4. Contents
Introduction 5
I V. Anand- v. Inkiov, Calcutta 1986 7
2 V. Anand- K. Ninov, World Junior Championship,
Baguio City 1987 1 2
3 V. Anand- S. Agdestein, World Junior Championship,
Baguio City 1987 16
4 V. Anand -J. Benjamin, Wijk aan Zee 1989 23
5 M. Thl- V. Anand, Youth vs Veterans, Cannes 1989 29
6 V. Anand - B. Spassky, Youth vs Veterans, Cannes 1989 34
7 M. Kuijf- V. Anand, Wijk aan Zee 1990 40
8 M. Petursson - V. Anand, Manila Interzonal/990 43
9 V. Anand - I. Morovic Fernandez, Novi Sad Olympiad 1990 49
10 A. Beliavsky - V. Anand, Munich 1991 54
1 1 A. Karpov - V. Anand, Candidates match (6), Brussels 1991 59
1 2 V. Anand - G. Kasparov, 1ilburg 1991 67
13 G. Kasparov - V. Anand, Reggio Emilia 199112 74
14 V. Anand -E. Bareev, Dortmund 1992 8 1
15 V. Anand - R. Hiibner, Dortmund 1992 87
16 V. Anand- I. Sokolov, SWIFT rapid, Brussels 1992 90
17 V. lvanchuk - V. Anand, Match (1), Linares 1992 92
18 V. Anand - G. Kamsky, Alekhine Memorial, Moscow 1992 98
19 V. Anand - V. Ivanchuk, Linares 1993 102
20 B. Gelfand - V. Anand, Linares 1993 106
21 V. Anand- E. Bareev, Linares 1993 1 10
22 V. Anand - F. Izeta, Madrid 1993 1 15
23 V. Anand - L. Ftacnik, Biel Interzonal 1993 1 19
24 L. 011 - V. Anand, Biel Interzonal/993 126
25 M. Adams - V. Anand, European Clubs Cup Final,
Hilversum1993 130
26 V. Anand - A. Be1iavsky, PCA Qualifier. Groningen 1993 136
27 J. Benjamin - V. Anand, PCA Qualifier. Groningen 1993 142
28 V. Anand - G. Kamsky, Linares 1994 148
5. 4 VISHY ANAND: MY BEST GAMES OF CHESS
29 V. Anand-J, Polgar, Linares 1994 153
30 V. Anand-G. Kamsky, PCA Candidates (3), Las Palmas 1995 157
31 V. Anand- G. Kamsky, PCA Candidates (9), Las Palmas 1995 164
32 V. Anand-G. Kamsky, PCA Candidates (11), Las Palmas 1995 171
33 V. Anand -J. Timman, Tal Memorial, Riga 1995 176
34 V. Anand - G. Kasparov, PCA World Championship (9),
New York 1995 184
35 V. Anand -B. Gelfand, Wijk aanZee 1996 189
36 V. Anand -J. Polgar, Amber Rapid, Monte Carlo 1996 198
37 V. Anand-V. Topalov, Donmund 1996 202
38 V. Anand-V. lvanchuk, Las Palmas 1996 207
39 V. Anand-A. Karpov, Las Palmas 1996 2 l l
40 V. Anand -J, Lautier, Biel/997 216
41 J. Lautier - V. Anand, Biel/997 220
42 V. Kramnik - V. Anand, Belgrade 1997 225
43 P. Nikolic- V. Anand, FIDE World Ch., Groningen 1997 234
44 V. Anand -A. Shirov, FIDE World Ch., Groningen 1997 238
45 V. Anand- A. Karpov, FIDE World Ch. Final (6),
Lausanne 1998 244
46 V. Anand-V. Topalov, Wijk aan Zee 1998 249
47 V. Ivanchuk-V. Anand, Linares 1998 253
48 V. Anand-V. Kr amnik, 1ilburg 1998 2S7
49 V. Anand-L. 011, European Clubs Cup, Belgrade 1999 262
50 D. Reindennan-V. Anand, Wijk aan Zee 1999 269
5 1 V. Anand-J. Piket, Wijk aan Zee 1999 273
52 V. Anand-P. Svidler, Linares 1999 277
53 V. Topalov-V. Anand, Linares 1999 283
54 V. Anand-P. Nikolic, Wijk aan Zee 2000 291
55 V. Anand-A. Khalifman, FIDE World Cup, Shenyang 2000 297
56 V. Anand-M. Adams, FIDE World Ch., New Delhi 2000 304
57 V. Anand - A. Shirov, FIDE World Ch. Final (4), Teheran 2000 3ll
Combinations
Solutions
Index of Opponents
Index of Openings
Symbols
316
326
335
336
336
6. Introduction
I don't want to spend toomuch time on biographical details, because this is a
book about my games, soI will be content with a brief sketch.
I was born on 11th December 1969 in Chennai (Madras), and learnt chess
atthe age of six from my mother. A year later I joined the TaJ. Chess club in
Chennai. A couple of years later I went to Manila when my father had an as
signment there and got caught up in the Philippines chess fever resulting
from the 1978 Karpov-Korchnoi World Championship match. Returning to
India,it was not until 1983 that I achieved a real breakthrough. In that year I
won both the national Sub-Junior (under- 16) and Junior (under- 19) titles,
and qualified for the (adult) national championship. The championship itself
was held the following year, and I finished fourth. In 1985 I gained my Inter
national Mastertitle;at the time I was the youngest Asian ever to achieve this
distinction. In 1986 I won the National Championship and became India's
youngest champion.
The following year, 1987, was special. After having played three times be
forein the World Junior Championship, finishing lOth,5th and 7th, I finally
won this title. In December of the same year I became the world's youngest
grandmaster (at that time).
In July 1990, I broke through the 2600 barrier. In the same month, I quali
fied for the Candidates from the Manila Interzonal. The first round of the
Candidates was held in January 1991, and I won my match against Dreev, but
in the quarter-finals I lost narrowly to Karpov in a match that hingedon the fi
nalgame. After this disappointment, I concentrated on tournament play with
somesuccess: I won outright at Reggio Emilia 199 112 (ahead of Karpov and
Kasparov) and in Moscow (November 1992).
In 1993 I qualified for both the FIDE and PCA Candidates cycles. The fol
lowing year I was successful in the PCA Candidates cycle, defeating first Ro
manishin and then Adams to reach the Final of the Candidates. In the FIDE
cycle I was less fortunate, losing to Kamsky after having been two games up
with three to play.
In early 1995 I defeated Kamsky in the Final of the PCA Candidates and
gained the right to face Kasparov for the PCA World Championship in New
York. As everybody knows, I started well but faded in the second half.
7. 6 VJSHY ANAND: MY BEST GAMES OF CHESS
However, losing this match was not the end of the world and in 1996 Iwas
back on the tournament trail, finishing joint first with Kramnik at Dortmund
and beating Kasparov in the Final of the Geneva Quickplay. However, the
most important event of 1996 had nothing to do with chess: on June 27th I
married Acuna in Chennai.
1997 was a successful year. In April I finishedjoint first with Kramnik at
Dos Hermanas and in May I became the first person to win the Amber tour
nament in Monaco twice. There followed a win in the Frankfurt Rapid, sec
ond place in Dortmund and outright first in Biel. Finally, I tied for first place
with Ivanchuk at Belgrade in November.
At the end of 1997 I participated in the FIDE World Championship held at
Groningen, and since I qualified for the final at Lausanne this event spilled
over into 1998. I reached the final against Karpov, but then tiredness took its
toll and, despite a 3-3 result in the match itself, I lost the tie-break. Despite
this early disappointment, 1998 was my most successful year up to that time.
I won the chess Oscar, and took first place in five major tournaments at Wijk
aan Zee, Linares, Madrid, Frankfurt and Tilburg.
1999 started well, since I scored +6 at Wijk aan lJ!e to finish half a point
behind Kasparov, but the rest of the year was rather disappointing. The prob
lems started at Linares, where I lost a crucial game to Kasparov. Whether
this was the sole cause is hard to say, but at any rate I struggled with my form
for several months thereafter.
The new millennium began with a modest success at Wijk aan ZJ!e, Where
I finished joint second, but once again Linares proved discouraging. How
ever, from this point on my form rapidly improved, and my play regained the
freshness and vigour which had been the foundation for my run of succes in
1998. First I won the advanced chess event in Leon and the Frankfurt rapid,
and then I was joint first at Dortmund. A further win in the FIDE World Cup
in Shenyang left me in excellent shape for the crucial event of the year- the
FIDE World Championship. This time there was no disappointment at the
last hurdle, and I achieved my greatest triumph to date with a decisive 3112-'h
victory in the final against Shirov.
Now I will sign off and leave the readers to enjoy the games.
Vishy Anand
Collado. Spain
June 2001
8. Game 1
V. Anand - V. lnkiov
Calcutta 1986
Sicilian, Richter-Rauzer
1 e4 cS
2 lLi3 d6
3 d4 cxd4
4 �d4 .!M6
s 00 �
6 i.gS e6
7 'l'd2 i.e7
8 0-0-0 0-0
9 ll:lb3
In 1986 this move was in vogue
thanks to the efforts of Mikhail Tal.
9 aS
10 a4 dS
11 i.bS
A move first played in Tal-Sis
mega, Taxco Interzonal 1985, even
though Tal ascribes the move's in
vention to Vitolin�. Tal won that game
and scored an even greater success
when he beat Korchnoi with it at the
1985 Montpellier Candidates Tour
nament. I figured that I could do
worse than to follow in Tal's foot
steps.
11 ll:lb4
Sisniega played l l...lt:lxe4 and
Kon:hnoi l l...dxe4. l l...i.b4 is an
other possibility; after 12 exd5 exd5
13 'l'f4, followed by exchanges on
c3 and f6, both sides will end up
with fract.•red pawn structures, but
it will be difficult for Black to get
at White's c-pawns, whereas White
will attack the d5-pawn. The move
l l...ll:lb4had been played before, in
Vitolins-Inkiov, Jurmala 1985.
12 l:the1 (D)
12 .•. dxe4
After 12...'illc7 13 e5 ll:le8 14
ll:ld4! Black is in a bad way. Due to
the weakness of the e6-pawn, Black
can't play ...f6 and his e8-knight is
badly placed.
13 'ii'xd8
Really the first new move of the
game. Vitolins played 13 ll:lxe4lllxe4
14 11rxd8 �xg5+ 15 'ii'xg5/t:xg5 16
h4, regaining the piece with an un
clear position.
9. 8 VJSHY ANAND: MY BEST GAMES OF CHESS
13 ... l:l.xd8
Mter l 3.....txd8 14 lllxe4 ..te7
(l4...lllxe4 15 ..txd8 lllxf2 16 l:l.d2
is very promising for White; Black
has no reasonable defence to the
threats of 17 ..txa5 and 17 ..te7) 15
lllxf6+ ..txf6 (l5 ...gxf6 16 ..th6 is
similar to the game) 16 ..txf6 gxf6
17 l:l.d6 Black can play neither ...b6
nor ...e5, and so has serious prob
lems developing his queenside.
14 lllxe4 lllbd5 (D)
15 c4
Apositionally ugly move, but the
main thing is to exploit White's lead
in development. To this end, White
must keep the d-file open.
15 - !i:Jc7
15...lllb4 occupies the 'hole' cre
ated by White's previous move, but
then 16 l:l.xd8+ ..txd8 17 l:l.d 1 ..te7
18 llld6 e5 ( l 8...b6 19 ..te3! also
nets a pawn) 19lllxc8 l:l.xc8 20lllxa5
wins a pawn.
16 l:l.xd8+ ..txd8
17 l:l.d1
Everything with tempo.
17 ..te7 (D)
18 ltlxf6+?!
White retains the advantage after
this move, but Black could have put
up more resistance. It was probably
stronger to play 18 llld6!. I would
not claim that this is a forced win
(Larsen once observed that all long
variations are wrong!), but Blackis
definitely in real trouble:
l ) 18...b6 19 ..tc6! (not 19 ..te3
because now Black can reasonably
play 19...lllxb5! with a satisfactory
position after either 20 cxb500!
or 20 axb5 transposing to line 2a be·
low) 19...l:l.b8 20 ..tf4! and White's
pieces occupy dominating positions.
2) 18...lllxb5 and now there are
two possible lines:
2a) 19 axb5 b6 (best; 19...a420
llla5! and 19...h6 20 ..txf6 ..txf6 2 1
lllc5 are more promising for White)
20 ..te3 a4 21 llla 1 llld5! 22 cxd5
10. ANAND - INKIOV, CALCUITA 1986 9
.bd6 23 dxe6 i.e? 24 exf7+ �xf7
with fair compensation for the pawn.
2b) 19 cxb5! (this concedes the
d5-square, but Black can't make full
use of it) l9... b6 20 ltlc4 (20 i.e3
�!as in line l) 20... llb8 21 ltld4
(intending lLlc6) 2 I ...i.b7 (2l...i.d7
22 ltle5 is also good for White) 22
l0xb6 i.xg2 23 ltlc4! (although
Black has the two bishops, the mass
of pawns on the queenside is the
most important factor in the posi
tion; 23 ltld7 is less accurate since
after 23...llc8+ 24 �bl i.e4+ things
are getting quite messy) 23...llc8 24
b3 (White only needs to play �b2
and ltlxa5 to decide the game with
his queenside passed pawns; how
ever, Black can try to win a pawn)
24...i.d5 25 �b2! i.xc4 26 llcl !
(when I checked this position with
Fritz, it saidthat Black was winning!
Surprised, I looked to see why)
26...:C5 (this is the reason, but after
some thought I found a solution) 27
llxc4 llxg5 28 :lc8+ i.f8 29 b6!
(Black is helpless) 29.. .:c5 (after
29... ltld7 30 b7 :ld5 31 ltlc6 Black
has no defence to both 32!De7+ and
32 :ld8) 30 b7 ltld7 31 ltlb5! and
Black cannot meet the threat of 32
:ld8.
18 gxf6
19 i.e3 lLlxbS
20 axbS fS!
Black fights back. This gives his
bishop some air and prevents White
forcing the exchange of bishops by
i.c5. 20... e5 is worse, as after 21
i.c5 i.xc5 22 ll:lxc5 i.g4 23 lld5
:lc8 24 b3 Black's queenside pawns
are in trouble.
21 &s (DJ
The alternative 21 f4 ambitiously
attempts to squash Black, but he can
free himself by sacrificing a pawn:
2l...e5! 22 i.c5 (22 fxe5?! i.e6 23
li:ld2 a4 prevents White supporting
his c4-pawn by b3, and then the c4-
pawn itself is attacked by ...i.b4 and
...:lc8) 22...i.f6 (not 22...i.xc5?,
when 23 ll:lxc5 gives White exactly
what he wants) and now 23 fxe5 fails
to 23...i.g5+ 24 �c2i.e6 and Black
is doing well as White cannot hang
on to all the pawns (for instance 25
�c3 :lc8, followed by some combi
nation of ...a4 and ... b6).
