
Soviet Chess Federation asked FIDE to exclude Korchnoi from the matches lead
ing to the world championship. Their motto was, perhaps, 'Even if it doesn' t help,
it can't hurt.'
Fortunately, it didn't help. But it meant that until further notice Korchnoi could
meet Soviet opponents only in matches. The confrontation peaked in 1 977, when
he successively met Petrosian, Polugaevsky, and Spassky. Petrosian went down
after a nerve-racking equal struggle, Polugaevsky never had a chance, and it
seemed at first that Spassky would be similarly run over: after nine games
Korchnoi had built up a lead of 6112-2112. At that point, Spassky began to exhibit a
remarkable pattern of behaviour: he appeared at the board only to make his
moves, and then he immediately sauntered backstage to muse over the course of
the game - a unique method of thought which appears to be not very efficient. In
stead of realising this, Korchnoi took it as a serious insult. He became extremely
upset and lost four games in a row. Only then did he slowly pull himself together.
With two draws and two more wins, he won the match by a comfortable margin
and went on to face Karpov.
This game is from the period before the relationship between Korchnoi and
Spassky was so radically disturbed.
1 1 c4 'flc7
12 e:xf6
w I play 1 2 cxd5 first, which amounts
to a transposition of moves.
12 gxf6
13 cxd5 lLlxdS
14 c3
But now the game takes another
route. A gainst Korchnoi I tried 14 c4
to aim for the endgame after 14 ... lLlde7
1 5 dxc5 0-0-0 16 j,c3 e5 1 7 1i'd6. The
continuation was 1 7 . . lLlf5 1 8 'ii'x c7+
.
This was first played by Korchnoi, <li>xc7 19 0-0 lLlfd4 20 lLlxd4 lLlxd4 2 1
against me, in Leeuwarden 1 976. j,d 1 c6 and now White should have
B lack provokes his opponent to open further opened the position with 22 f4
the centre at once, an idea formerly (instead of 22 j,xd4 as played), after
considered too dangerous because of which the chances would have re
White's bishop-pair. mained balanced.
In the old days Botvinnik used to The text-move is based on a differ
close the centre with 10 ... c4. Planinc has ent strategy: White holds d4 so as to
had particular success against 10 . . . c4 prevent Black from quickly freeing his
with 1 1 lLlg5, and even the great expert game with e6-e5 . The centralised po
Uhlmann could not find a satisfactory sition of the knight on d5 helps Black
reply; e.g., Kavalek-Uhlmann, Manila to rapidly carry out this advance any
1 976: 10 . . . c4 l l lLlg5 h6 1 2 lLlh3 0-0-0 way.
1 3 lLlf4 <ti>b8 1 4 0-0 lLlc8 15 lLlh5 14 0-0-0
.:.hg8 1 6 j,g4 lLlb6 17 .:.e 1 lLlxa4 1 8 15 0-0 .:.hg8
.:.e3 b 5 1 9 .:.f3 j,e8 20 .:.g3 lLle7 2 1 16 .:.e1 (D)
j, c 1 lLlg6 2 2 j,a3 aS 2 3 'ii'd 2 'ilc7 A very tense position. After this
24 j,h3 j,d7 25 j,d6 1i'c6 26 .:.n move the storm breaks, but 16 g3
j,cS 27 f4 f5 28 exf6 gxf6 29 f5 ! and would also have been answered by
White broke the position open. 1 6 . . . e5 17 c4 lLlf4 with complications
Later, Vaganian tried 1 0 . . . 'ii'c 7 and similar to those in the game.
reached a good position against Klo 16 .. e5!
van after 1 1 0-0 0-0 12 .:.e1 h6 1 3 j,f4 Very sharply judged.
lLlg6 1 4 j,g3 lLlce7 . Kuraj ica, against 17 c4
me, later improved White's play : 1 3 The standard reaction. White wants
'ii'c 1 f6 1 4 exf6 .:.xf6 1 5 1i'a3 ! c4 1 6 to establish a strong protected passed
lLle5 and White stood better. pawn in the centre. Spassky now had
Spassky - Korchnoi 151
conclusion was that Black would have suppose that he himself (Kholmov
little to fear. Perhaps 1 1 d2 was dis that is) doesn't take it very seriously.'
