Contents

Bibliography 7
Preface 9

1 Pawn Endings – Introduction 13

1: Cohn-Rubinstein, St Petersburg 1909 14
2: Brinckmann-Rubinstein, Budapest 1929 18
3: Randviir-Keres, Parnu 1947 21

2 Knight Endings – Introduction 24

4: Znosko Borovsky-Seitz, Nice 1931 25
5: Pillsbury-Gunsberg, Hastings 1895 29
6: Lasker-Nimzowitsch, Zürich 1934 31
7: Kramnik-Shirov, Tal Memorial, Moscow 2007 35

3 Bishop Endings – Introduction 38

8: Yanofsky-Pinkus, Ventnor City 1942 40
9: Averbakh-Furman, USSR Championship, Odessa 1960 43
10: Larsen-Polugaevsky, Le Havre 1966 46
11: Kotov-Botvinnik, USSR Championship, Moscow 1955 52
12: Kurajica-Karpov, Skopje 1976 56
4 Bishop versus Knight Endings – Introduction 60
13: Reti-Rubinstein, Gothenburg 1920 62
14: Plater-Botvinnik, Chigorin Memorial, Moscow 1947 65
15: Fischer-Taimanov, Candidates Match, Vancouver 1971 68
16: Torre-Jakobsen, Amsterdam 1973 74
17: Henneberger-Nimzowitsch, Swiss Ch’ship, Winterthur 1931 79
18: Smyslov-Tal, USSR Championship, Moscow 1969 82
19: Flohr-Capablanca, Moscow 1935 86
20: Karpov-Kasparov, World Championship, Moscow 1984 90
21: Botvinnik-Smyslov, World Championship, Moscow 1957 97
22: Flohr-Botvinnik, Moscow/Leningrad 1933 101
23: Botvinnik-Bronstein, World Championship, Moscow 1951 107

5 Rook Endings – Introduction 111

24: Alekhine-Capablanca, World Ch’ship, Buenos Aires 1927 113
25: Alatortsev-Chekhover, USSR Championship, Tbilisi 1937 118
26: Schlechter-Rubinstein, San Sebastian 1912 124
27: Spielmann-Rubinstein, St Petersburg 1909 126
28: Capablanca-Tartakower, New York 1924 131
29: Flohr-Vidmar, Nottingham 1936 137
30: Andersson-Miles, Tilburg 1981 141
31: Botvinnik-Fischer, Varna Olympiad 1962 145
32: Korchnoi-Karpov, World Championship, Baguio City 1978 151

6 Queen Endings – Introduction 156

33: Botvinnik-Minev, Amsterdam Olympiad 1954 158
34: Marshall-Maroczy, Ostend 1905 163
35: Pillsbury-Burn, Vienna 1898 166
36: Webb-Miles, British Championship, Morecambe 1975 169
37: Schlechter-Lasker, World Ch’ship, Germany/Austria 1910 173
38: Tal-Keres, Candidates T’ment, Bled/Zagreb/Belgrade 1959 180
7 Rook and Minor Piece Endings – Introduction 184
39: Timman-Velimirovic, Interzonal, Rio de Janeiro 1979 185
40: Alekhine-Lasker, St Petersburg 1914 190
41: Marshall-Lasker, World Championship, USA 1907 193
42: Andersson-Stean, Sao Paulo 1979 197

8 Multi-piece Endings – Introduction 200

43: Englisch-Steinitz, London 1883 202
44: Barcza-Soos, Varna Olympiad 1962 207
45: Petrosian-Botvinnik, World Championship, Moscow 1963 211
46: Fischer-Petrosian, Candidates Match, Buenos Aires 1971 216
47: Petrosian-Bannik, USSR Championship, Riga 1958 219
48: Carlsen-Azmaiparashvili, FIDE World Cup,
Khanty Mansiysk 2005 222
49: Karpov-Kasparov, World Championship, Moscow 1984 226
50: Kramnik-Leko, World Championship, Brissago 2004 230