21 ••• eS?
The best defence was 2l... f4! 22
i.d4, and now:
I ) 22...e5 23 i.xe5! i.xc5 24
lld8+i.f8 25i.d6 andWhite wins.
11. 10 VISHY ANAND: MY BEST GAMES OF CHESS
2) 22.. .a4 (by not allowing lt:la4,
Black gets some counterplay) and
now:
2a) 23lt:ld3 i.d7 (23...f6? is very
good for White after both 24 i.c5
and 24 tt:lxf4 e5 25 tt:ld5) 24 tt:lxf4
.:.C8 25 i.c3 i.e8 and Black has suf
ficient counterplay.
2b) 23 tt:le4 f5 (23...e5? 24 i.xe5
i.e6 25 c5) 24 i.c5 and White re
tains some advantage.
3) 22...f6 (a solid and sensible
defence; Black prepares ...e5 fol
lowed by ...�f7) 23 tt:la4! e5 24 i.c5
�n 25 i.xe7 �xe7 26 tt:lb6 .Z:.b8
with a slight advantage for White.
Thus 2 l...f4, while not equaliz
ing, would have restricted White's
advantage. The importance of driv
ing the bishop to d4 becomes clear
after the text-move.
22 tt:ld7!
From this excellent square the
knight virtually paralyses Black's
whole army. The game is already al
most over.
22 f4
Too late, as now the bishop need
not block the d-file.
23 i.b6 f6
24 i.c7!
Since 25 tt:lb6 is threatened, Black
can no longer delay capturing the
knight; the result is that White's rook
occupies the seventh rank.
24 i.xd7
25 lbd7 i.cS
26 i.d6 i.xd6
White also wins after 26...i.�f2
27 :xb7 :ds 28 :bS! (not 28c5?
i.xc5 29 i.xc5 .Z:.c8) 28..J[xb829
i.xb8 �f7 30 i.e? and the threatof
3 1 b6 forces Black to approach with
his king, whereupon 3 1 i.xa5 gives
White an overwhelming mass of
passed pawns.
27 :xd6 (D)
Black's position is lost. The active
rook, combined with White'squeen
side pawn majority, guarantees a
straightforward win.
27
28 b3
29 �b2
30 �xb3
a4
axb3
�
Just abandoning the b-pawn, but
30...f5 31 :d7 :bs 32 c5 is also
hopeless.
31 :d7+ �e6
32 lbb7 e4
33 :a7 e3
34 fxe3 fxe3
35 �c3 :ds
12. 36lW
37 bC'i!
38 b7
39 �b4
ANAND - INKIOV. CALCUTTA /986
40 b81t'
41 1tb7+
42 'i!t'dS+
43 c5
lba2
�d3
�e2
1-0
11
The 1987World Junior Championship was a breakthrough for me. It was
already my fourth World Junior Championship; 1 had played in every one
since 1984, finishing lOth in my first appearance, 5th in 1985 and 7th in
1986, a fairlyzigzagging pattern. Istarted the 1987 event in fairly good shape
but I drew 3out of my first 4 games, and at that point I didn't imagine that I
mightwin the tournament However, the following game was a turning point.
13. Game 2
V. Anand - K. Ninov
World Junior Championship, Baguio City 1987
Sicilian, Kan
1 e4 cS
2 lbf'3 e6
3 d4 cxd4
4 liJxd4 a6
s �d3 �cS
6 lbb3 fLa7
7 lbc3 lbc6
8 'ife2 d6
9 h3 �xe3
As a result of this game, playel'll
with Black started delaying this ex
change so as to leave the white
queen on e2 and thereby prevent the
�e2 line mentioned in the note to
Black's 13th move. In time, this led
to the development of a whole new
branch of opening theory.
10 'ifxeJ lbf6 (D)
w
11 g4!
I found this innovation over the
board. PreviouslyWhitehad contin
ued I I 0-0-0, butafter 11...0-0White
either has to prepare this advance
withllhgl, or play g4 as a pawn sac
rifice. The idea behind the immedi
ate g4 is quite simple: if you play it
when g7 is undefended, then it isn't
a sacrifice.
11 •.• bS
Black thought for a while, and
then decided simply to proceed with
his qutenside counterplay. However,
the tempo White saves by missing
out llhgl is quite important, and he
should have tried l l...lDxg4. After
12 'ifg3 lbf6 13 'ifxg7 llg8 14 'ifh6
�d7 Black will play ...'ife7 and both
sides will castle queenside. How
ever, when Black castles queenside
in the Sicilian,White is better unless
there is some mitigating factor,which
is not the case here. Indeed, after
'ife3 by White there is an awkward
weakness on b6 and the d6-pawn it·
self may become vulnerable later.
12 0.0.0 0-0
13 gS lbe8
A poor square for the knight,but
after 13...lbd7 14 �e2 Black has
14. ANAND - NINOV, BAGU/0 CITY 1987 13
immediate problems with his d6-
pawn, due to the exchange of dark
squared bishops and White's extra
tempo. Usually White doesn't go af
ter d6 in a Sicilian, but that doesn't
mean you should forget about the
possibility altogether!
After the text-move, there is no
point to :bgl, which would throw
away the advantage gained as a re
sult r:i White's innovation, so White
has to come up with an alternative
attacking plan.
One possibility is h4-h5 followed
by:dgl. Thearrangement with rooks
on gl and hi is very desirable, be
causetheywillbreak through almost
any kingside defence, but it is very
time-consuming to set up. Unlike
many similar positions in the Sicil
ian, Black's queenside counterplay
is rather slow here, which is the only
reason White can consider this plan,
but in the end I decided on a more
conventional approach.
14 f4 b4
15 lDe2
15 � would be a more posi
tional formula. White stops ...a5-a4
and threatens to invade on b6. After
l5...llb8White can continue 16 e5,
in order to clear the c5-square for the
Imight However, I preferred to play
for the attack, and for that the knight
is needed on the kingside.
15 aS
16 ltlbd4 ltlxd4
17 ltlxd4 (D)
White has the advantage. Nor
mally in the Sicilian, Black's the
matic queenside pawn advance gains
time because it hits minor pieces on
b3 and c3, but here White has evacu
ated these squares quite quickJy.
Moreover, the knight on e8 is very
badly placed for supporting the at
tack. Just about the only useful thing
it can do is to shore up the kingside
by ...g6 and ...ltlg7.
17 ••• 'ilfb6
At the time I felt that Black should
have gone in for 17...a4, but now I
don't think so. The line that worried
me was 18ltlc6 'iWc7 19 o!l'!xb4 a3 (if
Black doesn't play this, then White
plays a3 himself, followed by lilbl
and c3, and Black will never break
through) 20 b3 (threatening to cen
tralize with 'il'd4) 20...'iWc3 (after
20...llb8 21 o!i)a6 .>.xa6 22 .>.xa6,
followed by .>.c4, White's queen
side position is solid since the poorly
placed knight on e8 cannot displace
15. 14 VJSHY ANAND: MY BEST GAMES OF CHESS
White's bishop), but in fact 21 Wei!
ifh2+ 22 'it>d2 is clearly better for
White- Black has surprisingly little
compensation for the pawn.
These were still the pre-computer
days, when players were much more
intuitive. Nowadays everybody goes
home and checks everything with
Fritz. The use of computers has made
people more sceptical and now they
are more prone to go pawn-grabbing
unless there is definite compensa
tion.
18 eS .i.b7
19 mtfi(D)
19 dxeS
Or 19...a4 (1 9...:d8 20 fS! is simi
lar) 20 fS! dxeS (20...exfS 21 e6 is
very good for White) 21 fxe6! (not
21 'ibeS :aS!) 2l ...exd4 221Fh3 g6
(22...fS 23 l:txfS! gives White a win
ning attack) 23 exf7+:xt7 (23...�h8
24 'i'h6 �g7 2S :C6 wins) 24 :Xf7
�xf7 25 'i'xh7+ winning the black
queen.
20 fxeS l:td8? (D)
Allowing a pretty finish. 20... g6!
was best, when White could con
tinue21 �S 'llrxe3+ 22ll'lxe3 with a
pleasant endgame-his knight could
head for either f6 or d6. Still, Black
has some chances by playing his
knight to fS, and he shouldcertainly
have gone for this.
21 .i.xb7+! �xb7
22 g6+
Once again Black suffers because
of his miserable knight position
Here it prevents Black from playing
22...fxg6.
22 ••• �g8
Or 22...'>1i>xg6 23 'i1Vd3+ (stopping
...�h7) 23...fS (23...�h6 24 'i'h3+
�gS 2S :gl + �f4 26 :del with
mate next move) 24 exf6+ with ade
cisive attack.
23 'i1Vb3 .!ffii
23...fxg61oses to 24 l:txfl!+ 'iPxfl!
2Sll'lxe6+.
24 exf6 (D)
16. 8
ANAND- NINOV, BAGU/0 CITY 1987 15
It isn't every day you see two at
tacking pawns on f6 and g6! When
you get a position like this, you go
away feeling very pleased and have
a warm glow for the next few games.
24 ••• fxg6
25 fxg7 1-0
Since 25...ii>xg7 26 ltlxe6+ and
25...llxfl 26 'l'h8+ �f7 27 :Xn+
are decisive.
After this game, which was from round 5, there was a free day and the
players went on an excursion. I still hadn't broken free from the pack. The
following day, however, saw another good result. I was facing Agdestein,
who at2565 was the highest-rated player in the tournament, although he was
not the only grandmaster (Ivan Sokolov was also participating, although he
was not very successful).
Iprepared as well as I could because Agdestein is very unpredictable in the
openings, and sat down hoping to play a good game.
17. Game 3
V. Anand - 5. Agdestein
World Junior Championship, Baguio City 1987
Ruy Lopez
1 e4 tt'lc6
A provocative move. At the time
it was felt that this move had little in
dependent significance because so
long as White knew he shouldplay 2
tt'lf3, Black had nothing better than
2...e5, thereby transposing into stan
dard king's pawn openings. Subse
quently it was discovered that Black
could wind White up even more by
playing 2...d6, but I don't trust this
for Black!
2 tt'lf3 eS
3 i.bS a6
4 i.a4 b5
5 i.b3 tt'la5
Agdestein is fond of offbeat sys
tems. I didn't know much about this
one, although it is popular amongst
Norwegian players. I could only re
member a game between Spassky
and Taimanov (in fact from the 1955
USSR Championship, held in Mos
cow). Nevertheless, I was quite happy
to see it on the board. White can play
natural moves and there is not much
risk even if he commits a slight inac
curacy - a pleasant situation when
facing the top seed!
6 0-0
7 d4
d6
tt'lxb3
8 axb3 f6
9 tDc3 i.b7
I recalled that Spassky had played
lO tt'lh4 in the above-mentioned
game, with dxe5, 11Vf3 and l:tdl fol
lowing in some order, and that later.
Spassky sacrificed a piece by means
of lL!xb5.
10 tt'lb4 tt'le7 (D)
11 dxe5! dxe5
IfBlack plays 11...fxe5, then White
can strongly reply 12 f4,opening the
position up while Black's king is
still stuck in the centre. However,
taking back with the d-pawn retains
control of gS, so that f4 can be met
by ...exf4 followed by the fork ... gS·
12 1Wf3
18. ANAND - AGDFSTE/N, BAGU/0 CITY 1987 17
White avoids the exchange of
queens as most of his chances lie in
exploiting Black's poor develop
ment and centralized king.
12 •.• 'i'd7
13 l:.dl 'i'e6
By now I had worked out that this
was indeed what had happened in
the Spassky game, and that he now
continued with 14 i.e3. For a mo
ment I wondered whatTaimanov had
played that allowed the sacrifice on
b5, then! realized that it was 14...g5.
After 15 �xb5! axb5 16 W'h5+ 'ilf7
(16...�g6 17 �xg6 'l/lf7 18 l:.xa8+
.baS 19 1Wg4! also wins) 17 ltxa8+
.ixa818l:.d8+ �xd8 19 'ilxf7 gxh4
20 'l'xf6 White had a winning posi-
lion.
However, I was worried by the
idea of 14...h5!?, intending ...'I'g4. I
mulled over this for some time, but
couldn't see an easy answer. If
White plays 15 h3, then 15...g5 is
now possible because White has no
queen check on h5. If 15 li:ld5, then
Black just castles queenside. It ap
peared to me that if White was going
to play 00, then it would be better
to do it straight away, before Black
bad time to set up the threat of
...'l'g4.
14 li:ldS li:lxdS
If 14...0-0-0, then 15 c4 and White
is already starting to make Black's
king feel insecure.
IS exdS 'l'f7
16 c4?! (D)
A slight inaccuracy which gives
Black the chance to sideline the
knight on h4. 16 li:lf5 would have
been more accurate, because White
can play c4 at any time-Black can't
prevent it. Then 16...g6 17 li:lh6
.ixh6 18 .ixh6 0-0-0 19 c4 ltd?
would have led to a position in which
White has an edge, since Black has
no really constructive plan.
16 .ie7?
Black misses the opportunity he
has been given. 16...g6! would have
left the h4-knight misplaced It is
true that after 17 'l'e2 .ig7 18 f4 0-0
19 f5, follow� by 'l'e4, White has a
good centralized position and Black's
bishops are hemmed in. However,
after 18...0-0-0!? Black would have
chances of puhing his two bishops to
work. The b7-bishop functions as a
kind of 'Dragon' bishop, in that it
makes it hard for White's attack
down the a-file to strilce home prop
erly. Moreover, playing cxb5 will
19. 18 VISHY ANAND: MY BEST GAMES OF CHESS
leave d5 hanging.White is certainly
not worse, but Black would have far
more counterplay than he obtains in
the game.
11 lbrs (DJ
The exchange of inaccuracies has
favouredWhite, because Black's in
accuracy is actually quite serious -
his bishop doesn't belong on e7 at
all. NowWhite is clearly better.
17 ... J:d8
17...0-0 is impossible because of
18 ..th6.
18 ..te3 g6 (D)
More or less forced, because oth
erwise Black lacks a constructive
move. 18...0-0? still fails to 19 ..th6
while after l8.....tc8 White has a
choice of promising lines:
l) l9l:ac l 0-0 20 lbxe7+(not20
cxb5 ..txf5 21 'i'xf5 axb5 22 J:xc7
J:xdS! andBlack escapes) 20...'i'xe7
21 cxb5 J:fe8 22 b6 cxb6 23 ..txb6
J:d7 24 'i'd3! ..tb7 25 d6givesWhite
a clear advantage.