cussed briefly and dismissed on the But not everyone was of this opin
grounds that Korchnoi would quickly ion. Dvoretsky emphatically recom
get into time trouble due to the inevi mends the line given by Dolmatov:
table complications, and in that case 15 . . . g4 16 cxd5 e5 . He then con
the correctness of the black moves siders 17 i.f4 3+ 18 i.xd3 exd3 1 9
would not play too great a role. 0-0 e2+ 2 0 xe2 dxe2 2 1 'ii'xf5 to
Korchnoi, however, showed his be the best line for White.
best side: no time trouble, but a clear Chances are then roughly equal .
headed, strong reaction. The question remains whether Black's
lS i.c2 attack after 1 7 exd4 i.xd4 18 i.e3
Good enough to keep a clear advan i.xe3 1 9 fxe3 11fc5 is really so strong
tage, but the Steinitzian retreat 1 5 as Dvoretsky wishes it to appear.
i.fl ! would have led to a winning po After 20 dxe4 'ii'x e3+ 2 1 i.e2 in
sition after some complications. For my opinion Black has insufficient
instance: compensation for the piece. In order to
1) 15 . . . g4 . In my original anno substantiate this assessment I give the
tation I stated : 'Given by Kholmov in following variations:
Shakhmaty v SSSR. He continues 1 6 1 a) 2 1 . . .i.g4 22 g3 (above all,
xd5 e5 ? ! 1 7 exd4? e3 ! 1 8 'ifxf5 not 22 11fc 1 ?? 3+) 22 ...l:lac8 23 'ifc2
f3 +! ! 1 9 'ii'xf3 exd2 mate (exclama l:tc4 24 'it'd2 1i'b6 25 d6 and Black's
tion and question marks by Kholmov). attack no longer amounts to much.
Despite the three successive piece sac 1 b) 2 1 . . .g4 (the best chance) 22
rifices, the variation is rather clumsy l:td3 ! 'ii' b 6 (not 22 . . . i.xe4 because of
and pedestrian, and is scarcely rele 23 l:txe3, winning; 22 . . . 'ii'f4 23 l:tf3
vant since the black knight's move to 'tli'e5 24 :XC5 ! 'ii'xf5 25 i.xg4 'ii'xg4 26
e5 is intended merely to tempt White 0-0 also offers no chance) 23 l:tf3 (not
to capture on d4 . Moreover, 1 5 . . . g4 23 l:tg3 f2 ! ) 23 . . . i.g6 (or 23 . . . e3
is just a blow in the air after the simple 24 l:txe3 Wxe3 25 6+ gxf6 26 1i'xf5
1 6 cxd5 . Then the c4-square becomes 'ii'c 1 + 27 d 1 , followed by 28 0-0
available to both the knight and the with a winning advantage for White)
king's bishop while the knight on d4 24 l:tg3 ! and White keeps the upper
stays in the trap.' I was not the only hand in all cases; e.g.: 24 . . . ffi 25
one to think this. In his Het matchboek l:txg6, or 24 . . .f5 25 We t ! .
Karpov-Kortchnoi Hans Ree observes 2) 1 5 . . . i.xa3 . The most important
with regard to Kholmov's variation: move. White has the following possi
'Really very nice, but the distribu bilities after the forced 16 i.xf6 gxf6
tion of punctuation marks leads one to (D):
158 The Art of Chess Analysis
16 'ii'xc2 (D)
44 <Rxf3 fj;f7 (D) l:d5 ! (D), the most obvious move and
the strongest. Black has two reason
able replies:
45 l:c8
Keene, in his book on the match,
writes that Korchnoi 's team analysed l ) 4S . . . i.bS (to keep White's king
the adj ourned position until the last out of the centre for the moment) 49
possible moment and wasted a lot of exf5 i.a7 (threatening to simplify to a
time on the unnecessary piece sacri drawn rook ending) 50 f6 ! (a charac
fice 45 l:dS <Re7 46 l:d7+ <ReS 47 e4 teristic tactical twist) 50 .. . gxf6 5 1 /0e4
l:b5 4S exf5 l:xc5 49 l:xg7. They with an easy win.
could not find a clear winning line af 2) 4S . . . fxe4+ 49 <Rxe4 i.bS 50
ter 49 . . . h6 ! . Korchnoi chose the text <Rd4 (or 50 <Rd3) 50 . . . i.a7 5 1 <Rc4
move at the last moment; since no l:b6 (on 5 l .. .l:bl or 5 l . . .l:b2, 52 1tlb3
concrete winning variation had been is immediately decisive) 52 /0e4 ! (D).
found, he decided to hold the position
as it was.