Index of Games 233

 Preface

I was delighted when Everyman asked me to write a book on the greatest ever
endgames, because this phase of the game has always been my favourite aspect
of chess. Although we all enjoy seeing spectacular sacrificial attacks, it is in the
endgame that the true depth and subtlety of chess is revealed. The complexities
and fine points that can be hidden in positions with only a small number of pieces
on the board never cease to astonish. The older masters, such as Capablanca, were
absolutely right to recommend that beginners start their study of chess with the
endgame, because it is by studying simplified positions, with only a few pieces on
the board, that one can develop an understanding of the capabilities of the pieces,
and what they do and do not “like”.
In this context, I recently came across an interesting story, which shows that
chess is not the only game where such a study of the basics pays off. Walter Lin-
drum, the Australian star, who in the 1930s and 40s developed the game of bil-
liards to such a level that he practically killed it as a competitive event, was tu-
tored in his youth by his martinet father, himself an excellent player. Young Wal-
ter was forced to spend up to twelve hours each day, practicing. It seems that for
the first three months, his father only allowed him one ball, the cue ball! As one
can imagine, by the time he had spent that long, with just a cue ball, Walter could
do just about anything imaginable with it, short of making it sing and dance, and
this expertise was the foundation of his later immortality at the game. In chess,
Vasily Smyslov is an excellent example of a player who learnt his chess expertise
by starting with the endgame.
Obviously, there are many textbooks on the endgame which deal with the basic
theoretical positions. The present book is somewhat different. As with other books
in this series, The Greatest Ever Chess Endgames aims to entertain, and also to in-
struct the reader, by presenting some of the finest practical examples of endgame
play in master chess. Through the medium of 50 annotated examples, we will look
at the different types of endgame, and see how they have been played by the
world’s greatest masters.

9
The Greatest Ever Chess Endgames

I should say a word about the selection. Clearly, with a book such as this, a lot of
the examples presented will be familiar to many readers, and will be games which
have been published in other endgame books in the past. However, I have tried, as
far as possible, to pick at least a few slightly less well-known examples, and there are
bound to be some games missing from the book which readers might expect to see
here. To take just two examples, Rubinstein’s famous rook endings against Lasker
(St Petersburg 1909) and Alekhine (Karlsbad 1911) have both been omitted, because
I felt them to be so well-known that it would be impossible to say anything original
about them. Where I have chosen other famous examples, I have tried, as far as pos-
sible, to add something new, or at least to pull together recent analytical discoveries
published in more obscure sources. One example is Capablanca-Tartakower, New
York 1924 (Game 28), where I have included a summary of some fascinating analysis
published in the Russian journal 64 a few years ago. Where appropriate, I have also
made use of the seven-man endgame tablebases, which have only recently become
available. These have enabled me bring a degree of certainty to some positions that
was previously unachievable.
One other thing I should comment on, as far as the selection is concerned, is
the age of the games presented here. The reader will note that, with a few excep-
tions, most of the games were played no later than the 1980s, and many long be-
fore that. To some extent, this reflects my own interest in chess history, and rever-
ence for the great masters of the past. However, I think there is also another, more
objective reason, which is the change to tournament regimes and time-limits over
recent years. A great many of the games in this book were played in the days of
slow time-limits (40 moves in two and a half hours being the norm), and with ad-
journments. Whilst I would accept that adjournments, and the use of third parties
to assist with analysis of adjourned games, does contradict the basic philosophy of
chess, as being a battle between two individuals over the board, I think it is clear
that abolishing adjournments has had a hugely negative impact on endgame
play. One only has to look at the epic analytical achievement of Jan Timman and
his second Ulf Andersson in Game 39, to see what we have lost by getting rid of
adjournments. In recent years, this process has been taken even further, with the
adoption of Fischer-style increment time limits, which almost all authorities agree
have further damaged the standard of endgame play. The paucity of games from
the past 20 years to be found in this book reflects this sad fact.
At the beginning of each chapter, I have added some introductory remarks,
which summarize the main principles and characteristics of the different types of
endgame. For now, there are three general endgame principles, applicable to all
endings, which are worth mentioning here:

10
 Preface

1. The principle of two weaknesses

In almost all of the endgames you will see in this book, the key to winning the po-
sition is for the stronger side to create a second front, so as to stretch the de-
fender’s resources beyond their limits. This may involve creating another passed
pawn, opening a file, or simply weakening a single square, so as to allow penetra-
tion (or, quite often, merely the threat thereof) by a piece. Whatever the precise
means, a second front is nearly always required. A single weakness is rarely
enough to lose by itself.