2) 19 c5! (even more forcing)
l 9.....i.xf5 (19...0-0 20 d6 wins a
piece) 20 11Vxf5 :Xd5 21 'l'c8+and
after 2l.....td8 2211fxa6 or 2I...l:d8
22 J:xd8+ ..i.xd8 23 l:ldl 'ife7 24
1!fxa6 White wins a pawn while re.
taining a positional advantage.
19 lbh6!
White could exchange on e7, but
then Black's king could castle or
moveton and he wouldhavefair de·
fensive chances.White's qu.enside
majority is an asset, but it will not
win the game by itself because the
c-pawn is tied to the defence of d5.
Instead, I wanted to keep Black's
king on e8, while I broke through on
the c- and d-files.
19 'fig7
Intending ...f5, when the knighl
might be in trouble.
20 'l'g3?!
This move, introducinglbf5 ideas,
is not so strong as I imagined during
the game. Sometimes, when you
20. ANAND - AGDESTE/N, BAGU/0 Crrr 1987 19
havea good position, the temptation
is justto play easy, comfortable moves
and wait for the position to win it
self. Thecorrectmovewas 20 cxb5!,
when 20...axb5 fails to 21 :a7. Dur
ing the game I rejected it because of
20...f5 21 bxa6 .i.a8, with the threat
of ...f4. However, with ten years'
hindsight I don't see Black's com
pensation for the two pawns. In par
ticular,once White has played a7 the
bishopon a8 willeffectively be dead,
sincetaking on d5 will always allow
a combination involving the promo
tion of the a-pawn. One line is 22
'tlh3! f4 23 .i.d2 .i.g5 24ll:lg4 h5 25
�xe5 and wins.
It is worth noting that 20 h4 was
playable. After 20....ic8 we trans
pose into the game, but White has
saved a couple ofmoves.
20 ... .i.c8
Threatening ...g5, soWhite's next
move is forced.
21 h4 .id6 (D)
If Black attempts to play actively
by 2I...f5 22 .i.g5! f4 then:
I) 23 'i'c3 b4 24 'ilfd2 (24 '6'f3
.txg5 25 hxg5 'fle7 26 ll:lg4 '6'xg5
27 �xe5 0-0! is unclear) 24....txg5
25 hxg5 'i'e7 26 .l:tel �f8! and
Black has counterplay.
2) 23 •h2!. A paradoxical move
which nails down Black's kingside.
Now White simply threatens :el
followed by doubling or even g3
(meeting ...f3by g4). However Black
continues_ his e5-pawn will come
under fire and his king is trapped in
the centre. The point of .wt-.2 is that
after 23....i.xg5 24 hxg5 '*le7 White
can defend the g5-pawn and main
tain his kingside bind.
w
22 '*if3
Other moves are inferior, for ex
ample 22 c5? .txc5 or 22 lLcl e4
23 .tf4 '*ixh6! 24 .txh6 .txg3 25
fxg3. Inthislatter lineWhite keeps a
slight edge as the opposite-coloured
bishops (which normally have a
drawish influence)makeBlack's de
fence a bit more difficult, whenthere
are still a fairnumber ofmajorpieces
on the board. However, I thought
that keeping the queens on offered
even more.
The move 'lin is mainly to pre
empt the threat of ...e4. Now that
White has got the knight firmly en
trenched on h6, he can proceed with
:aci followed by c5 and d6. The
manoeuvre '6'f3-g3-f3 may appear
odd, but Black has also wasted time
21. 20 V!SHY ANAND: MY BEST GAMES OF CHESS
with the manoeuvres ...iLb7-c8 and
...iLf8-e7-d6.
22 iLe7
Black would like to play ..J:I.f8, but
the immediate 22...l:l.f8 allows 23 c5,
because after 23...iLxc5 24 i.xc5
the rook on f8 is hanging. So Black
returns to e7, still preventing c5 and
getting ready to play ...l:l.f8 next
move. The net effect of both players'
oscillations has been to add the
moves h4 and ...iLcS, which benefits
White.
At this stage I felt that my posi
tion was very comfortable, but I had
(and still have) a great deal of re
spect for Agdestein and I wasn't
counting on victory yet.
23 l:l.acl
Retaining the option of either cxb5
or c5 followed by d6.
23 ••• bxc4
Agdestein finally decides to re-
move the cxb5 option.
24 bxc4 :t8
25 c5 f5
26 iLgS iLxgS
27 hxg5 'ii'e7
28 'l'g3! (D)
White had a more complex alter
native in 28 d6 'ii'xg5 29 'l'c6+ i.d7
(29...l:l.d7 30 'ii'a8! l:l.d8 31 :tel
should win for White) 30 'ii'xc7
'l'xh6 31 c6. This would also have
been quite promising, as Black would
have to return the piece for one
pawn, still leaving White with a dan
gerous passed pawn. However, given
that an effective, solid alternative
existed, I preferred to play safe.
In a way it is quite strange that
White is not winning already, since
Black's king is irrevocably since
in the centre - not only call it not
castle, but even f7 is denied to it.
However, the reason is the knighton
h6. It is of course doing a wonderful
job, but when it comes to landing a
killing blow in the centre,White is
effectively a piece down.
28 f4
29 'ii'h4 e4
30 d6 'ii'eS!
The best chance. After 30...cxd6
31 cxd6 'ii'e6 (3l ...'l'e5? 32 llxcS!)
32l:l.c7 e3 (32...iLd7 33 �g4) White
can continue 33 fxe3 fxe3 (after.
33...'1'xe3+ 34 �hl there is node·
fence against 35 :te l) 34 'l'c4!l:l.f2
35 'ii'xe6+ iLxe6 36 l:l.e7+ �f8 37
l:l.xe6 with an extra piece.
31 dxc7 llxdl+
32 :Xdl (D)
22. ANAND - AGDESI"EIN, BAGUIO CITY 1987 21
32 e3?
Blackdecidestocounterattack, but
now White's win is fairly straight
forward. Black's only chance was
32...Wxc7, when we can see that it is
not so easy for White because his
knightisonh6.Ifit were on any nor
mal square then, for example, lZc3
or �4. heading for d5 or d6, would
be decisive. Nevertheless, after 33
ll:lg4 (heading for f6) 33...i.xg4 34
Wxg4 :tf5 35 b4! (White needs to
pause for this move; it strengthens
c5 andindirectlysupports:td6; after
35 'l1Ve2lhc5 36'liV:u4+Wt7White's
attack has got a bit stuck) White has
a large advantage. He intends 11'e2
and :td6, taking aim at e4 and a6,
when Black's exposed king causes
continuingproblems.lf35...e3, then
36 fxe3 fxe3 37 11Fe4+ picks up the
e3-pawn (37....:.es 3S 'IlVaS+ �7 39
'iiVhS is even worse).
33 l:td8+ 'lie7
34 ltlg8+ lhg8
35 lhg8! ..te6
Or 35...e2 36 11'xh7+ 'lie6 37
.l:l.eS+and wins.
36 c811'
36'l1Vxh7+ J..f7 37 .l:l.eS+<lixeS 3S
cS11'+ �7 3911'b7+ <lids 4011FhS+!
is a prettier win, but I preferred the
prosaic text.
36 ••• exfl+
37 •xn!
This game has been published in
some magazines with the move 37
<lixf2, but thatallows mate in two!!
37 ..bc8
38 lhc8 1-0
This win over the top seed put me wellon the way tobecoming World Jun
iorChampion.My run ofwins continued with further victoriesoverKlinger,
Ivanchukand Blatny, after which I was in clear first place, aposition I heldon
to until the end of the tournament.
Winning the World Junior was my big breakthrough. Nonnally a player
from India would have to waste a lot of time playing in mediocre open tour
naments, gradually improving his Elo rating and hoping to get some invita
tions. However, the two factors of my World Junior title and my GM title,
which I gained shortly after this event, enabled me to short-circuit the pro
cess. I got an invitation to a pleasant open tournament in Lugano (but after
23. 22 V!SHY ANAND: MY BEST GAMES OF CHESS
my play there, I didn't get invited again!). Then I was asked to bea commen
tator at the World Cup event in Brussels (April1988). In Brussels I was able
to meet some tournament organizers and this led to my first top-class invita
tion- to Wijk aan Zee 1989.
I had a tough year in 1988, just after gaining my GM title. I lost about 40
rating points in two tournaments at Biel and Blackpool- I still don't under
stand why, although many GMs have told me that they had similar experi
ences after gaining their title. I had a break after Biel and returned to active
play near the end of the year at the Thessaloniki Olympiad, making 8lh
points out of 12 games. This score was sufficient to regain 10 of the Elo
points I had lost, but later I was surprised to discover that FIDEhad notgiven
me any points at Thessaloniki. The reason, I found out, was that the Indiau
team had arrived late, after the first round, and had therefore lost one match
by default. FIDE had counted this as a normal loss, thereby wiping out my
ten point gain from the rest of the event! Later on I had my ten points rein
stated.
I then scored 6/9 at the GMA Open in Belgrade and subsequently I played
at Reggio Emilia, another good invitation resulting from my World Junior
success. This tournament started well with wins against Ivanchuk and Sax,
but after that I lost some games and finished on 4/9. However, this event was
very useful for me; it was the first time that I had played such a strong field.
My next event was the long-awaited trip to Wijk aan Zee. This was my first
really big event and I was thrilled to be there. I had a zigzag course in the
tournament. I won my first two games,just as at Reggio Emilia, then lost to
Tseshkovsky (who used to be a nemesis of sorts!). lalso lost to Van der Wiel,
but then won against Ivan Sokolov. I was still on '+I' when the following
game was played in the penultimate round.
24. Game 4
V. Anand - J. Benjamin
Wijk aan lee 1989
Sicilian, Richter-Rauzer
1 e4 cS
2 lt:f3 d6
3 d4 cxd4
4 tt:lxd4 tt:lr6
s w tt:lc:6
6 �gS e6
7 'l'dl i.e7
8 0-0-0 0-0
9 tt:lb3
Two of the first fourgames in the
bookreachthis position!I have done
pretty well with this line so far.
9 •.• 'il'b6
By the time this game was played
it hadbecomeclear that 9...a5 wasn't
that great a move.
10 f3 IUS (D)
11 <!Jb1
There is an interesting story relat
ing to this move. I used to get a lot
of my theoretical infonnation from
Ivanchuk. For example, during the
1985 World Junior Championship in
Sharjah he had shown me a tremen
dous idea in theDragon, whichturned
an existing evaluation upside down;
I was amazed (and thankful!) that he
was so open and generous about
showing his ideas to me. I would try
to give him sometitbit in return, but
unfortunately my novelties were not
that good!
At the Reggio Emilia event the
month before Wijk aan Zee, there
hadn't been much to do in the eve
nings. so one day I went to Chucky's
room. By this time he had recovered
from his first round loss to me and
was on a respectable score. He was
also feeling bored; we went outfor a
walkand then returned tohis room. I
asked him "Why does everybody
play the Catalan - it seems such a
boring opening." He replied that it
was not boring at all and proceeded
to show me an interesting idea. The
introductory moves were 1 d4 lilf6 2
c4 e6 3 tilf3 d5 4 g3 i.e7 5 .i.g2 0-0
6 �3 dxc4 7 lL!e5 c5 8 dxc5 and
25. 24 VJSHY ANAND: MY BEST GAMES OF CHESS
now Vassily said that 8. ..'ifc7 was
supposed to be a good move. He ex
plained that after 9 �xc4 'l'xc5...
(D)
...everybody plays 10 'l'b3, but he
had found a very strong idea: 10 b3!.
The point is that after IO...:ds White
can play 1 1 .llLa3 :Xdl+ 1 2 J:l.xd1
'ilt'c7 13 �b5 trapping the queen. I
was very impressed by this line,
which I had never seen before. He
explained that this was the reason
why ...'ilt'c7 was not good in this par
ticular line of the Catalan, but was
good in the similar lines in which
Black plays ....llLb4 and later retreats
the bishop to e7 in response to a3.
The reason, of course, is that the
pawn on a3 prevents the move i.a3.
A nice idea, but apparently not much
use to me as I didn't play the Cata
lan. I hope Chucky will forgive me
for revealing this piece of analysis!
Returning to the Benjamin game,
at this time I didn't really study the
openings too deeply. I looked at all
the theory, but didn't really go be
yond that. After lO...:ds I suddenly
didn't feel very happy with my posi.
tion and couldn't fmd a continuation
I felt comfortable with. Then I sud
denly brightened up, becauseI sawa
little trick after 11 q;,bI.
11 ••• ciS?! (D)
If Black plays ll...a6, then 12
.i.e3 Wfc7 13 'ilf2 and White gains
time owing to the threat of i.b6, so
q;,b1 is really to provoke ...a6. Later
on it was discovered thatthis doesn't
really matter, because Black'scoun
terplay consists of ...a6, ...lOfl and
...b5 in any case, but we didn't know
that in 1989!
After 11 Wbl Benjamin looked
surprised, bKause he didn't under
stand the point of the move. Perhaps
he was thinking 'Vishy doesn't know
that Black's threat is...d5'. ,
when he played ...d5 I checked my
idea carefully, although there isn't
much choice because after anything
else White is clearly worse.
26. ANAND - BENJAMIN, WIJK AAN ZEE 1989 25
12 .hr6 dxe4?
Benjamin falls for it hook, line
and sinker. He could still have bailed
out by 12...i.xf6 13 exd5 i.xc3 14
•xc3 exd5, although after 15 'il'c5
or 15 i.d3 White has some positional
advantage. However, as he admitted
after the game, he simply hadn't seen
the idea at all.
13 i.xe7
Not 13 i.d4 .fud4 14 �d4 e5.
13 •.• :Xd2
14 lilxd2! (D)
14 exf3
After 14...lilxe7? White wins by
15 lllc4 'ilc7 16 lllb5 and Black's
queen is lost almost exactly as in Iv
anhuk'sCatalan idea. It isn't trapped
here as it was in the Catalan, but the
threat of mate on d8 means that it
amounts to the same thing. At the
next tournament where I met Ivan
chuk, he came up to me and said "I
see you used my idea in the Cata
lan!".
After the game Benjamin said that
he wanted to resign at this point, but
decided to play a few more moves;
however, I then staned to play so
badly that he couldn't bring himself
to resign any more.
15 gill?
Here's the bad move. Later I real
ized that I5lilc4! would have won on
the spot. 15...'1ff2 loses to I6lile4,
and after 15...'1fc716 �d6 fxg217
�xg2'1Fd818�g3'1Fe719l:l.hel all
White's pieces are active and his
knights are going on a queenside
rampage (lilb5 or llld5, coupled
with ltld6). The factthatWhite has a
lost a pawn is irrelevant.