However, Keene and Stean were on
the right track. The introductory moves B
of the variation, 45 l:dS <Re7 46 l:.d7+
<ReS 47 e4 l:b5 are completely logical.
Maybe they underestimated White's
chances. After a brief look, Andersson
and I came to the conclusion that
White must be winning and that a
forced winning variation, hidden or
not, existed. After nearly an hour's
search, our analysis continued with 4S
Korchnoi - Karpov 165
not 50 . . . i.xc5 5 1 b8'ii' .J:lxb8 52 .J:lxb8 win. The main variation runs 48 . . . e7
i.xe3 53 g6 i.d4 54 .J:lg8 and White 49 .J:ld7+ e8 (otherwise he loses ma
wins, but 50 . . . f7 and if 5 1 e4?? then terial : 49 . . . f6 50 .J:ld5 i.b8 5 1 .J:ld8
White is suddenly mated with 5 l . ..g6) i.c7 52 b8'it' or 50 . . . i.c7 5 1 lba6 !
50 . . . .J:lb4 ! 5 1 e4 (if 5 1 c3, now .l:xb7 52 .:td7 winning a piece) 50 .J:ld5
5 l . . .i.xc5) 5 l . ..i.e5 ! (D). i.b8 5 1 .J:lxf5 and White wins the
black h-pawn also after 5 l . . .i.d6 52
lLle4 . If his pawn were still on h7,
Black would have good drawing
w chances after 52 . . . i.e7 followed by
53 . . . .J:lxb7.
48
gS
Desperation.
49 g4
The surest way. White is not misled
by 49 lLld3 , which leads to a draw af
ter 49 . . . g4+ 50 hxg4+ hxg4 5 1 t2
e7 ! 52 lDxe5 xd8 53 b8'ii'+ .J:lxb8
54 lDc6+ c7 55 lDxb8 xb8 5 6 e4
This last variation was given by An c7 57 e3 d6 (Tal).
dersson. It is difficult to see how White A good move is 49 lLld7, given by
can make progress after 46 . . . .J:lb5 . Kholmov. After 49 . . . g4+ 50 hxg4
47 .J:lg8 f7 Black has two ways of recapturing:
48 lld8 (D) 1) 50 . . . fxg4+ 5 1 e4 i.xg3 52
b81t' i.xb8 5 3 .J:lxb8. Kholmov stops
here. If Black does not exchange
rooks, 54 lLle5+ will be decisive, and
53 . . . .J:lxb8 54 lLlxb8 h4 55 lDc6 f6
56 f4 h3 57 lDe5 g3 58 lDg4+ fol
lowed by 59 xg3 also leads to an
easy win for White.
2) 50 ... hxg4+ 5 1 e2 .J:lb2+ 52 3
i.xg3 53 b8'it' i.xb8 54 .J:lxb8 . Khol
mov again gives no further analysis.
This time, avoiding the exchange of
rooks is bad because of 55 .J:lb6 fol
lowed by 56 lDe5 , and 54 . . . .1:xb8 55
With White's pawn on h3 and lDxb8 f6 56 lLlc6 g5 57 lDd4 g3 5 8
Black's on h5 this leads to a fairly easy e2 g4 5 9 fl i s also hopeless for
Korchnoi - Karpov 1 67
1 e4 cS 6 .tel bS
2 lbf3 e6 This advance is completely sound
3 d4 cxd4 now that White has developed his
4 lbxd4 a6 king's bishop.
Polugaevsky does not play this sys 7 .tf3 .ib7
tem often. He prefers the Najdorf and 8 0-0 lbc6
sometimes the Scheveningen. 9 lbxc6 dxc6 (D)
5 lbc3 'fkc7 A known theoretical position. Most
5 . . . b5 is known to be somewhat of the examples now continue 10 a4 ;
premature because of 6 .id3. A recent for example, 1 0 . . . .id6 1 1 axb5 cxb5
example is Tal-Ljubojevic, Montreal 12 e5 .ixe5 1 3 lbxb5 axb5 14 l:txa8+
1 979: 6 ....ib7 7 0-0 lbc6 8 lbxc6 .ixc6 .ixa8 15 .ixa8 .ixh2+ 16 h 1 .id6
9 1i'e2 lbf6 10 e5 lbd5 1 1 lbxd5 .ixd5 with roughly equal play (Estrin-Polu
1 2 a4 1i'a5 1 3 l:td 1 with advantage to gaevsky, US SR 1 964). A more cau
White. tious approach is 1 1 g3 . White won
Kasparov - Polr.gaevsky 169
10 eS! ?