2. Do not hurry!
Time and time again, in this book, you will see players delaying decisive action, in
favour of taking minor prophylactic measures, playing small strengthening
moves, marginally improving the position of a piece, etc. Such play is absolutely
typical of the endgame. The majority of the time (rook endings are a frequent ex-
ception), the pace of the endgame is slower than the middlegame, and there is not
the same pressure to take immediate action, or to avoid any suggestion of a
wasted tempo. Instead, patient manoeuvring and gradual strengthening of the
position tend to be the order of the day.

3. Endgames always become concrete!

This rule, which to some extent may seem to contradict point 2 above, was taught
to me years ago by GM Jonathan Levitt. Despite the slower pace of endgames, and
the need for patient manoeuvring, nearly all endgames eventually reach a phase
where they become concrete, and exact calculation of specific variations is re-
quired. Many players tend to assume that tactics and calculation belong in the
middlegame, and have little place in the ending, but this is a major fallacy. If you
look through the endings in this book, especially the minor piece endings, you will
see that a long bout of patient manoeuvring usually culminates in a specific, con-
crete tactical finish, which needs to be accurately calculated. This may be as sim-
ple as counting a passed pawn race in a king and pawn ending, or it may involve
much more complex calculations, such as in Henneberger-Nimzowitsch (Game
17), but if you cannot or will not calculate variations, you will not be a successful
endgame player.

This has been an immensely enjoyable book to write. Spending my days going
through some of the finest practical endings ever played has been a great way to
pass the last few months, and I hope the reader will gain a similar enjoyment from
reading the book. If it serves to stimulate an interest in the endgame amongst

11
The Greatest Ever Chess Endgames

players who hitherto thought it dull, I will be even more delighted.

Last, but by no means least, acknowledgments are due to the following: John
and Byron of Everyman, for their advice, technical support and general encour-
agement; Calle Erlandsen, Per Skjoldager and Jørn Erik Nielsen, who between them
dug out and supplied details of Larsen’s Danish commentary on Game 10; Gerard
Welling, for supplying photocopies of Dutch-language material pertaining to
Game 16; Bernard Cafferty, who clarified the probable repetition rules applying in
Game 35; Matthew Sadler, for some interesting analytical suggestions in Game
25; Jakov Konoval, who provided results from his seven-man endgame tablebases
(and also John Nunn and Harold van der Heijden, who put me in touch with Jakov
in the first place); and finally, Laithwaites Wine, the eventide imbibing of whose
products provided such splendid refreshment from my daily labours on this book!

Steve Giddins,
Rochester, UK,
January 2012

12
 Chapter Three
Bishop Endings

Bishop and pawn endings come in two other colour, the player can establish a
distinct categories, depending on degree of control over those squares
whether the bishops operate on the too.
same colour squares, or squares of op- 2. If the pawns become fixed on the
posite colours. In both cases, the key same colour squares as the bishop, its
characteristic of bishops is their colour- activity is restricted and it risks becom-
blindness – a bishop can only cover ing a “bad bishop”.
squares of one colour, which means
that its influence is always limited to Game 8 shows a typical case, where
only half the squares on the board. the defender suffers from the bad
Same colour bishop endings are the bishop. The consequent weakness of
more straightforward. The usual end- Black’s dark squares, combined with
game assets are important here too – White’s more active king, allows Yanof-
passed pawns, pawn weaknesses, king sky to win a position with equal mate-
activity, etc. A particular issue in rial and symmetrical pawns.
bishop endings is the placement of the Game 9 sees White able to win by
pawns. In same colour bishop endings, virtue principally of the much more
one should almost always strive to active king, although here, too, Black is
place one’s pawns on squares of the handicapped by his queenside pawns,
opposite colour from those on which fixed on the same colour squares as his
one’s bishop travels. This serves two bishop. This game also shows a typical
purposes: device in such ending, viz. the sacrifice
of the bishop for several pawns. Finally,
1. The colour-blindness of the Game 10 is a virtuoso display by Larsen,
bishop means that it can only control who exploits the smallest of positional
half the squares on the board. By plac- advantages, against a world-class op-
ing the pawns on the squares of the ponent. The ending illustrates espe-