15 ••• e5
16 �h4?! (D)
Another mistake. 16 �a3 would
have been much better, keeping con
trol of key squares such as c5 and d6.
In this case White shouldstill win in
the long run, although thanks to
White's previous error it is likely to
be a laborious process.
27. 26 VISHY ANAND: MY BEST GAMES OF CHESS
16 ... .te6
17 lMe4
White already has to take care:
not 17 .i.c4? .i.xc4 18ltJxc4 'iWb4.
17 lM4 (D)
18 .i.g2?!
After this, the position becomes
rather murky, although it is possible
that White can still retain an advan
tage by accurate play. 18 .i.f2 was a
much safer way to prove that White
is better.
18 ••• .l:c8
Suddenly I couldn't find a line
that worked for White. The threat is
...f5 followed by ....l:xc3, and it isn't
easy to find a good defence. Finally I
found a line that seemed to work for
White but I was very nervous be
cause Black has all his pieces aimed
at White's queenside.
19 .i.f2 fS!
20 f4!
Not 20 ltJg5? .l:xc3 21 ltJxe6
1Wxe6 22 .i.xd4 (22 bxc3? 1Wb6+ and
Black wins) 22...exd4 23 bxc3 dxc3
24 q;>a1 11te2! (24...'itb625l:lbl 'iff2
26 .i.fl 11txc2 27 .i.c4+ followed by
:Xb7 wins for White)25.l:hgt11txc2
26.1:bl1ltd2 27 a3 and Black is even
slightly better.
B
20 fxe4
21 fxeS (D)
21 ••• :C4
2 l ....l:d8! would have made life
much harder for White:
1) 22 ltJe2? .i.xa2+ 23 �xa2
'li'a6+ wins for Black.
2) 22 .l:he1?! .i.g4! 23 .l:d2'1116!
24 .i.e3 (24 .l:xd4 .l:xd4 25 .i.xd4
'Wd2 favours Black) 24...Wh4 25
.i.f2 with a draw by repetition.
3) 22 .l:d2! .i.f5 (not 22...ltJb3?
23 axb3 and White wins) 23 .i.e3!
'ilrg624 :hd 1 and, surprisingly, there
is little Black can do to prevention
followed by taking on d4.
While 2I. ...:I.d8 might not have
been any better than the text-roove
against perfect play, 22l:td2! and 23
28. ANAND - BENJAMIN, WIJK AAN ZEE 1989 27
.A.e3! aren't easy moves to find over
lhe board.
22 l:tbel? (D)
Missing 22 &2! l:l.a4 (Black's
sacrificialattempts fail, for example
22...'l'xb2+ 23 �xb2 l:txc2+ 24
�al �xe2 25 i.el or 22...l:l.xc2 23
bd4 'il'c6 24 �4 and White wins
in both cases) 23 b3! l:txa2 (if
23....1xb3, then simply 24 axb3) 24
.1xd4 and wins.
B
22 ••• .11b4?
Benjamin was in time-trouble by
now, and commits another mistake.
The best line was 22....ig4! 23 .!:d2
(not 23 l:txd4? l:txd4 24lt:lxe4 l:txe4
25 .1xb6l:txel#) 23...'fih6! 24lt:lxe4
(or 24 :Xd4 l:txd4 25 i.xd4 'ilt'd2)
24...�6 and Black has avoided los
ing a piece. Despite Black's slight
material advantage, I don't think
White is orse, since B ack' k
fairly exposed and his pieces dis
jointed.
23 '.tel!
It would also have been good to
play 23 b3!, which looks a bit para
doxical as Black has so many pieces
ready to sacrifice on b3. However,
after 23....i.xb3 24 axb3 l:txb3+ 25
<t>cl White should win.
23 ... i.g4
If 23...lt:lb3+ then 24 axb3 1Wxf2
25 lld8+ �f7 26 :n picks up the
queen.
24 .!bdS 'ilt'cS
25 lt:lxb4 .i.xdl
26 �xdl! (D)
Not 26 l:l.xdl? �e2+ and Black
wins, nor 26 l:txe4lt'lb3+! 27 <t>xdl
'ilt'xf2 28 axb3 'ilt'xg2 29 l:te2 and
only Black can be better.
After the text-move it suddenly
dawned on me that I was completely
winning. I had been struggling to
contain Black's queenside initiative
for so many moves that the realiza
tion caught me by surprise!
26 e3
26...'ili'xb41oses to 27 l:txe4.
29. 28 VISHY ANAND: MY BEST GAMES OF CHESS
r1 lhe3 lMS
28 �dS+
An important intermezzo. 28 .l:tf3
is amistake because of 28...'i!fxb4 29
:Xf5 W'g4+.
28 �
Forced; 28...Wh8fails to 29 l:l.c3
29 .:tf3 'il'xb4
30 .:txfS+ WeB
31 e6 1-0
After 31...11Vxb2 32 �c5 it is all
over.
Before the last round I was in a tie for fiCSt place with Nikolic, Ribli and
Sax. I had the feeling that a draw in the last round would be enough to main
tain the status quo, and indeed when my game with Douven ended in a draw
my feeling turned out to be justified, as the other leaders also drew.Thus we
ended up in a four-way tie for first. It was my first success in one of the
world's major international events.
In 1989, Bessel Kok, thechairman of SWIFf and agreatchess benefactor,
organized a Youth vs Veterans event in Cannes not long afterWijk aan Zee.lt
was held in conjunction with the annual Cannes Games Festival and was a
very enjoyable event. The veterans were Tal, Spassky, Andersson, Csom and
Larsen; the youth team consisted ofLautier, Renet, Adams, Miralles and my
self. As the only GM in the youth team, I was acconunodated in an excellent
hotel; indeed, the conditions at this event were the best I had experienced un
til then.
30. Game 5
M. Tal - V. Anand
Youth vs Veterans, Cannes 1 989
English
I met MishaTal forthe first time when I visited theWorld Cup tournament in
Brussels in 1988. He was the most popular player by far, captivating every
one with his personality and his brilliant chess. This was my first game
against my childhood hero and I was obviously quite excited.
1 c4 cS
2 ltl£3 lt:c6
3 00 lt:d4
How do you explain a move that
violates the rule not to move the
same piece twice in the opening?
Well, I can't really find a general
principle thatjustifies it, but it does
seem to work! One possible expla
nation isthat when White recaptures
onf3 with his queen, he has lost con
trol ofthe important d4-square. Then
Black can bring out his other knight
via h6 and f5 to fight for d4.
Inany case, 3...�4 is a provoca
tive move which leads to a more un
balanced type of position than is
usual in the Symmetrical English.
4 e3 lt:xf3+
S 'i'xf3 g6 (D)
6 b3
Or 6 d4 .i.g7 7 dxc5 (after 7 'Wd1
�f6 White will have to play d5 and
then lose a tempo with e3-e4 in order
to get his customary space advan
tage) and now:
1) 7... 'i'a5 8 e4! 'i'xc5 9 lt:d5!
(stronger than 9 .i.d3 .i.xc3+ 10
bxc3 d6) with a slight advantage for
White.
2) 7....i.xc3 +!? 8 bxc3 11fa5 9 e4
11fxc5 10 .i.d3 d6. Normally Black
should not give up his bishop like
this in an 'Indian structure', but due
to White's doubled c-pawns this po
sition should be compared with the
Nimzo-Indian rather than the King's
Indian Defence. Admittedly Black
has already played ... g6, but it's not
clear how White can exploit this.
31. 30 VISHY ANAND: MY BEST GAMES OF CHESS
6
7 .i.b2
8 g3
9 .i.g2
Two years later, I beat Karpov
with 9...llli'6; the tactical justifica
tion is that 10 liJd5 .i.g4! equalizes
immediately. Actually, it is much
more natural than 9...ltJh6. Whatever
can be said about the objective mer
its of 3...liJd4, I scored 212 against
Tal and Karpov with it!
10 'ild1 0-0
1 1 0-0 .i.d7
To supportthe ...b5 advance.
12 a4
After 12 d4?! Black's plan of ex
erting pressure·on d4 comes to frui
tion: 12...cxd4 13 exd4 llli'5! 14 d5
(14 liJd5 b5! gives Black the edge)
14...b5 ! with good counterplay.
B
12 .i.c6
13 d4 .i.xg2
14 �g2 (D)
14 l:l.c8
After 14...cxd4 15 exd4 ltJf5 16
d5 a6 17 l:l.bl ! the position is slightly
better for White. He has a space ad
vantage and a weak black e-pawn to
play against On the otherhand, his
pieces on the queenside are awk
wardly placlld.
15 'ild3?!
Tal latermentionedthe possibility
of 15 d5! and this move does seem
to offer White a stable if minuscule
edge. Black will hardly be able to
achieve ...b5 and playing ...c7-e6
would create weak pawns on e6 and
d6.
15 - ad4
16 exd4 lllrs
NowBlackis fine.
17 d5 (D)
After 17 lDe2d5 18 c5 a5 Black is
slightly better. White's queenside
pawns are crippled and the pawn on
d4 is weak. White's best line was
probably 17 �d5! e6 18 ltle3, just
playing for the exchange of knights
and equality.
B
32. TAL - ANAND, CANNES 1989
17 1fb6 25 c5
31
The reason why 'ifd3 was bad - There is no other defence against
b3 is undefended. 25...ll:ld4.
18 t0d1 i.xb2 25 ••• 'ifxcS
19 <fub2 eS! 26 'ffxb7+ .l:.c7
Now Black is justified in playing 27 'lfdS 'ffb4
actively. The open f-file counts for Not 27...'1Vxd5+'? 28 l:txd5 l:tc2
morethan Black's slightly weak cen- andWhite can hang onwith29 ll:lc4.
tral pawns. 28 lUd1 .l:.cS
20 dxe6 Exploiting the weakness of b3.
Forced. If White doesn't do this, 28...ll:ld4 is less clear after 29 ll:lc4
Blackcan aim for...e4-e3. Moreover, l:tc5 (not 29...ll:lxb3'? 30 ll:lxd6!) 30
his knight can settle comfortably on 1Wa8.
d4. 29 'IVai
20 ••• fxe6 If 29 ll:ld3'? then 29...ll:le3+! 30
21 :Sd1 ltt6 fxe3 11Vxd2+ winning the exchange.
Preparing to double rooks on the 29 ... 'lixb3
f-lile. 30 ll:ld3 (D)
22 :d2 eS! (D) After 30 '1fxa7+ l:tf7 31 11t'a6 (3 1
23 11Fd5+
After 23 ll:ldI ltld4 Black wins a
pawn as White has to meet the threat
of 24...'i'c6+.
23 ...
24 'ltbs
'lfa8 l:tc2 is similar, e.g. 32 ll:ld3 'lic3
33 l:txc2 1Vxc2 winning material)
31...l:tc2! 32 ll:ld3 l:txd2 33 l:txd2
'lid5+ 34 �gl e4 White loses the
pinned knight.
30 l:tc2?!
33. 32 VISHY ANAND: MY BEST GAMES OF CHESS
30...l:l.c7! would have been bru
tally effective- Black simply main
tains his a-pawn and threatens both
3l ...'lWxa4 and 3 l...�d4.
31 'lWe4?!
White could havemadelifeharder
by playing 31 'lWxa7+ .l:l.f7 32 'lWa5
(D) (32 'liaS 11fc3! wins). I don't
know ifTal spent much time consid
ering this apparently greedy con
tinuation. I must admit that it was
only after some time checking the
position with Fritz thatI began tore
alize that Black's task was not so
easy. Still, Black does have a way to
win:
1) 32...e4'!! and now:
Ia) 33 .l:l.xc2 'i'xc2 34 .l:l.d2 'ifc6
35 �b4 'ilfc3! wins for Black.
lb) 33 �f4 'iff3+ (33...l:l.c5 34
1Wd8 is unclear) 34 �gl e3 looks
strong, but after 35 fxe3 'ifxe3+ 36
�hI there is nothing clear-cut, for
example 36...'iff3+ 37 �gl �d4 38
.l:l.el .
lc)33.1Wb4! 1Wxb4 34 &b4 saves
White.
2) 32...1i'h7+?! 33 �gl lbd2 34
1Wxd2 lDd4 35 �1 andWhite avoids
disaster.
3) 32...l:rxd2 33 'lWxd2 (33 l:txd2?
1i'b7+ 34 �gl 'lWbl+ 35 �g2 ta3+
wins) 33...1Wxa4 wins a pawn, but
there is still a long way to go.
4) 32...�3+! 33 fxe3 Vxd3 34
.l:l.xc2 ile4+ 35 �gl (35 �h3 :t'5!
and there is no way to stop ...l:l.hS#)
35...11'xe3+ 36 �g2 'lWf3+ '57 Wh3
1rxdl and White's king is hope
lessly exposed.
31 _ .l:l.c4
32 'lidS 1Wd
Besides the extra pawn, the differ
ence in strength of theknights can be
seen.
33 .l:l.b2 :.14
34 .l:l.b7+ wh6
35 'lrbS (D)
35 'i6'gS loses to 35...'i6'c6+.
35
34. TAL - ANAND, CANNES 1989 33
36 <i>gl
Or 36 fxe3 'l'c2+ 37 �h3 :h4+!
38gxh4:f3+ 39 �g4 'l'g2#.
36 -·
'fic2
37 lin llxd3
0-1
After 38 1i"d7 Black mates by
38...1Fxf2+!.
I was very happy to win in Tal's own style, although I recognized that his
poor play in this game was due to illness. In fact, he withdrew from the tour
nament after the first half for medical treatment. Still, I am proud to have
played at least one game against Misha Tal.
35. Game 6
V. Anand - B. Spassky
Youth vs Veterans, Cannes 1989
Ruy Lopez, Breyer
1 e4 eS
During the course of his career,
Spassky has played just about every
opening there is, but in recent years
he has tended to stick to dual king
pawn openings, so this was nota sur-
prise.
2 ttlf3 ttlc6
3 �bS a6
4 �a4 ttlf6
S 0.0 �e7
6 l:e1 bS
7 �b3 d6
8 c3 0-0
9 h3 ttlb8
The Breyer Defence, for decades
Spassky's main weapon against the
Ruy Lopez.
10 d4 ttlbd7
11 c4
Round about this time there had
been a modest revival in this old
move. I felt that Spassky's knowl
edge of the main lines would be
much greater than mine, so I tried,
with some difficulty, to find a rela
tively unexplored continuation. 1 1
c4 enjoyed popularity in the 1960s
and was probably no surprise, but at
least it didn't have such a huge body
of theory as 1 1 ttlbd2.
11 ••• c6
12 a3
This was an idea I came up with
over the board At the time I hadno
idea if it had ever been played be
fore, but I have since found acouple
of earlier games. The immediate 12
ttlc3 is met by1 2...b4.
12 ..• bxc4!