Good o r bad? Kasparov and Sakh
arov put a ' ? ! ' (dubious) after the move
in their notes in the Informant. My
analysis indicates that it is neither bet
ter nor worse than the normal theoreti
cal paths. Nearly all the variations
lead finally to equal positions.
10
'ii'xeS This move ought to do in any case.
Naturally, he accepts the offer. 1 2 . . . g6 is bad, of course, because of 1 3
1 1 l:lel 'flc7 'ii'd4.
Black can force an ending with 13 Axe6 g6
1 1 . . .11Vd6, since after 12 'ii'e2 f6 he The sharpest: Black aims to win
would simply be a pawn ahead. Barle material. 13 . . . f6 is quieter. Kasparov
Miles, B led/Portorof 1 979, continued and Sakharov reject that move be
1 2 .tg5 11Vxd 1 1 3 Aaxd 1 .te7 1 4 lbe4 cause of 1 4 e4, with the point that
f6 1 5 i.xf6 gxf6 16 .i.h5 Aa7 ! 1 7 14 . . . xh5 15 'ii'xh5 0-0 1 6 i. f4 leads
170 The Art of Chess Analysis
21 c7
The most obvious, although there
Oddly enough, after this powerful is also an idea behind the alternative
move it seems that White has run out 2 l . . .c8 . He must meet 22 fi)d5 by
of ammunition. The queen sacrifice 22 . . . J.xd5, as in the game (22 . . . J.d6 is
2 1 J. xe7 :txg3 22 J.xf6+ f8 23 refuted by 23 fi)b6+ c7 24 l:te8), and
fi)e4 :tg6 is insufficient, and after 2 1 after 23 llxd5, fi)f6 is possible. If now
llxe7+ xd8 2 2 'tlt'b8+ xe7 White White thoughtlessly plays 24 llf5 ,
can ' t even force a draw because the Black achieves the desired co-ordina
black king escapes via f5 . tion with 24 . . . d7 . Much stronger,
Andersson's attempt to improve the however, is 24 llde5 and after 24 . . .J.d6
white play with 19 1i'f4 (instead of 1 9 only now 25 llf5. The important black
J.c7) also fails, because of the laconic f-pawn falls, and Black does not get a
reply 1 9 . . . d7 . dangerous enough initiative; for ex
18 'ii'g6 ample, 25 . . . fi)g4 26 :xn c4 27 llf5
19 J.xd8 1Vxg3 with a winning position. Black must
20 hxg3 xd8 therefore play more modestly.
Kasparov - Polugaevsky 173
been playing the Exchange Variation the usual 1 4 l:[b 1 (after the pawn ex
against the Griinfeld ever since his change). Even now, after 1 3 . . . cxd4 1 4
youth, has never shown any interest in cxd4 White's position remains very
the exchange sacrifice. He chooses an good and he can sacrifice his d-pawn
other continuation, which is associ with little to worry about. You can see
ated with great risk. that after 14 . . . i.. b5 15 l:tb1 i.. xe2 1 6
12 i.. dS i.. d7 (D) 'ihe2 i.. x d4+ 17 h 1 Black has seri
ous problems.
13 i.. bS
14 l:tb1 i..a6
Not so good is l4 . . . 1i'd7, because
after 15 l:txb5 'ii'x b5 1 6 i.. xe7 White
clearly has the better chances. But
now Black is ready for l 5 . . . 'ii'd7 fol
lowed by 1 6 . . . e6, after which White's
position would collapse . Therefore,
the following supersharp attacking at
tempt is born of necessity.