38
 Bishop Endings

cially well the battle between mutual his bishop. This secures them from at-
passed pawns. tack, since his own bishop can defend
Opposite-colour bishop endings are them, whilst the enemy bishop cannot
notorious for their drawing tendency, attack them.
although this should not be exagger- In order to win an opposite-colour
ated. The fundamental concept in such bishop ending, the stronger side usu-
endings is that of the fortress. Since ally needs to create a second passed
bishops only control squares of one pawn, several files away from the first.
colour, it follows that opposite- This will usually force the defender to
coloured bishop endings allow the de- block one pawn with his king and the
fender the possibility of setting up a other with his bishop. The stronger side
fortress based on the colour squares then needs to get his king over to sup-
which his bishop controls. By defini- port whichever pawn the enemy
tion, he will have an extra piece over bishop is stopping.
his opponent when it comes to the In this chapter, we have two classic
fight on those squares, and this fre- opposite-colour bishop endings, which
quently allows the defender to draw illustrate this. Game 11 is a famous
positions two, or even more pawns example of a combinational break-
down. through, where Black is prepared to
An important detail about opposite- sacrifice several pawns, in order to cre-
colour bishop endings is the pawn ate his second passed pawn. Game 12
placement. Contrary to the usual rule shows Karpov, one of the finest-ever
in same-colour bishop endings, in posi- handlers of opposite-bishop endings,
tions with bishops of opposite colour, weave a magical win against a grand-
the defender should usually put his master, from a completely drawn posi-
pawns on the same colour squares as tion.

39
The Greatest Ever Chess Endgames

Game 8
D.Yanofsky-A.Pinkus
Ventnor City 1942

1 d4 Ìf6 2 Ìf3 d5 3 c4 e6 4 Ìc3 Íe7 5 on d5, which blocks in his own prelate.
Íg5 0-0 6 e3 Ìbd7 7 Ëc2 c6 8 a3 Îe8 9 Black also has two of his kingside
Îd1 Ìf8 10 Íd3 dxc4 11 Íxc4 Ìd5 12 pawns on white squares, and Yanofsky
Íxe7 Ëxe7 13 0-0 b6 14 Îc1 Íb7 15 hurries to capitalize on this. The first
Íd3 Îec8 16 Ìxd5 exd5 17 Íf5 Îc7 18 stage of his plan is to fix all three black
b4 g6 19 Íd3 Ìe6 20 Ëb2 Îac8 21 kingside pawns on light squares.
Íe2 c5 22 bxc5 bxc5 23 dxc5 Ìxc5 24 43 Íd3 f5
Ëb4 Ìe6 25 Ëxe7 Îxe7 26 Îxc8+ Coming rather quietly, but Yanofsky
Íxc8 27 Îc1 Îc7 28 Îxc7 Ìxc7 29 Ìd4 argues that the move is practically
Íd7 30 Êf1 Êf8 31 Êe1 Êe7 32 Êd2 forced anyway. The threat is g2-g3, fol-
Êd6 33 Êc3 f6 34 Êb4 Íe8 35 Íb5 lowed by h4-h5. Black can try to bring
Íf7 36 a4 Íe6 37 Ìb3 Íc8 38 Ìd4 his bishop over, to defend the kingside
Íe6 39 Íd3 Íd7 40 h4 Ìe6 41 Íb5 pawns, but then he allows the white
Ìxd4 42 exd4 Íc8 king into a6, when the a-pawn or d-
W________W pawn will fall. Yanofsky gives the varia-
[WDbDWDWD] tion 43...Íd7 44 g3 Íe8 45 Íb5 Íf7 46
[0WDWDWDp] Êa5 Êc7 47 Êa6 Êb8 48 a5
[WDWiW0pD] W________W
[DBDpDWDW] [WiWDWDWD]
[PIW)WDW)] [0WDWDbDp]
[DWDWDWDW] [KDWDW0pD]
[WDWDW)PD] [)BDpDWDW]
[DWDWDWDW] [WDW)WDW)]
W--------W [DWDWDW)W]
Yanofsky notes that the game was [WDWDW)WD]
adjourned in this position, and the [DWDWDWDW]
other competitors in the tournament W--------W
all thought it was drawn. However, by and Black is close to zugzwang.
very fine play, Yanofsky proves other- 48...g5 (48...Êa8 49 Íc6+ Êb8 50 Êb5,
wise. White has the advantage of the followed by Êc5, wins the d-pawn) 49
more active king and the better bishop, f4 (in view of the next note, 49 Íd3 h6
principally because of the black pawn 50 Êb5 and Êc5 is simpler) 49...h6 (the