The best reply. After1 2...b7 ll
ttlc3 I think White is slightly better,
since White's structure is a bit more
comfortable in this type of 'Old In·
dian' position. Black can of course
continue with ..."Wic7, ...l:ac8 and so
on, but White can gain more space
with an eventual �a2 and b4.
13 hc4 (D)
8
36. ANAND - SPASSKY, CANNES 1989 35
13 ••• dS
Black is not worse after the text
move. but 13...li:lxe4 would have
equalized straight away:
I) 14 dxe5?d5 15 .i.xd5 cxd5 16
1l'xd5 l:lb8 (16...li:lxf2 may also be
good for Black) 17 'ilfxe4 li:lc5 18
'lre2 ltlb3 19 l:la2 .i.e6 and White
loses the exchange by force.
2) 14 :.Xe4 d5 with two possibili
ties:
2a) 15 ltlxe5 (Spassky suggested
this, but Black has an escape route)
15...dxe4 (not 15...li:lxe5 16 .i.xd5
cxd5 17 :XeS.i.d6 18 :eJ and White
can behappy because, compared to
theMarshall Attack, his piece devel
opment is easier as ltlc3 is possible)
16 ltlxc6 'i'e8 17 .i.d5 .i.d6 18
ltle7+ 'i'xe7 19 .ba8 li:lb6 20 .i.c6
j,c7 (theexposedbishop on c6 gives
Black's queen a free tempo on its
wayto h2) 2I ltlc3 'i'd6! (21 ...f5 22
'i'b3+ �h8 23 'i'b4 is annoying for
Black) 22 he4 f5 23 .i.f3 'i'h2+ 24
�fl l:le8 25 g3 f4 and White is cer
tainly not playing for the advantage
- indeed, Black is probably slightly
better.
2b) 15 l:lel dxc4 16 'ilfe2 (after
16 dxe5 ltlc5 the outposts at d3 and
b3 giveBlacktheedge) I6...:es with
equality, as Black's weak pawns are
balanced byhis active pieces. In fact
thishad beenplayed as long ago as
1971, in the game HUbner-Lengyel
from Wijkaan Zee.
14 exdS
14 dxe5? ltlxe4 transposes to line
1 of the previous note.
14 cxdS
15 .i.a2 e4
16 .!Des .i.b7 (DJ
Here I was quite happy, as it
seemed to me that my pieces were
well placed to exen pressure on d5.
While there is some truth in this,
Black has so many pieces available
to defend d5 that the inconvenience
is not serious.
17 lLlcl li:lb6
By protecting d5, Black sets up
various threats based on moving the
f6-knight. The first is to expel the
e5-knight by ...IDeS followed by
...f6, in whichcase he would proba
bly be better. The second is the sim
ple ...li:lfd7, when Black can ex
changeon e5 without having aknight
attacked after dxe5. Hence White
must react quickly.
18 f3! l:lc8
19 .i.b3
37. 36 VISHY ANAND: MY BEST GAMES OF CHESS
White could have maintained
some slight pressure by 19 .ig5, as
Spassky pointed out after the game.
19....ixa3? 20 bxa3 l:l.xc3 is bad af
ter 21 .id2! l:l.c8 (2 1....:Xa3 22 .ib4
favours White) 22 .ia5 ! with a very
awkward pin.
19 --· .ia8
20 .igS l:l.c7
Black's position is basically safe;
for example he could have played
20...exf3 2 1 'i'xf3 �c4 with equal
ity, and indeed this would have been
the natural way to make use of
....ia8, which protected the bishop.
21 .:tel li:Jrd7
22 .if4 .igS (D)
23 .ixgS
After 23 �xd7 the line 23 ....ixf4
24 �xb6 .ixcl 25 �xa8 .ixb2
(25...l:l.xc3 26 bxc3 is also very good
forWhite) 26 �xc7 .ixc3 27 �xd5
.ixel 28 'i'xel exf3 29 gxf3 clearly
favours White. However, Black can
improve by 23...l:l.xd7! 24 .ixg5
1i'xg5 25 fxe4 dxe4 26 �xe4 'lg6
and he regains the pawn since 27
.ic2 fails to 27...f5.
23 --
24 fxe4
'ifxgS
dxe4
Black had a good alternative in
24...�xe5 25 dxe5 dxe4 andnow
I) 26 2d4 �d7! (not 26...e3 27
l:l.e2 attackingb6 and e3) 27 .6 (not
27 �xe4?? .ixe4! and Black wins)
27...�e5 28 exf7+ �h8! 29 'ifd6
l:l.cc8 and Black has dangerous king
side threats.
2) 26 e6. During the game, I be
lieved this was good for White, but
Black simply continues 26...'lc5+
27 �hI fxe6 28 .ixe6+ Wh8 and the
position is just unclear.
The game continuation is also
roughly equal.
25 'ifg4 'ifxg4
Black cannot get away with
25...'ifd2, when White must decide
how to take on f7:
I) 26 .ixf7+ (obvious, but this is
in fact an error) and now:
Ia) 26....:.Xf7 27 �xf7 li:i6 28
'i'e6 'ifxd4+ (28..Jbf7 29 'ifxb6 e3
30 �e4! wins) 29 �hi J:xf7 30
l:l.cdl and White is winning.
l b) 26...�h8! 27 �xd7 (not 27
.:tedI ? ttlxe5 28 dxe5 'ifxb2 and
Black wins) 27 ...'1Vxd4+ 28 �hi
�xd7 with an unclear position.
2) 26 �xf7! �f6 (26...'i'xd4+
27 Wh1 �c5 28 J:cdI 'li'f6 29 J:fl !
wins, whileafter26...'ll'xb2 27 'le6!
Black is in considerable difficulties)
38. ANAND - SPASSKY. CANNES 1989 17
27 �g5+ 'it>h8 (27...�d5 28 'I'd! !
With aclear extta pawn) 28 'iVg3 and
White is clearly better.
26 lllil:g4 g6?
Black's first step downhill. After
26. ..�h8 27 it)e3 f5 Black activates
his kingside majority, which should
provide enough counterplay to main
tain the balance.
r1 �2!
Wllite now wins apawn, although
in view of the reduced material this
does not necessarily guarantee win
ning thegame.
r1 ••.
28 d5
29 it)fxe4
:es
�g7
lllil:dS (D)
At first I couldn't believe this
move; it looks as though White must
win material after 30 .txd5 .txd5
31 liXI6, as Black will end up being
threatened with �xd5 and lt:le8+.
However, Spassky had worked out a
defence. AfterI had calmed down, I
did'nt see any way toforcea decisive
material gain, so I just went for a
pawn.
30 it)d6
30 .i.xd5 .i.xd5 31 �d6 is an
sweredby 3 J...:Xel+ 32 :Xel :cs,
meeting both ofWhite's threats. It is
easy to miss that the c3-knight is no
longer protectedtwiceand therefore
b2-b4 is impossible. White can try
33 :e7 �e6 34 �f7 <M6 35 :Xd7
lhc3! 36 bxc3 .i.xd7 37 it)d6 �e5,
but although he has won a pawn,
Black's king becomes too active and
he draws easily.
30
31 lbel
32 :e7
33 :xt7+
34 �c4
J:x:el+
�f6
:c6
�h6
:e6
Black has managed to get some
counterplay; White's f7-rook is sur
rounded and Black can activate his
king via g5.
35 'M2
36 �2
37 �dl
�,s
.tel»
liS (D)
39. 18 VISHY ANAND: MY BEST GAMES OF CHESS
38 �b3?!
Thiswouldhave been the moment
to settle for 38 l:tg7! �e5 (38...�e8
39 �de4+! Wh6 40 l:th7+! Wxh7 41
�g5+ leaves White aclearpawnup)
39 J:l.a7 and White's rook gets out of
the trap! In fact this is quite awkward
for Black as a6 is weak. Black's
mass of pieces appears menacing,
but White has all the critical squares
controlled. The sacrifice 39...�eg4+
doesn't work here: 40 hxg4 �xg4+
4 1 Wgl (but not 4 1 Wg3? h4+! 42
Wh3 l:te3+ 43 �f3+ .i.xf3 44 gxf3
l:txf3+ and Black wins) 41...l:tel+
42 �fl defends.
I thought that the text-move was
even stronger, but I hadn't really
seen Black's 39th move-in fact, I'm
quite lucky that it doesn't cause seri
ous damage!
38
39 l:ta7
llJes
�eg4+! (D)
This works because White can't
play l2lfl any more.
40 hxg4
41 �
42 wgl
43 l:txa6
44 l:txe6?
�g4+
�
�c2
..td7
Though this sets up the nice fin
ish to follow, it is a mistake. Two
knights can't win by themselves and
there aren't that many pawn left !
White shouldhave played44l:ta5+!,
followed by J:l.d5-d2. By keeping
the rooks on, White has much better
chances of pushing the queenside
pawns and in fact he still has excel
lent winning prospects.
44 ·- ..txe6 (D)
45 llJcS ..tc4
46 a4 Wf4?
One reason why I exchanged
rooks earlier was the combinnation
which now occurs, butin the interim
I had realized that46...'�•f5 prevents
it: 47 a5 l2lb4 48 b3 iin 49 li:ld3
l2lxd3 (a forced move, as 49...li:la6
50 b4 iic4 51 l2lb2 is good for
40. ANAND - SPASSKY, CANNES 1989 39
White) 50 a6 .ie8 5 1 itld5 �e6 52
lOc7+ ci.>d6 53 itlxe8+ �c6 and the
king catches the a-pawn. Therefore
White has to settle for 47 'iii>f2, but
Black has good drawing chances.
47 a5
Now we get to see an elegantfin
ish - a lone knight dominating two
minor pieces !
� ••• itlb4
48 b3 .if7
49 1Dd3+!!
Spassky had only seen 49 a6??
w650itlxa6.ixb3 andWhite has
no winning chances since he has
only one pawn left.
49
50 a6
itlxd3
£e8
51 itldS+ 1-0
After 5 l ...'iii>e5 52 itle7 (D) the
positiondeserves a diagram:
Black can't stop the pawn - a nice
bit of domination!
In Cannes Imadethe bestscore in thejunior team, 61/2110, butthe 'Senior'
Andersson made the best score overall, with 7112 points.
ThenextgameisagainfromWijk aanZee, theyear after mysuccess in the
1989 event.
41. Game 7
M. Kuijf - V. Anand
Wijk aan lee 1 990
Ponziani Opening
1 e4 e5
Round abouta month before Ihad
beaten Kuijf in the tournament at
Groningen. That game had been a
Closed Sicilian, and at one stage it
could have been very dangerous for
me. This time I didn't want the same
'excitement' so I decided to play
more solidly. Hence my choice of
first move.
2 ll:lf3 ll:lc6
3 c3
A real surprise.
3 ••• lllf6
I played this move instantly and
while he was thinking about his re
ply I had to spend afew minutes try
ing to remember the name of the
opening!
4 d4
S dS
I don't really understand what
Kuijfwas aiming forwith his choice
of opening. It is harmless and only
useful ifWhite is aiming foradraw.
6 lOxeS ll:lg6
7 i.d3
However, this indicates that White
is not aiming for a draw, towards
which he could have made substan
tial progress by 7 'iWe2 'fle7 8 'flxe4
(8 ll:lxg6?! hxg6 is a little betterfor
Black) 8...'iWxe5 9 ll:ld2 (or alterna
tively 9 1i'xe5+).
7 ... ll:lxeS
7...ll:lxf2? isn't even a difficult
!rap:White wins by 8�xg6li:lxd1 9
i..xf7+�e7 10 i..g5+�d6 1 1 1iJc4+
�c5 12 i.xd8.
8 i..xe4 i..cS
9 1lfhs d6
10 i..gS? (D)
10 h3 was a much bettermove,aJ.
though even in this case Black can
play for an advantage.
10 i.g4!
Black can play 10...'i'd7 11 0-0
'i'g4, with boring equality, but I had
seen that the text-move is much
42. M. KUIJF - ANAND, WIJK AAN ZEE 1990 41
stronger. Technically, it may be a
aovelty, but I am reluctant to call it
that To my mind, novelties should
be at least alittlebit difficult to find.
If you play the most obvious move
andthen discoverthatby an accident
of history nobody has played it be
fore, I am not sure that it deserves
any special appellation.
11 111'h4?
Losing on the spot. The lines l l
11fxh7 111'xg7 and ll 111'xg4 lLlxg4 12
..ixd8 llixf2 l3 l:l.fl lLlxe4 14 ..ixc7
lc8 15 .ia5 b6 16 ..ib4 ..ixb4 17
cxb4 lc2 also offer White no hope.
This leaves l l .ixd8 ..ixh5 12 ..ig5
(12 ..ixc7 l:l.c8 l3 ..ia5 b6 14 b4
..ixf2+ 15 �xf2 bxa5 and 12 ..ih4
0-0 followed by l3 ...l:l.ae8 are also
verypromising for Black) as the only
realistic way for White to play on.
Even here Black has a range of
tempting options. He could simply
playforthetwobishopsby l2.....ig6
13 llid2 llid3+, but 1Lf6 is proba
bly stronger. Then 14 ..i£4 is impos
sible, 14 ..ih4 leaves thebishop shut
out on the kingside and 14 ..ie3 al
lowsBlack to shatterWhite's pawns.
Black could also consider l2...f6,
wins ith sim Jar id
11 ... f6
Now White has no reasonable
continuation.
12 ..ict
After 12 ..id2 111'e7 l3 0-0 g5
White has the unpleasant choicebe
tween:
1 ) 14 11'h6 00 15 1fg7 1l'xe4 16
l:l.el ..ie2 17 b4 ..ib6 18 ..ie3 <:Je7 19
l:l.xe2 (l9lDd2 1l'd3) l9...l:l.ag8 wins.
2) 14 1l'g3 f5 15 h3 (the only
chance, or else ...f4 traps the queen)
15...f4 16 'llfh2 ..id7 and although
material is even. White is playing a
whole qu.en down for all practical
purposes. After ...0-0-0 and a subse
quent ...g4 the attack should over
whelm White.
If White retreats his bishop to e3
the lines are even simpler: 12 ..ie3
g5 l 3 11fh6 (l31l'g3 f5 14f4..ixe3 15
fxe5 ..if4 and Black wins) l3...11'e7
14 0-0 llif7 15 1l'g7 0-0-0 16 ..ixc5
l:l.dg8 17 l:l.el ..id7 and White loses
his queen.
12 'fle7 (D)
13 0-0
'This is forcedto meetthethreatof
...ll:ld3+, because l3 f3 is met by
l 3.....ixf3.