15 f4 'ii'd7
Black can insert l5 . . . h6 here, at a
13 i.. g5 moment when White is forced to
The usual move is 1 3 l:tb1 , as in move his bishop to h4 . But my intui
Hort-Timman, Niksic 1 97 8 . After 1 3 tion warned me that it was too risky,
l':.b1 'ii'c 8 1 4 dxc5 , Black could have and a closer look shows that this was
had a good game with 14 . . . l:td8, as the correct evaluation; after 15 . . . h6 1 6
Hort suggested after the game . Need i.. h4 'ii'd 7 1 7 f5 gxf5 White launches
less to say, taking on c5 meant that an offensive full of bold sacrifices: 1 8
White's strategy was a failure. There tLlg3 ! e6 1 9 ttlh5 ! ! (D) and now:
were no good alternatives, however, 1) 19 ... i.. xfl 20 i.. f6 ! exd5 2 1 i.. xg7
because Black threatened to get the and White's attack quickly becomes
better position with 14 . . . e6 15 i.. b3 decisive. The imperturbable manner in
tLlxb3 . which White's attack is conducted in
Spassky therefore chooses a con this variation is rather typical of this
tinuation that lets him keep his bishop position; Black's pieces are active and
on d5 for a while. It is a remarkable well co-ordinated, but they are not
idea because the move i.. g 5 in con well placed for defence.
nection with the pawn exchange on d4 2) 1 9 . . . f6 20 l:txf5 ! exd5 2 1 i.. xf6
(and therefore played on move 1 4 ) is i.. xf6 22 tLlxf6+ l:txf6 23 l:xf6 'ii'g 7
never mentioned as an alternative to 24 e5. Although the attack has been
178 The Art of Chess A111l lysis
temporarily halted and the material to 19 i.h6 l:.f7 20 lilf4 ! exd5 2 1 lilh5
situation has not turned out too badly with a winning attack) 19 lilf4 exd5
for Black, White has the better chances 20 'ii'h 5 fxe4 2 1 l:.h3 1i'f5 22 g4 ..ie2
because Black's minor pieces cannot 23 lilxe2 1i'g6 24 1i'h4 e3 25 l:.fl , with
be brought onto the battlefield very the unavoidable threat of 26 lilf4, I re
easily. alised that 1 9 exf5 'iWxd5 20 f6 pro
16 rs gxrs duces a much more plausible win.
17 l:.f3 (D) Remember this long variation, how
It took Spassky quite a while to find ever, for it will come up again.
this unsophisticated continuation of Choosing1>etween the text and the
the attack. Now 1 7 lilg3 e6 1 8 lilh5 is insertion of 1 7 . . . h6 gave me the big
not sufficient because of 1 8 . . . f6. Al gest headache. After the subtle reply
though 19 l:.xf5 is still possible (as in 1 8 .te l ! Black's position is critical .
variation 2 above), under these circum Taking on e4 is now forced, and after
stances it is just too fantastic : Black 1 8 . . . fxe4 1 9 l:.g3 (D) I investigated:
reacts cold-bloodedly with 1 9 . . exd5.
1) 19 ... 'ii'x d5 20 .txh6 'fih5 2 1 set-up is the fact that his e- and f
.t xg7 'ii'x e2 2 2 'ii'c 1 ! with the devas pawns are not yet advanced, but aside
tating threat of 23 'fih6. from that it is surprising how few de
2) 1 9 . . . h7 20 lLlf4 .td3 21 'ii'h 5 fensive possibilities Black has. The
.txbl 22 l:txg7+ xg7 23 ltle6+ and main variation is 20 llc l 'ii' f5 2 1 'ifh5
White has a mating attack. (threatening 22 .th6) 2 1 . . .e6 22 llh3
18 llg3 (D) exd5 23 g4 .te2 24 ltlxe2 'fi'g6 25
'ii'h4 and we get almost the same posi
tion as in the note after Black's 17th
move. After 25 . . . h6 26 lLlf4 'ii'h 7 27
B .txh6 .txh6 28 1i'f6+ g8 29 llxh6
White's attack has reached storm pro
portions.
19 'ifxe2 h8
20 .txe4
One of the ideas of Black's defence
becomes clear after 20 'ii'xe4 f5 2 1
'ii' h 4 'ii'x d5 22 llh3 'ii'g 8 and Black
wins.