40
 Bishop Endings

computer points out 49...gxh4! 50 gxh4 47 Íe2!

Íg6! which probably draws, as White’s With the enemy bishop no longer
king gets stuck on a6) and now 50 Ía4! able to reach the e8-h5 diagonal, White
is the full “Volkswagen”, as Nigel Short can now prepare his h4-h5 break.
calls it: 50...Íe6 51 Íe8! followed by 47...Êc6
Êb5-c5. 47...h5 would obviously create a fa-
44 f4 tal weakness on g6, and after 48 Íb5
Immediately fixing the pawns on there would already be no way to avoid
light squares. loss of material, in view of the threat
44...Íe6 45 g3 Íd7 46 Íb5! Íe8. The king and pawn ending is of
At present, the black bishop is ide- course lost after 48...Íd7 49 Íxd7
ally placed on d7, from where it is able Êxd7 50 Êc5 Êe6 51 Êc6.
to swing over to either wing. White’s 48 h5
main plan is to play h4-h5, but if he W________W
prepares that with 46 Íe2, then Black [WDbDWDWD]
has 46...Íe8. Yanofsky therefore drives [0WDWDWDp]
the enemy bishop off its ideal square,
[WDkDWDpD]
forcing it to choose an inferior post.
[DWDpDpDP]
W________W [PIW)W)WD]
[WDWDWDWD] [DWDWDW)W]
[0WDbDWDp] [WDWDBDWD]
[WDWiWDpD] [DWDWDWDW]
[DBDpDpDW] W--------W
[PIW)W)W)] 48...gxh5
[DWDWDW)W] If Black does not capture, White will
[WDWDWDWD] have a pleasant choice between ex-
[DWDWDWDW] changing pawns, creating a fatal
W--------W weakness on g6, or pushing on to h6.
46...Íc8 The latter creates various tactical
Black chooses to maintain the guard threats involving the promotion of the
of a6, but now the bishop loses the h-pawn, for example 48...Íe6 49 h6
chance to come to e8, defending the Íc8? 50 Íb5+ Êd6 51 Íe8 and there is
other wing. If instead 46...Íe6, keeping no defence to the threat of Íxg6.
the option of coming to f7, then White 49 Íxh5
moves in on the queenside instead: 47 Now Black has a further weakness
Êa5 Êc7 48 Êa6 Êb8 49 Íe8 as in the on f5.
note to move 43. 49...Êb6 50 a5+! Êc6