13 gS
14 'iWgJ rs
43. 42 VISHY ANAND: MY BEST GAMES OF CHESS
The threat is ...f4, so White re- 17 � J:tg8
sorts to desperation. 18 l:!.ael h6
1S .hi'S 19 :XeS+ dxeS
15 h3 f4 16 'irh2 i.d7 shuts the 20 'IfxeS+ �d7
queen completely out of play and 21 i.e3 l:1ae8
Black wins by ...0-0-0 and a pawn 22 'i!Vr4 11Vxd5
push on the kingside. 23 lbr3 i.d6
1S .hi'S 24 'i!Fa4+ bS
16 i.xgS 'i!Ff7 0-1
This game was finished while some of the others were still in the opening;
I spent ten minutes and my opponent a little under half an hour.
It was a nice miniature to play in the first round, but the rest of the tourna
ment didn't go as well as the year before. I lost in the second round to Nunn
and continued unevenly throughout the event, finishing on 50%.
Shortly after Wijk aan Zee, I played an open tournament in Rome. This
started well, but I lost a miniature to Miles in the penultimate round and fin
ished with 6/9. ThenI went back toIndia forsome rest before travelling to the
Zonal tournament in Qatar. I was by far the highest rated player,but even so I
was happy to win the tournament convincingly. As a result of this and some
other tournaments I gained quite a lot of rating points, and on the 1st July
1990 list I stood at 2610-I hadbroken through the 2600banier. I took part in
an open tournament in Manila, went back to India and then returned to Ma
nila for the Interzonal. Based on my new rating I was certainly a potential
qualifier, but in an Interzonal you cannot take anything forgranted - an Inter
zonal is a tough tournament even for the top seeds.
My results in the Interzonal followed a fluctuating course: I won in the
second round, lost in the third, won in the fourth and lost in the fifth. After a
draw with Chandler in the sixth round, I faced the Icelandic grandmaster
Margeir Petursson in round 7.
44. Game S
M. Petursson - V. Anand
Manila Interzonal 1990
Queen's Pawn
1 d4 d6
2 c4 eS
This was part of my usual open
ing repertoire at the time. I liked
l ...d6 because it is such a complete
system in itself. If you play the Pirc
thenyouhave no reason to fear 2 e4,
and 2 lLlf3 can be met by 2.....1lg4.
Since then, however, White has found
ways tokeep some pressure.As are
sult ! lost faith in the system and had
tolearn a decent defence to I d4.
3 lLlc3 exd4
4 'l'xd4 ll:lf6
An important finesse. If Black
plays � • . e:lc6,then 5 'l'e3+ is a little
annoying as 5...il..e7 6 ll:d5 snares
the two bishops.
5 g3 lLlc6
6 'ild2 g6
White's pieces are not badly
placed - for example, the queen is
quite useful on d2 ifWhite intends to
play b3and il..b2.On the other hand,
Blackhasgained atempo because of
'ilxd4 and 'ild2.
7 il..g2
8 lLlh3
9 lLlf4
il..g7
0-0
White is aiming to establish a grip
on d5. He has played the opemng
accurately, forexample by not play
ing b3 too early, which sometimes
allows a tactical ...d5! by Black.
w
9 a5 (D)
This is designed to prevent White
from easily developing his queen's
bishop, for if 10 b3, then IO...a4 I I
llbl (11 lLlxa4 lLle4 wins the ex
change, while II ..llb2 may be met
by l l ...a3 1 2 i.e! i.f5 or l l ...ll:a5)
1l ...axb3 12 axb3 lle8, followed by
...i.f5, with active play.
10 0-0 l:le8?!
A slight inaccuracy. IO...a4 was
better, simply preventing b3.
11 l:lel?!
White misses his chance. 11 e4
was correct, and after l l ...a4 12 llbl
45. 44 VISHY ANAND: MY BEST GAMES OF CHESS
�5 l 3 b3 White has secured his
queensidepawn structure. Ofcourse
the immediate 1 1 l:l:b1 is answered
by 1 l....if5.
11 •.• a4
12 l:.b1 llld7!
A good move. By the time White
has played b3, Black has arranged
to occupy active squares with his
knights.
13 b3 axb3
14 axb3 (D)
14 ...
�eS!
If 4...lt'Kie5!? (intending to play
lS ...llld4) then 15 lllbS! is slightly
better for White. It is the c6-knight
whichbelongson eS, sothat theother
one can occupy cS.
15 'l'c2 lllcS
16 b4?
Premature, because with ener
getic play Black now gets the better
position. After the correct 16 .ib2
c6 we reach a position typical of the
Fianchetto King's Indian. White has
the centre, whilst Black has open
lines forhis rooks and activesquares
for his knights. Chances are about
equal.
16 ... .irS!
An important intermezzo.
17 e4
After 17 lbe4 llla4, followed by
l 8...ltlc6,Black has excellent control
over the long dark-square diagonal.
17 ... lt:le6! (D)
Margeir had missed this tactic.
18 lllxe6
Black's play is tacticallyjustified
by the variation 18 exf5 llld4 (·at
tacking the queen and threatening
19...lbef3+) 19 1fe4 (19 'l'd1 li:lxf3
20 .ixf3 :Xe1+ 21 1Wxe1 li:lxfJ+
andBlack wins) 19...gxf5! (the queen
is trapped) 20 1Wxb7 (20 11ie3 ltlc2)
20...:b8 with a clear advantage to
Blackas White mustnowgive up his
queen.
If 18 .i.e3, then 18 ...i.g4 19 12lxe6
:Xe6 and f3 is very weak.
46. PETURSSON - ANAND, MANILA INTERZONAL 1990 45
18 ••. .be6
Now Black is much better; his
pieces are active and the c4-pawn is
weak.
19 �5
19 c5 dxc5 is very good for Black
after:!J bxcH Vd3 ! or 20 J:[dl 'iVf6!.
19 bS!
Winning the c-pawn and forcing
White to search for some sort of
compensation.
20 .tb2 (D)
Not:!J cxb5? .txd5 21 exd5llJf3+
winning.
20 ... :.2!
I prefer this to the line 20...llJxc4
21 .ixg7 �xg7 22 'iVc3+ f6 23 llJr4
c6! (23....tf7 24 e5 llJxe5 25 .txa8
1fxa8 26:e3 :cs intending ...c5 is
unclear) 24 llJxe6+ :xe6 25 .tfl
llle5. Here Black is a pawn up, but
White has some compensation due
1D the weak queenside and the open
7th rank, which would enhance the
strength of any rook penelration by
White. Note that 26f4can be met by
26...'it'b6+ followed by 27...llJg4.
21 llJcJ :xb2!
22 :xb2 .txc4!
Not 22...llJxc4 23 :a2 with an un
clear position. Mter the text, Black
has extremely active pieces and
strong dark-squared pressure in re
turn for his small sacrifice. The im
mediate threat is 23...llJd3.
23 :eJ (D)
After 23 J:[dl 'iVf6 24 llJd5 llJf3+
25 .txf3 'iVxf3 White is pretty close
to being lost.
23 .th6?
A really awful move whereby, in
one stroke, Black throws away all
hisadvantage.Theproblem was that
I was so excited about the way all
Black's pieces were working well
together that I forgot White could
still develop counterplay. The cor
rect line was 23...llJg4! 24 l:lf3 (24
J:[el .id4 25 'ifcl 'iff6 26 llJdl .i.xb2
wins)24....td425 'ifd2 c5, followed
47. 46 VISHY ANAND: MY BEST GAMES OF CHESS
by 26...lbe5, and Black has a large
advantage.
24 f4!
More or less equalizing.
24 ••• lllg4
25 IU3 �g7
26 'l'dl
26 :tbl !? was also possible.
26 ••• "l'a8
27 h3 "l'a1+ (D)
The tactics don't work for Black
after 21..."tl'a1+ 28 �hl "l'al+ 29
:tbl �xc3 30 li'c2! 'l'a2 31 lhc3
'i&'xc2 32 l:l.xc2 �d3 33 hxg4 .bc2
34 .:tel �xe4 35 l:l.xc7 �xg2+ 36
>fi>xg2 .:te4 and the resulting rook
ending is probably a draw in view of
White's active rook position. I was
quite surprised that Black had noth
ing better in this line; with a King's
Indian bishop and a knight rampag
ing around, you expect something to
work but in this case there was noth
ing.
21 �n �xn
29 lW1 "ifa3
Once again White canhang on af
ter 29..."i'a7+ 30 >fi>hl lbe3 by 31
llcl ! lllc4 32 :ta2'irb7 33 'Wd5!.
30 lU3 lbf6
31 lle3 (D)
31 dS
Black can regain the exchange by
3 l...lbd5?! 32 lllxd5 'l'xb2 33 'l'xb2
�xb2, but after 34 lllxc7 Black is
suffering, for example 34...l:l.b8 35
lld3, 34...llc8 35 lllxb5 llb8 36l:l.b3!
or 34...�d435 lllxe8 .ixe3+36�g2
�d2 37 ll:lxd6 .ixb4 38 lbxb5 and
White is better in every line.
32 li::JxdS
White should avoid 32 e5 d4!.
which gives Black a clear advantage
after 33 'irxd4 'l'xb2 34 exf6 :.Xe3
35 'itxe3 �xf6. However, 32 ll:ld!!
'fila? 33 e5 was a perfectly good a!·
temative to the text; after 33...ll:le4
the position is unclear.
32 .•• "ifa1+
33 l:l.e1 1Wa7+! (D)
48. PETURSSON - ANAND, MANIU INIEKUJNAL 1990 47
I could have forced a draw by defensive moves subconsciously re-
33...li:'lxe4 34 l:lxal ltlxd2 35 l:txd2 !axes just when his troubles appear
J.xal 36 li'Jxc7, but even though to be over, and commits a further er
things had notgone according to plan, ror. In this case time-trouble proba-
1 win as st 11 hop ng to w n. Someti es ly a so pla e
it is amistaketoplayunder theinflu- 34 ••• lLixds
ence of your former advantage, but 35 l:ta2 1rd4!
inthis case Black's optimism proved 36 11'xd4 i..xd4
justified. Petursson had overlooked that
34 �h2?
Petursson returns the favour with
this blunder. The alternatives were:
1) 34 �hi?! ltlxe4 35 l:la2 1Wb7 !
is also bad foe White.
2) 341We3!'l'xe3+ 35ttlxe3lDxe4
36 l:lc2 li:lxg3 37 ttlg2 l:d8 38 .l:l.xc7
�f5 with a near-certain draw. White
has a nominal material advantage,
but Black's pieces are well coordi
nated and White's pawns are dis
jointed, so he has no chance of
putting it to use.
It often happens that a player
who has fought backfrom a bad po
sitionwith a long series of accurate
there was no way to catch one of the
minor pieces, even though they are
temptingly lined up on the d-file. 37
.l:l.dl fails to 37...li'Jc3 and 37 .l:l.d2 to
37....J.c3.
37 .:as ltlc3
38 eS .J.b6
39 .l:l.a3 ttlds
40 .ltb3 (D)
40 ttlxb4!
Simplest. Now if 41 .l:l.xb4, then
4I...i.a5 42 .l:!.ebl .i.xb4 43 :Z:xb4
:Z:b8, and 44...c5, whentheconn��eted
passed pawns will romp home.
41 .l:l.dl cS
42 l:ld7 l:lb8
49. 48
43 g4
44 l:a7
45 l:a6
VISHY ANAND: MY BEST GAMES OF CHESS
46 l:a3
47 �g3
0-1
.i.dl
b4
This game put me on + 1, but although it put me in the right direction, it
was a late winning streak in which I beatLautier, Miles and M. Gurevich in
consecutiverounds that made me a qualifierfor the Candidates.
My next majorevent was the Novi Sad Olympiadtowardsthe endof 1990.
I have always enjoyed playing in Olympiads. Dubai 1986 and Manila 1992
were my favourites; in both cases the organizers went all-out tomake the
players feel comfortable. However, the Novi Sad Olympiad was also pleas
ant, despite the cold and dismal weather. In general I enjoyed playingevents
in the former Yugoslavia because of the great public interest, which always
ensured a good turnout of spectators.
I started the Olympiad with a good win against Olafsson, but lost arather
silly game to Bouaziz in round 3. Later on things started to get better, and I
was satisfied with my final score of71h/12. The followingeventful gamewas
played in round 7.
50. Game 9
V. Anand - I. Morovic Fernandez
Novi Sad Olympiad 1990
Sicilian, Maroczy Bind
1 e4 cS
2 lllf3 lllc6
3 d4 cxd4
4 lllxd4 g6
5 c4 .tg7
6 i.e3 lllr6
7 llJc3 0-0
8 .tel d6
9 0-0 llld7 (D)
I had played the white side of the
Maroczy Bind before, including a
good win against Larsen from the
Cannes event mentioned previously.
However, in that game Larsen played
9....i.d7 10 'itd2 li:lxd4 ll .txd4 .tc6.
Morovic adopts a rather unusual
move, re-deploying his knight to the
queenside.
10 llb1
I was just casting around for a
logical move, and since Black in
tends to play ...lllc5, it seemed rea
sonable toprepare to meet it with b4.
It perhaps looks a little odd toput the
rook on bI rather than c I or (after
'l'd2)dl, but in factWhite often puts
his rooks on bl and c1 in this varia
tion, in order to support a queenside
pawn advance. Typically Wbite con
tinues l:l.ci-bl to play a2-a3. Black
responds with ...Wb6-b4. White then
playsl:l.fcI to support the c3-knight.
Play then revolves around White's
ability to get a3 and b4 in, and
Black's ability to stop it. The idea
behind the text-move is to save time
by going to b1 directly.
10 ••. lDcS
When he played this anyway, I
wondered what the idea was, as he
seemed to be running into b4. Then I
saw that after I I b4 llle6 1 2 lllxe6,
he could play1 2...fxe6! (12....ixe6?!
1 3 llld5 is clearly better forWhite, to
be followed by 'trd2, l:l.d1 and possi
bly c5) 13 lllb5 e5. However, 14 c5
would then give White a slight ad
vantage, so this would have been a
valid alternative to the text-move.
51. 50 VISHY ANAND: MY BEST GAMES OF CHESS
11 'ifd2
A flexible alternative; Whitekeeps
open the possibility of b4 while de
veloping his pieces.
11 ••• liJxd4
Black decides to prevent b4 by
playing ...a5. Ifimmediately l l...a5,
then 12 liJdb5; the plan is f3, fol
lowed by liJd5, and White will play
b3, a3 and b4 at his leisure. Black's
c5-knight can eventually be kicked
away, while the knight on b5 is there
permanently. Hence Black's deci
sion to swap knights before advanc
ing the a-pawn.
12 �xd4 aS
13 b3
White is aiming for an eventual
b4, but the immediate 1 3 a3 allows
1 3...a4.
B
13 �d4
14 'iVxd4 (D)
14 b6?!