20
.. rs
18 .txe2! 21 .trJ
A necessary exchange, because the Quite rightly refraining from fur
white knight threatened to enter the ther attacking tries and placing the
thick of the fight by going to f4. Nev bishop on the long diagonal where it
ertheless, after 18 . . . h8 1 9 lLlf4 Black can exert the most pressure.
is not lost because he still has the fi 21
.. cxd4 (D)
nesse 1 9 . . . .td3 . White's attack is not There is no time to protect the e
yet strong enough for him to sacrifice pawn; after 2 1 . . .e6 22 dxc5 White's
a rook: 20 1i'h5 .txbl and if now 2 1 bishop-pair comes to life and after
.txf7 llxf7 22 'ii'xf7 llf8 or 2 1 .th6 22 . . . .txc3 he would have more than
.txh6 22 'ii'xh6 llg8 23 .txf7 e3 or 2 1 sufficient compensation for the pawn .
.te6 .txd4+ ! 2 2 cxd4 'ii'xd4+ 2 3 h 1 22 .txe7!
.txa2, White's attack is repulsed and The only way to stay in the game.
Black's material plus will be decisive. The endgame after 22 'ii' xe7 'ilxe7 23
But the rook sacrifice is unneces .txe7 llfe8 24 .tb4 ltlc6 offers White
sary. White's pieces occupy ideal at little chance of survival.
tacking positions and he can take the 22
.. d3
time to play 20 llc l . The most impor The beginning of a forced simpli
tant strength of B lack's defensive fication leading to a roughly equal
1 80 The Art of Chess Analysis
40 l:.b2+?
It happens quite often that a player
makes a mistake on his 40th move.
The text-move helps only White, and
Black's winning chances melt like
snow in the sun.
40 . . . e5 is no more effective, be
cause of 4 1 .ie2. Black should have After putting up a heroic resistance,
realised that f5 , not e5, is the ideal Spassky went wrong with 52 l:.h6+
place for the knight and that 40 . . .3 ! and had to resign after 52 . . .e5 53
i s the correct move. After 4 1 l:.d7 h6 l:.b6 4 54 l:.e6+ 'ii?d4 55 l:.e4+ c5
Black maintains winning chances; for 56 l:.xa4 l:.a l . Later analysis showed
Spassky - Timman 18 3
sition: 50 b3
During the post-mortem analysis 51 .:.xh7 b2
Spassky suggested 56 Wd2 h5 57 .:.bt , 52 .:.b7 1fz.1/z
Game Twenty-three
Ti m man - Tsesh kovsky
8/ed/Port oroi 1979
English Opening
The Vidmar Memorial Tournament is held every two years. For some reason or
other, the fifth of the series, in 1 979, attracted me immediately. Not that I have
ever played through a game of Vidmar's - at least, never a game he won; but prob
ably I have seen a number of his losses printed among the collected games of
Alekhine, Capablanca, and Euwe. Frankly, this splendid tournament is a rather
exaggerated mark of honour for a not very brilliant chess player.
The frrst six rounds were played in Bled, Yugoslavia, beside a Jake of serene,
almost sterile beauty. The last nine were played in Portoroz, a rather mundane
bathing resort with a casino which accepts only Italian lire. Venice is two hours
away in a fast boat. Yet it was not even these attractive locations that made the
tournament so tempting from the frrst moment. It was something else: a tourna
ment that seemed cut out for me to win. Though not as strong as Montreal, it was
strong enough to make a first place honourable. Larsen, at the opening ceremony,
seemed to be thinking the same thing. He had just arrived by train, and the re
freshing white wine being passed around was obviously doing him some good -
and me too, for that matter. 'Who is going to win the tournament?' one of the or
ganisers asked us. Politely, I made a noncommittal reply, whereupon Larsen,
bursting with self-confidence, swallowed a good mouthful of wine and declared,
'I am going to win the tournament ! ' I must admit that I was at first taken aback by
this display of naked optimism, and I often recalled it during the tournament.
The struggle for frrst place did indeed take place between Larsen and me, after
Ribli dropped out of the running by losing spectacularly to Marjanovic in the
eleventh round. Larsen was still half a point ahead of me after that round, but the
situation was reversed when he lost to Chi and I beat Tseshkovsky.
That victory meant a great deal to me. Not only was I clearly at the top of the
crosstable for the first time in the tournament, but also it had been a long time
since I last beat a Russian grandmaster. A psychological factor entered into the
game, too: although Tseshkovsky did most of the playing, so to speak, I went into
the complications with a healthy measure of optimism and confidence.