41
The Greatest Ever Chess Endgames

Yanofsky points out that 50...Êa6 This condemns another pawn to

leads to a comic state of helplessness oblivion, but if 60...Íe6 61 Êa6 Êc6
for Black after 51 Íf7 Íb7 52 Íe6. then White frees his king by means of
51 Íe8+ Êd6 52 Êb5 62 Ía4+.
White is gradually able to inch into 61 Íd1 Êc6
the enemy camp. W________W
52...Êe7 53 Íh5 Êd6 54 Íe2 Íb7 55 [WDbDWDWD]
Íd3 Íc8 [IWDWDWDW]
W________W [WDkDWDWD]
[WDbDWDWD] [)WDpDpDp]
[0WDWDWDp] [WDW)W)WD]
[WDWiWDWD] [DWDWDW)W]
[)KDpDpDW] [WDWDWDWD]
[WDW)W)WD] [DWDBDWDW]
[DWDBDW)W] W--------W
[WDWDWDWD] 62 Ía4+
[DWDWDWDW] 62 Íxh5 Êb5 63 a6 would also have
W--------W won (63...Íxa6? 64 Íe2+), but Yanof-
56 Íc2! sky prefers not even to allow Black that
Zugzwang again. The king cannot forlorn hope of counterplay.
move without allowing the white king 62...Êc7 63 Íe8 Íe6 64 a6 1-0
into c5, the bishop cannot go to b7 be- Megabase gives the final move as
cause the f5-pawn drops, and if the 64 Êa6, but Yanofsky’s own book gives
bishop moves to d7 or e6, it allows the the text. Either way, the position is ob-
white king into a6. There is only one viously hopeless for Black. After 64 a6,
other move. Yanofsky gives the further line 64...Íc8
56...h6 57 Íd3! 65 Íxh5 Íe6 66 Íg6 Íc8 67 Íf7 Êc6
But that only staves off the moment 68 Íg8! with another zugzwang. The
of truth by one move. Now Black must pawn promotes after 68...Êd6 69 Êb6.
make a concession. A splendid piece of precision play by
57...Íd7+ 58 Êa6 Êc7 59 Êxa7 Íc8 the Canadian Champion, who was a
Black has been forced to surrender a noted endgame expert. His best game
pawn. He tries to make things difficult, collection (alas, a bibliographical rarity
by boxing in the white king, but once these days) contains a number of finely
again zugzwang defeats him. played endgames, and can be highly
60 Íc2! h5 recommended.

42
 Bishop Endings

Game 9
Y.Averbakh-S.Furman
USSR Championship Semi-final, Odessa 1960

1 c4 c5 2 Ìf3 Ìc6 3 e3 e6 4 Ìc3 Ìf6 5 while, White will penetrate with his
d4 d5 6 cxd5 exd5 7 Íe2 Íd6 8 dxc5 king on the weakened light squares on
Íxc5 9 0-0 0-0 10 b3 a6 11 Ìa4 Íd6 12 the kingside (two weaknesses!). It
Íb2 Ìe4 13 Îc1 Îe8 14 Ìc3 Ìxc3 15 seems unlikely that he should be able
Íxc3 Íc7 16 Ëd3 Íe6 17 Îfd1 Îc8 18 to do so, but with the aid of zugzwang,
Ëb1 Ëe7 19 Ëa1 f6 20 Ìd4 Ìxd4 21 his king makes it all the way to h8. Fi-
Íxd4 Íd6 22 Íf3 Îc6 23 Íb2 Îec8 24 nally, we see another typical device in
g3 Ía3 25 Îxc6 bxc6 26 Íxa3 Ëxa3 27 minor piece endings – the sacrifice of a
Îd4 Ëa5 28 Îa4 Ëb6 29 Ëf1 Îa8 30 piece, to net a winning collection of
Ëd3 Íf7 31 b4 Íe6 32 Ëd4 Ëxd4 33 pawns.
exd4 Êf8 34 Íe2 Íc8 35 f3 Êe7 36 Averbakh’s handling of the end-
Êf2 Êd6 37 Îa5 Íb7 38 Êe3 Êc7 39 game is technically flawless.
Êd2 Êb6 40 a3 Îe8 41 Îc5 Íc8 42 a4 49 g4!
Íd7 43 Íd3 h6 44 a5+ Êb7 45 Îc1 Íc8 Fixing weaknesses – now the enemy
46 Íg6 Îe7 47 Îe1 Îxe1 48 Êxe1 Êc7 kingside pawns are fixed in their place,
W________W and White can prepare to infiltrate
[WDbDWDWD] them with his king.
[DWiWDW0W] W________W
[pDpDW0B0] [WDbDWDWD]
[)WDpDWDW] [DWiWDW0W]
[W)W)WDWD] [pDpDW0B0]
[DWDWDP)W] [)WDpDWDW]
[WDWDWDW)] [W)W)WDPD]
[DWDWIWDW] [DWDWDPdW]
W--------W [WDWDWDW)]
In almost any endgame, it is a sig- [DWDWIWDW]
nificant advantage to have the more W--------W
active king, and this is a perfect illus- 49...Êd6 50 Êf2 Êe7 51 Êg3 Êf8 52
tration of that rule applying in a bishop Êh4 Íe6 53 Êh5 Íc8 54 Íd3!
endgame. Black has two weak pawns Tying the black bishop down to the
on a6 and c6, to the defence of which defence of a6.
his bishop will be tied down. Mean- 54...Êf7