There are some lines of the Mar
oczy Bind in which ...b6 and ...�b7
is played, but normally on! where
Black is aiming for central and king
side play by ...liJh5 and ..f5 - it's a
completely different type ofplanto
that Black has adopted here. In this
position ...b6 is an inaccurate move,
which reduces Black's options. His
queen can no longer occupy b6 and
in some lines where Black plays
...e5, it is useful to have ...l:ta6 de
fending the weak d6-pawn.
14...�d7 was probably better,al
though Wbite has the interestingline
15 e5 liJe6 16 '1Fe3 dxe5 17 :tbdl!
and Black still faces difficulties (if
1 7...f6, then 18 �f3).
15 J:lfe1
Now White has a very comfort·
able position. He has a space advan
tage and Black has no chance of
playing ...b5 or ...d5, the two breaks
which normally give Black counter·
play in theMaroczy Bind. Moreover
...a4 is no danger, as Black cannot
back it up by ...'llrb6, so White can
always reply b4.
The only question is how White
arranges to play h4-h5 to step up the
pressure on Black's poorly defended
ldngside.
15 ••. �b7
16 .l:[bd1
The rook has done its duty on bl.
inducing Blackto weaken hi queen·
side. Now the need is forplay in the
centre and in the changed circum
stances White reacts by moving his
rook to the half-open d-file.
52. ANAND - MOROVIC FERNANDEZ, NOV! SAD OLYMPIAD 1990 51
16 ••• f6
17 .i.g4
A nice move, activating White's
bishop. Hedoesn't mindexchanging
bishops, because in the Maroczy
Bind White doesn't need many mi
nor pieces to prosecute his advan
tage. Black cannormally hold a pure
major-piece position, but even one
pair of minor pieces can be enough
forWhite to exploithis space advan
tage.
17 ... .ic6
18 b4 'fkc7
19 'i'eJ (D)
19 h5!? is possible, but I didn't
want to cornrnit my pawns to light
squares too quickly; Black might
stillset upsomesortofdark-squared
blockade by continuing ..,<jJg7, ...g5
and ...h6.
19 ... 'iWb7?!
Black shouldplay 19...<jJg7!, when
after 20 h5 g5. followed by ...h6,
Blackhasset up thesort of blockade
mentioned above. In this case White
would have toprepare h5, forexam
ple by playing llld5 orf4. It's only a
small point, but in such positions
you have to put asmanyobstacles in
your opponent's path as possible.
20 bS
White can push his h-pawn with
out more ado, because ...g5 can al
ways be met by h6, both creating a
permanent danger to Black's king
and making f4 much stronger.
20 ·- M7 (D)
Now it is too late for 20...<jJg7 as
White can play 21 h6+ <jJhS 22 f4,
preventing ...g5.
21 lllds
Preparing to step up the pressure
on g6 by lllf4. 21 hxg6 would be
premature; Black can defend after
2I ...hxg6 22 'ifh6 ltg7.
21 ... gS
22 h6
As intended. Now theg5-pawn is
vulnerable to f4, and Black has to
53. 52 VISHY ANAND: MY BEST GAMES OF CHESS
look after e7, as the rook can always
be driven away by ..th5.
22 ••• �h8
23 a3
Black's position on the kingside is
creaking and now Wbite inconven
iences him further by resuming his
queenside play.
23 e6
24 b4 axb4
25 axb4 exdS
26 exdS?!
A mistake, not because it doesn't
win but because 26 bxc5 ! was much
simpler. If Black takes on c4 or e4,
White obtains a tremendous passed
pawn on d6, while after 26...d4 27
:lb.d4 dxc5 28 .l::!d6, followed by e5,
Black's king will come under a dev
astating attack.
26 ..ta4(D)
27 bxcS?
Seduced by beauty, I allowed my
selfto be distracted from the process
of gaining a point! I had seen the
winning move 27 ll:al !,whereupon
27.../i',d7 28 .:lb.a4 l:l.xa4 29 •e8+
lt:lf8 (29..1lf8 30 1le7 forces mate)
30 'ilfxa4 is hopeless for Black, butI
unwisely decided to 'win' in more
flashy style.
27
28 c6
..txdl
1le7! (D)
Somehow this move had escaped
my attention. White stillhas the ad·
vantagebecause ofhis powerfulpro
tected passed pawn, but of course I
had spoilt my winning position.
29 'li'c3
White has a slight advantageafter
29 .l::!xdl 'li'xe3 30 fxe3 f5, but the
pawns appeared so strong that I
thought I couldplay on with a queen
against two rooks and a bishop.
29 'li'xel+
30 "ifxel hg4
31 'li'b4?
Another error. 31 "tfe4! .i.h5 32
'li'e6 wins the d-pawn (32...l:ld8? 33
c7 wins) without allowing Black to
54. ANAND - MOROV/t FERNANDEZ. NOV/ SAD OLYMPIAD 1990 53
take the c4-pawn in return,which
should be enough foc a clear advan
tage.
31
32 "ilxd6
33 c7?!
.ie2
.ixc4 (D)
33 'i'c6 is a better try:
1) 33.. .:Z.ff8? 34 'il'e7 wins im
mediately.
2) 33...�g8 34 c7 .ixd5 (or
34....ia6 35 d6 .ic8 36 'ii'e8+ l:tf8
37 d7 .ixd7 38 "ilxd7 and wins) 35
c8'ii'+ (35 "ilxd5 .l:te836 "ild8 is also
effective) 35....:Xc8 36 1Wxc8+ .l:tf8
37 'ii'd7 .if7 38 "ile7 wins.
2) 33....l:tafll! 34 "ile4 .ia6! (a dif
ficult move to see; after 34....ixd5
35 "ilxd5 Black can't immediately
double rooks against the pawn, for
example 35....l:tc7 allows 36 "ild6)
35 d6 .ic8 36 'ii'd5 �g8 37 d7 .ixd7
38 cxd7 .l:td8 39 "ile6 Wf8 and Black
draws.
33 •••
.l:te8??
A time-trouble blunder. Black
could force a draw by 33....l:taf8! 34
"ilc6 .ixd5 35 "ilxd5 .l:txc7 36 "ild6
.l:tcf7 37 11xb6.
34 We6!
Grabbing my chance.
34 ... .l:tfl'8
35 c8'1f!
A pretty win.
35 .:XeS
36 'ii'e7 1·0
InApril 1991 I played in a tournament held in Munich. While I very much
liked thecity, I have less happy memories of the chess. In fact, my greatest
pleasure was the blitz tournament held at the end, which I won with 14/15,
21h points ahead ofthe next player.
Inround 1 I lost to Nunn, and in round 2 I was fortunate todefeat Zsuzsa
Polgar. In round 3 I met Beliavsky, and the result was one of the few good
games I played at Munich.
55. Game 10
A. Beliavsky - V. Anand
Munich 1991
Pirc Defence
In our previous encounter (Linares
1991) I had built upa totally won po
sition only to perpetrate a form of
hara-kiri. I was very pleased to get
revenge in this game!
1 d4 d6
2 e4 liJf6
3 liJc3 g6
4 (4 ..tg7
5 liJf3 0-0
6 ..te3 b6 (D)
When I was studying this line
from White's point of view, I won
dered why 6...b6 wasn't a more
popular reply. Black forces through
...c5, and ifWhite is to try for an ad
vantage he has to push all his pawns
forward in the centre, which is very
committal. If the pawns eventually
tum out to be weak, the weakness
wiJI probably be serious.
7 e5 liJg4
8 .i.gl cS
9 h3
White avoids a little trap: 9 dxc5
bxc5 10 'ill'd5 'ill'b6 I I ifxa8 losesto
I I.....tb7 1 2 liJd5 ifxb2.
9 liJh6
10 d5 i.b7 (D)
I knew the theory, but at this point
I decided to ignore it andjust look at
the position. It seemed to me that
Black could play very natural moves.
The point of this one is to play ...e6
andcompletelydestroy White's cen·
tre. After the resulting exchanges
56. BEUAVSKY - ANAND, MUNICH 1991 55
Black may be left with a weak pawn
(for example, on e6) but it doesn't
matter because Black has generated
so much active play for his pieces.
While the specific move IO.....i.b7
was thoughtupoverthe board, I had
looked at these lines before and the
ideas I had during this earlier analy
sis germinated into this 'innov
ation'.
11 "i'd2
After 1 1 "i'e2 ..i.a6 (I I...a6 and
I2...b5 is also possible) I2 'l'f2
i.xfl 13 "i'xfl lLlf5 Black equalizes
comfortably. If White tries to shut
the h6-knight out by I I g4, then
Black plays I l...dxe5 I2 fxe5 e6 I3
i.c4 exd5! (13...b5 I4 ..i.xb5 exd5
IS i.xc5 favours White) I4 ..i.xd5
with an unclearposition.The knight
on h6 isbad, but thee5-pawn is weak
and gi-bishop is also oddly placed.
11 ...
lLlfS
12 ..i.h2
After I2 ..i.f2 dxe5 I3 fxeS e6
White doesn't have time to castle
long owing to I4 0-0-0 .i.h6. Hence
the text-move.
12 dxeS
13 fxeS e6!
14 0-0-0
Other moves are ineffective, e.g.
I4 g4 i0114 is very bad for White, I4
d6 lLld7 leaves e5 collapsing and fi
nally 14..i.c4 .ih6 (14...exd5 15
i.xds lLlc6and I4...b5 I5 lLlxb5 exd5
16 i.d3 are also possible, in both
cases with an unclear position) IS
..i.f4 ..i.xf4 I6 1Wxf4 exd5 I7 0-0-0
d4 IS lLle4 lDd7 leaves White with
out enough for the pawn.
14 exdS
1S lLixdS lLc6
16 c3 (D)
Underestimating Black's initia
tive. If White had time for ..i.c4 and
l:hei then he would have a clearad
vantage, but Black's counterplay is
so fast that he has no time for the
necessary consolidation, e.g. I6 ..i.c4
lLlcd4 I7 lLlxd4 cxd4 attacking the
d5-knight and, when it moves, creat
ing the possibility of ...lLle3.
16 ••• lLlcd4
17 lLr6+
I7 cxd4 11fxd5 IS �bi :adS is
verygood for Black.
17 ..i.xf6
18 cxd4
IS exf6 .'Llxf3 I9 gxf3 11fxf6 gives
Black a safe extra pawn.
18 ..i.g7
19 dS (D)
57. 56 VISHY ANAND: MY BEST GAMES OF CHESS
19 c4!
Essential. IfWhitecouldplay.i.c4,
then Black would be in some trou
ble. Now Black locks in the bishop
on f1 and frees the c5-square for the
manoeuvre ...l:lc8-c5. All Black's
moves are very natural; he has the
initiative and must make use ofit be
fore White completes his develop
ment, so heneed only consider active
moves.
20 .tel?
In the critical position White goes
wrong. Heshouldplay 20 �bl , with
the possible continuation 20...l:lc8
21 d6 (21 .i.e2 .i.h6 22 .i.f4 .i.xf423
'l'xf4 .i.xd5 transposes to the game)
and now Black can choose between
21...1lrd7 and 21...b5. Black has a lot
of trumps, not least of which is his
lead in development. Theposition is
unclear but I prefer Black.
20 ... l:lc8
21 �b1
21 g4 Ci:Je7 22 d6 00, threaten
ing 23...c3, favours Black.
21 ...
.1h6
22 .i.f4
After 22 11rc3 Black can also
safely take the d5-pawn.
22 Jud4
23 11rxf4 .i.xdS! (D)
-
w •
24 h4?!
There is no way Whitecanexploit
the d-file pin, so he tries fora king
side attack. The alternatives were:
1) 24 Ci:Jd4 c3 (24...Ci:Jg7 25 ttlb51
is bad, but 24...'llrb4 25 .i.g4 1Llxd4
26 llxd4 llcd8 is a reasonable alter
native for Black) 25 bxc3 'i'h4 26
.i.g4 Ci:Jxd4 27 l:lxd4 .i.xg2 28 :gl
.i.xh3 29 .i.xc8 11rxf4 30 J:lxf4AxeS
is excellent for Black. He has two
connected passed pawns forthe ex
changeand White'sremainingpawns
are weak.
2) 24 lld2 c3 (alternatively, after
24...llc5 25 llhdl 11ra8 Black keeps
his extra pawn) 25 bxc3 J:lxcJ 26
llhdl Ci:Je3 and Black is clearly bet
ter.
58. BEUAVSKY - ANAND, MUNICH 1991 57
3) 24g4 tiJg7 25 'l'h6 tiJe6 26 h4
c3 27 bxc3 (27 �g5 c2+ 28 �a!
cxdl'l'+ 29 .l%xdl �xg5 and Black
wins) 27...i.e4+ 28 �b2 'ilc7 fa
vours Black. White's pawns are weak
and his king exposed.
24 ••• c3
25 bxc3 .l%xc3
26 bS lLle3!
Black's attack is much faster than
anything White can muster.
21 lt)gs
After 27 hxg6 fxg6 28 'l'h6 "flc7
Black defends h7 and White will be
mated, while after 27 "flh6 ll!xdI 28
hxg6 (28 ll!g5 .i.xa2+ forces mate)
28....i.e4+ 29 �a! .i.xg6Black wins
easily.
27 •••
28 ll!m7
0-1
flc7
l:lb3+!
It is mate after 29 axb3 flc2+ 30
�a! 'ilc3+3 1 �bl "flxb3+ 32 �al
lllc2#.
A few days afterqualifying from the Manila Interzonal, I received an invi
tation to play in Linares (1991) from Seiior Rentero. On my way to Linares
(which was a couple of months before the Munich event mentioned above) I
stopped offfor a couple of days inAmsterdam.There I receivedthe news that
FIDE had made the pairings for the quarter-finals. Instead of the simple 1 vs
8, 2 vs 7, etc., they had changed the system such that anyone in the top half
could be paired against anyone from the bottom half .
Karpov, Timman, Yusupov and Short were in the top half of the draw, with
Gelfand, Ivanchuk, Korchnoi and myself in the bottom. I was paired against
Karpov, who at that time was a formidable opponent. Just at that moment I
was quite annoyed by this pairing, but later I took the view that you couldn't
become World Champion by avoiding people - you just have to take oppo
nents as they come.
I washeartened by the fact that I beat Karpov quite easily in Linares after
he misplayed a promising position. Between then and the match in August
neither of us had produced any inspiring results. My own performance in
Munich was not very satisfactory, while Karpov had drawn a match 2-2 with
Agdestein.
Inthe firstgame of the match itself I played an insipid system; to be honest
we (my second in this match was M. Gurevich)knew that itgave White noth
ing against best play, but we decided to try it anyway. because Karpov had
failed to find the correct solution in a previous game. He got a bad position
with an isolated pawn and suffered a lot, but defended very well and, indeed,
outplayed me completely. However, he threw away all his good efforts by
misplaying the ending.