14 'ii'x b5 lik6. Black has quite enough with 1 8 .i.h6. Black's reaction is ade
compensation for the pawn after 1 5 quate, however.
.i.xc6 bxc6 1 6 'ii'xc6 h3. 16 l:b8
12
c6 17 lDB4
Naturally. Without this pawn sacri Again threatening 1 8 .i.h6, but
fice, White would get a clear posi Black has a strong reply which takes
tional advantage. all the sting out of the white strategy.
13 d3 a6! (D) Perhaps the pawn sacrifice 17 .i.h6 is
White's best chance. After 17 . . . .i.xh6
18 'i!i'xh6 l:xb2 he has sufficient com
pensation for the pawn. White's best
would then be to force the exchange of
Black's rook with 19 l:fbl .
17 h5! (D)
I played this quickly, uncertain Black's king is now cut off from the
whether or not the pawn ending after queenside and White's can approach
32 . . . d6 33 l:td4+ ri;;c7 34 l:txd8 xd8 the weak pawns unhindered. 33 l:ta4 is
was a win. Black, in time trouble, un less accurate because of 33 . . . l:te8+ 34
derstandably did not want to chance it. d2 l:ta8 35 c3 e5 36 c4 ri;;d 6
Closer analysis, however, showed and the black king is in time to prevent
that it is a draw. The main variation his colleague from penetrating.
runs: 35 d4 rl;e7 (or 35 . . . rl;d7, but 33 l:tb8
not 35 . . . c7 because of 36 h4 f5 37 e4 34 c3 .z:tb5
and White gets two passed pawns. Af 35 l:tc4 (D)
ter either 35 . . .<>d7 or 35 . . . e7, then
36 h4 f5 37 e4 would be satisfactorily
answered by 37 . . . rl;e6.) 36 rj;c5 a5 37
ci>c4 d6 38 b3 d5 39 a4 (Black 8
would have winning chances after 39
e3 c5) 39 . . . d4 40 xa5 rj;e3 41 b6
<i>xe2 42 rl;xc6 xd3 43 ri;;d5 e3 44
rj;e5 (D).
23 .Z:.xcS 1i'xa2
24 .Z:.xgS
B This is the position the World
Champion was aiming for. His advan
tage is not great, but it is quite endur-
ing.
24 ... 1i'bl+?
Preparing to exchange queens.
However, the rook ending looks un
tenable. 24 . . ..Z:.f5 is correct. The point
is that after the exchange of rooks a
draw would be unavoidable: 25 .Z:.xf5
'ilkf7 White can ' t avoid the endgame, 'ii' b l + 26 g2 'ikxf5 27 'ikxa7 'ike4+
since his f-pawn is hanging. But the and recovers the pawn on e2. This
rook ending after 24 'ifc8+ 'ikf8 25 means that 25 .Z:.g4 would be the only
flxf8+ xf8 26 e4 .Z:.e5 27 .Z:.c4 is no winning try, but Black would not stand
forced draw; e.g. : much worse after 25 . . . a5 . His passed
1 ) 27 . . . g4 28 h3 h 5 2 9 hxg4 hxg4 pawn would ensure counterplay.
30 fl f7 3 1 e2 f6 32 e3 g5 25 gl 1i'b6
33 .Z:.a4 .Z:.e7 34 .Z:.a5 .Z:.c7 35 d3 and 26 .Z:.e5
Black gradually runs out of good White has absolutely no objection
moves. to a rook endgame.
2) 27 . . .e7 28 f4 exf4 29 gxf4 .Z:.h5 26 'ikxe3
30 g2 (much stronger than B ohm's 27 .Z:.xe3 (D)
30 e5, which Black meets by 30 ... .Z:.h4
3 1 g2 g5) 30 . . . d6 3 1 .Z:.a4 c6 32
.Z:.xa7 and now Van Wij gerden 's con
tinuation 32 . . . c4 33 .Z:.xg7 c3 34 f3
looks hopeless for Black; e.g., 34 . . . c2
35 .Z:.g 1 c5 36 .Z:.c 1 .Z:.xh2 37 a4 and
the white passed pawns decide. Much
tougher, however, is 32 . . . g6 to prevent
White from getting two connected
passed pawns. Black would then have
reasonable drawing chances.
All in all, Hort had little reason to
go in for this. The text-move is more
solid since White's remaining pawns The most critical position of the
will all be on the kingside. game. Hort now defends his weak
Karpov - Hort 1 99
pawn on a7 with his rook on the sec How does White make progress?
ond rank, but, as the game shows, this He gains nothing tangible with 1 g5+
method fails to build a tight defensive f7 2 l%a7+ f8. nor with 1 l%a7 g5 .
line. Most commentators recommend The only try is to bring the king
27 . . . n 28 l%a3 h5 as the best de nearer. After, say, I e2 l%b3 2 d2
fence. B ohm writes that the resulting 1;{7 3 l::tc 5 l::t a3 4 g5 e7 5 l%c7+ f8
four-against-three endgame does not 6 l::tc 3 l::ta4 with the threat of forcing a
look lost, and Van Wijgerden claims draw immediately with 7 . . . e5 8 fxe5
that sooner or later White must play l:[g4, White makes no progress.
g3-g4 with a probable draw. This method therefore seems to
I will subject this ending to a have little chance of success. White's
closer examination, continuing after king fails to penetrate.
29 l%xa7+ f6 30 l%a5 . Black has two
plausible replies. Method Y
1) 30 . . . g6 is the most solid: Black White aims for this position:
tries firmly to hold his ground.
I will now show the several meth
ods White has at his disposal so that
we can form a good idea of this ending w
Method X
White aims for the pawn structure e3-
f4-g4, as given by Van Wijgerden and
later achieved in the game. He easily
reaches the following position:
30 . . . g6'? A brief variation shows that for the sake of convenience; Timman
he comes out one tempo short: 3 1 >f3 Meulders was actually at move 43
l%b8 32 f4 l%b2 3 3 l%e5 and now not here), Black loses in a studylike man
33 ... g5+ 34 :Xg5 l%xe2 35 l%xh5 lW2+ ner: 35 f5 ! gxf5 36 exf5 :n 37 l%g6+
36 g4 and White wins, but 33 ... l%b4+! h7 (D) (or 37 ... f7 38 J:lg5 e4 39
34 e4 l%b2 with a draw. White lacks l%xh5 e3 40 l:th4 and wins).
the extra move h2-h4.
Method Z
White's king tries to penetrate via h4, B
and he plays 3 1 f4 l%b8 32 e4 ! l%b2+
33 h3 (D).
8
A very refined defence. The point is
that the white rook is tied down after
34 e5 + f5 35 :as g5 36 :f8+ e4 .
White can win a second pawn with
check after 36 fxg5 (instead of 36
:f8+) 36 ... xg5 37 :gS+ h6 3S
:es, but it is insufficient for victory
after 3S . . . lle2 39 llxe6+ g5 . The
white king stands too poorly; e.g., 40
Z l a) 37 . . . f6 3S lla6+ f5 39 h3 ! :es h6 4 1 e6 g6 42 e7 g7.
and the threatened 40 g4+ cannot be On 33 . . . llf2 White quietly replies
adequately met; for example, 39 . . . :el 34 :b5 (D). Remarkably, Black is in
40 g4+ hxg4 41 hxg4+ xf4 42 :f6+ zugzwang.
202 The Art of Chess Analysis
Andersson v Ljubojevic 1 28
Bronstein v Ljubojevic 78
Fischer v Larsen 26, Petrosian 38, Spassky 47, 56, 69
Geller v Spassky 120
Gligoric v Portisch 20, 1 14
Gulko v Timman 107
Hort v Karpov 1 94
Karpov v Hort 194, Korchnoi 92, 1 54, Spassky 88, Timman 136, 1 42
Kasparov v Polugaevsky 168
Korchnoi v Karpov 92, 154, Spassky 149
Larsen v Fischer 26
Ljubojevic v Andersson 128, Bronstein 78
Mecking v Polugaevsky 1 3
Petrosian v Fischer 38
Polugaevsky v Kasparov 1 68, Mecking 13
Portisch v Gligoric 20, 1 1 4, Smyslov 9
Smyslov v Portisch 9
Spassky v Fischer 47, 56, 69, Geller 1 20, Karpov 88, Korchnoi 149,
Timman 176
Stein v Taimanov 33
Taimanov v Stein 33
Timman v Gulko 1 07, Karpov 1 36, 142, Spassky 1 76, Tseshkovsky 185
Tseshkovsky v Timman 1 85
I n d ex of Open i n g s