59. 58 VISHY ANAND: MY BEST GAMES OF CHESS
In the second game I outplayed Karpov completely, but then went wrong
and had to acquiesce to a draw. In the third game I could have mated him in a
few moves, but somehow just didn't see it. Then I lost the founhgame. In
both game two and game four I had played the Meran Defence, which I pre
pared especially for this match; I felt it was a dynamic opening and that Kar
pov wasn't particularly good against it. In games two and four I had played
after I d4 d5 2 c4 c6 3 tbf3 tbf6 4 tbc3 e6 5 e3 tbbd7 6 .ie2 .id6 7 0-0 0-0 8
'i!Vc2 dxc4 9 .ixc4, a plan involving ...a6. In game four Karpov had found a
pretty good line against this system, so in game six I decided to switch to
9...'i!Ve7, which also formedpart of my preparation. I should addthat game
five was unfinished when the following game was played, but I waswinning
the adjourned position.
60. Game 11
A. Karpov - V. Anand
Candidates match (6), Brussels 1991
Semi-Slav
1 d4 dS
2 c4 c:6
3 lllf3 �f6
4 0.c3 e6
5 eJ �bd7
6 'i'c2 i.d6
7 i.e2 0-0
8 0-0 dxc4
9 hc4 'ii'e7
10 h3
To avoid the exchange of knights
after ...�e5, but it is not a very ambi
tious move.
13 ••• iDeS
14 .i.b3 i.d7
15 i.eJ (D)
Ingame eight he finally found the
rightrecipe, which is to play 10 a3.
B
10 ••. cS
Ingame fourI hadplayed 10...a6.
11 dxcS i.xcS
12 e4 (D)
12 • • •
13 �d4
i.d6
15 ••• �g6
15...l:tfd8? is a loss oftime; White
continues 16 f4 �c6 17 e5 �xd4 1 8
i.xd4 i.c5 19 l:tad1 (after 1 9 i.xc5
'ihcS+ 20'ii"f2'ihf2+ 21 :x£2 iDeS
22 l:td1 i.c6 23 l:tfd2l:txd2 24 l:txd2
�f8 Blackshouldholdon)and now:
1) 19...i.c6?20exf6l:!.xd4(White
also wins after 20...i.xd4+ 21 l:txd4
1Wc5 22 lbe2!) 21 fxe7 l:txdl+ 22
�h2 l:txfl 23 'i'd3 i.gl+ 24 �h1
and White is winning.
61. 60 VISHY ANAND: MY BEST GAMES OF CHESS
2) 19....ixd4+ 20 :Xd4 ll:le8 21
l:l.fd1 and White is clearly better.
After the text-move Black has
more or less equalized.
16 J:l.ad1 J:l.fd8
17 ll:lr3 .ic6
18 J:l.fe1 l:tdc8
This looks a little odd, but I felt
it was important to inconvenience
White's queen by playing a rook to
c8. The aS-rook can't go there be
cause the a7-pawn is hanging, so it
has to be this one. White must move
his queen because 19....ixe4 is a
threat.
19 'l'bl (D)
. 0,
-B -
19 NO .ib4
Everything is based on threats
against the e4-pawn.
20 .id2
After 20 .id4 e5 the e-pawn
would be in serious trouble. In this
position only White can be worse,
because of his exposed e-pawn.
20 l:td8
The rook returns now that White
has been forced to block the d-file.
21 a3 .icS
22 ll:la4 .id6 (D)
Black is more or less committed
to this piece sacrifice, as 22....ixa4
23 .ixa4 would give White a slight
advantage based on his two bishops.
However, I had no objections as I
feltthat it was promising for Black.
23 ll:lc3?
After this cop-out, Black's pieces
are more harmoniously placed. The
criticalline is 23 e5! .ixf3 and now:
1) 24exf6 (bad) 24...'1'xf625 gxf3
.if4! and now:
Ia) 26 .ic3 "irg5+ 27 �h1 (27
�fl "irb5+) 27...1Wh5 28 �g2 lllh4+
and wins.
1b) 26 .ixf4 ll:lxf4 also wins.
1c) 26 .ib4 'ii'g5+ 27�fl 'li'b5+·
followed by 28...a5, with a distinct
plus for Black.
2) 24 exd6 (best) 24...'1'xd6 25
gxf3 ll:lh4 and now:
62. KARPOV - ANAND, BRUSSELS CANDIDKTES 1991 61
2a) 26 .tc3?! lllxf3+ 27 �g2
®14+ 28 1i>fl (28 'iii>g1 'ifc6 is dead
lost) 28_.'i'a6+ 29 l'Z.e2 'il'c6 and
Blackshould win.
2b) 26 .i.e3? lilxf3+ wins.
2c) 26 l'Z.e3 (D) and now:
2c l) 26...1i'f4?! 27 l'Z.d3! lilxf3+
28 �g2 lilh4+ (28...lilxd2 29 l'Z.1xd2
'i'gS+ 30 �fl 1i'xd2 3 1 l'Z.xd2 l'Z.xd2
gives White an edge) 29 �fl 'il'h2
30 .tg5! (30 .i.c3 lild5 31 .i.xd5
l'Z.xd5 32 lilc5 is unclear) 30...l'Z.xd3
31 'i'xd3 'il'g2+ 32 �e2 lilg6 33
l'g3 favours White.
2c2) 26...1i'c6 27 .i.c3 .!l'lxf3+ 28
�fl b5 with a final branch:
2c21) 29 l'Z.xd8+ l'Z.xd8 30 .i.d1
IOd4 (30...lild2+ 31 .i.xd2 l'Z.xd2 32
.tn defends, while 30...lilh4 31
.txf6 gxf6 32 l'Z.g3+ �h8 33 .!l'lc3
'i'hI+ 34 �e2 lilf5 35 l'Z.d3 lild4+ 36
�d2 is at least equal for White) 31
i.xd4 l'Z.xd4 32 lilc3 1i'h1 + 33 �e2
lllh5 34 J:f3 1i'h2 35 .i.c2 lilf4+ 36
�3 lilg2+ 37 Wxd4 1i'd6+ 38 �e4
'il'c6+ 39 �e5 and this exciting line
ends in a draw by perpetual check.
2c22) 29 lilc5 lilh2+ 30 �e2
'il'xc5 31 l'Z.xd8+ l'Z.xd8 32 .i.xf6
gxf6 33 1i'g1+ �h8 34 'il'xh2 'il'cl is
unclear.
The conclusion is that White may
be able to hold the balance by ac
cepting the sacrifice, but it would
have been very difficult for Karpov
to find all this at the board!
23 ••• 'il'c1 (D)
If we look at the total effect ofthe
last five moves, White has played
.i.d2, 'ifbl and a3, while Black has
achieved ...1i'c7. Thus White's queen
and bishop have been pushed back,
while Black has improved his queen
position and now controls e5 -
Black's manoeuvrecan be counted a
success.
Now Black is slightly better be
cause he can expand on the queen
side, while in the meantime White's
pieces are only crawling back to
their earlier positions.
63. 62 VJSHY ANAND: MY BEST GAMES OF CHESS
24 �a2
It's hard to say what this move is
for. Perhaps he was worried that af
ter ...j_cs, there might be a threat of
...j_xf2+ and ...'l'b6+, butit's only a
guess.
24 •.• a6
25 j_eJ b5
White's 24th move is revealed as
a mistake. It may have solved one
problem, but now the bishop has to
return in order to meet ...'l'b7 by
j_c2.
26 j_b3 j_b7
27 .:tel 'l'e7 (D)
I was very proud of this game, be
cause I felt that I had outplayed Kar
pov in themanoeuvring phase of the
game. The idea now is ...�7 fol
lowed by either ...lllcS or ...lLldeS
heading for c4.
28 .1b6 l:.dc8
29 .1d4
Now 29 e5 .ixf3 is good for Black,
as 30 exf6 1i'xf6 leads to variations
similar to those in the note to
White's 23rd move, while 30 exd6
'l'xd6 attacks the bishop on b6 so
there is no win of a piece.
29 ·- �7
30 l:.cdl
Over the last few moves White
has only been moving his pieces
backwards and forwards, simply re
sponding to Black's various threats,
while Black has gained space on the
queenside and created an outpost at
c4. Nevertheless, the symmetrical
nature of the position exerts adraw
ish tendency. Black's position is
more comfortable and easiertoplay,
but one cannot say more than that
30 liJgeS
31 liJxeS lillce5 (D)
32 liJe2?
An error, overlooking the reply.
32 ••• 1i'h4
33 f4
After 33 lt:lc3 Black continues
33...lllf3+ 34 gxf3 1i'xh3 35 l:.d3
64. KARPOV - ANAND, BRUSSEU CANDIDKI'ES 1991 63
.A.h2+ 36 �hl .i.£4+ 37 'iilgl :tc6,
and the possibility of playing ...e5
and switching the rook to the king
side givesBlackaneasy win. The al
ternative 33 lLlg3 is similar; then
33...<�f3+ 34 gxf3 .bg3 35 'iilg2
.A.f4 doesn't lead to a winning at
tack, but White's pawn structure is
significantly damaged.
The text-move is therefore forced,
but it is a move White certainly
doesn't want to play with Black's
bishops pointing at the kingside.
33 •••
lLle4
34 .if2 .icS
35 .!xeS :XeS (D)
36 .l:l.el?
36.l:l.d4was abetterdefence, when
36...ll:lxb2? loses to 37 :tfl lLlc4 38
.ixc4. However, Black can continue
36....1:1.ac8 with a clear advantage.
36 •.• li)d2
37 Wd3 li:Jxe4?!
One of my weaknesses during
this match was myinefficiency in
converting technically winning po
sitions. Too often I played moves
which just maintained my advan
tage, instead of pressing it home
forcefully. On the other hand, Kar
pov, although he only gained a large
advantage twice, pushed it home
both times, and this effectively de
cided the match.
Here 37....l:td8! 38 'ire3 lLlxb3 39
'irxb3 .ixe4 would be a much im
proved version of what happens in
the game.
38 'ife3
39 .l:l.edl
.l:l.ac8
h6
39...g6 would have been slightly
more precise.
40 :n
41 lLlc3
lLlf6
'ii'hS (D)
Black has a clear extra pawn, and
despite White's stern resistance, the
win shouldonly be amatteroftime.
42 :1d6 .I:I.Se6
43 .l:l.xc6 .ixe6
44 :1f2 li)d7
65. 64 VISHY ANAND: MY BEST GAMES OF CHESS
45 �d2
I decided not to try to find a mid
dlegame win, but just to liquidate
down to an ending.
46 'i'xe5
47 .tel
48 <M2
49 �e3
50 g3
51 ..tdl
ltJxe5
Wf8
�e7
liJd7
�6
ltJe8 (D)
At some point I should play ...g5,
so that if White plays h4, then Black
can reply ...gxh4 followed by ...J:tg8,
when White has the possibility of a
g-file penetration to worry about.
52 .tel
53 .i.d3
ltJd6
.i.b7
53...g5 would still be good, but
this is the last chance as White now
prevents it (of course, he could have
played h4 earlier himself).
54 h4
Now any attempt to prepare ...g5
will just lead to a lot of simplifica
tion.
54 "" ltJc4+?!
A real mistake. Again it doesn't
give away Black's advantage but it
makes the win fairly complicated
1n endings with bishop and knight
each, exchanging dissimilar pieces
(i.e. a bishop for a knight) has the
effect of increasing the defender's
drawing chances. The superior side
should therefore only exchangesimi
lar pieces.
55 he4 lbc4
Now White has much betterpros
pects of setting up a dark-squared
blockade.
56 :td4 :tcS (D)
57 a4
At this point I decided to sit down
and really calculate, because I could
see that the win was no longer going
to be trivial.
57 ••• eS
Over the next few moves Kar
pov's resistance starts to weaken.
58 l:lb4
66. KARPOV - ANAND, BRUSSEI.S CANDIDATES 1991 65
Already 58 fxe5 l::txe5+ 59 'it>f2
was more accurate, when Black has
a long way to go to create a passed
pawnon the kingside.
58 .i.c6
59 axbS axbS
60 �2? (D)
This was White's last chance to
play 60 fxe5 l::txe5+ 61 'it>f2.
B
60 f6!
Now Black retains control of the
key dark squares d4 and f4, and the
Another bad move. White should
have tried 6 l ll'lc3.
61 ••• 'it>e6
62 l:la3 l:lc2
Now Black is winning.
63 fxe5 fxe5
64 l:la6
The only line I needed to calcu
late was 64 l:lc3 l:lxc3+ 65 ll'lxc3
'it>d6 66 b4'it>e667�d3'it>£5 68 �e3
�g469 �f2 e4 70ll'ldI 'it>£5 71 �e3
�e5 72 ll'lc3 g6 (but not 72...g5? 73
hxg5 hxg5 74 g4 and White escapes
with a draw) 73 g4 g5 74 h5 .i.d7
and Black wins.
64 ••• �d6
65 b4
65 l:la7 l:lxb2 66 l::txg7 b4 is also
an easy win.
65
i:ia7
:C4
66 .i.d7
67 l:la6+ �7
68 l:lg6 �7
69 l:ld6 .i.g4
70 l:ldS l:lc2
white knight will feel a lack of good 0-1
squares. As the knight is trapped after 71
61 :b3? ll'lgl l:lg2.
This game was finished after theconclusionofthefifthgame; the two ad
journament were played on the same day.Unfortunately , in the adjourment
ofthe fifth game Ifirst made a simple win rathercomplicated and then, when
I need to calculate a long forced line in order to win , I made a mistake that
allowed him to gain a tempo.Thegameended in adraw. Here Karpov's expe
rience alsoplayed a part. He had two lost positions (games five and six), but
he concentrated all his efforts on finding resources in game five (where
had a better chance to save the game) and eventually managed to salvage a
half-point
67. 66 VJSHY ANAND: MY BEST GAMES OF CHESS
I missed a forced win in the seventh game and then he won the eighth game
totake the match. Cenainly he defended very well in thismatch,buthedidn't
demonstrate the form which would have taken him to the world champion
ship, and he subsequently lost to Short in the semi-finals.
The two matches I played in this Candidates cycle were virtually the first
matches I had ever played. Before, I had only played a not especially serious
four-game match with Levitt. When the next Candidates cycle came around,
the experience of match play that I gained against Dreev and Karpov turned
out to be very useful.
The 1991 Tilburg tournament was adouble-roundevent. Priorto this tour
nament I had only played Kasparov once, the game ending in a draw. In the
first cycle at Tilburg I lost, but gained my revenge in the game immediately
following. I won again in Reggio Emilia (see Game 1 3) but, to date, that was
the last I saw of a plus score against him!