43
The Greatest Ever Chess Endgames

W________W 62...Íd7 63 f4
[WDbDWDWD] Gaining further ground. White sees
[DWDWDk0W] that his bishop will no longer need the
square f5, so he can advance his pawn
[pDpDW0W0]
there, to squeeze Black further.
[)WDpDWDK]
[W)W)WDPD] W________W
[DWDBDPDW] [WDwDWiWI]
[WDWDWDW)] [DWDbDW0W]
[DWDWDWDW] [pDpDW0B0]
W--------W [)WDpDWDW]
55 h3!
[W)W)W)PD]
Zugzwang. Now Black’s king must [DWDWDwDP]
give way, since 55...Íb7 results in his [WDWDWDWD]
bishop being completely immobilized [DWDWDWDW]
after 56 Íf5. W--------W
55...Êf8 56 Êg6 Êg8 57 Íf1 63...Íc8 64 f5 Íd7 65 Íh5!
Preparing a further zugzwang. Setting up the final breakthrough.
57...Êf8 58 Íe2 Êg8 59 Íd3! 65...Íc8
Once again, Black is in zugzwang W________W
and must give ground. [WDbDWiWI]
59...Êf8 60 Êh7 Êf7 61 Íg6+ Êf8 62 [DWDWDW0W]
Êh8
[pDpDW0W0]
W________W [)WDpDPDB]
[WDbDWiWI] [W)W)WDPD]
[DWDWDW0W] [DWDWDWDP]
[pDpDW0B0] [WDWDWDWD]
[)WDpDWDW] [DWDWDWDW]
[W)W)WDPD] W--------W
[DWDWDPDP] 66 Íe8!
[WDWDWDWD] A beautifully elegant final blow. The
[DWDWDWDW] bishop sacrifices itself, on an empty
W--------W square (always the most aesthetic form
White’s king has made it to the end of sacrifice), in order to destroy the
of the earth. Now the next stage is to black kingside pawn structure.
force Black to undefend the pawn on 66...Êxe8 67 Êxg7 h5
g7. Desperation. If 67...Êe7 68 h4! (not

44
 Bishop Endings

68 Êxh6?? Êf7 with a draw) 68...Íd7 74 Êf5 Êg8 75 g7 Íf7 76 Êe5 Íg6 77
and now simply 69 g5! and the f-pawn Êd6 Íd3 78 Êxc6 Êf7 79 Êd6 Íf5
will promote. If the bishop stays on the a6-f1 di-
68 Êxf6 hxg4 69 hxg4 agonal, to prevent White’s next, then
The two connected passed pawns after 79...Íc4 White wins by 80 g8Ë+!
will divert the black king and allow Êxg8 81 Êe7.
White’s own king to penetrate to the W________W
queenside – two weaknesses again! [WDWDWDWD]
69...Êf8 70 g5 Êg8 71 g6 Êf8 72 Êg5 [DWDWDk)W]
Íd7 73 f6
[pDWIW)WD]
W________W [)WDpDbDW]
[WDWDWiWD] [W)W)WDWD]
[DWDbDWDW] [DWDWDWDW]
[pDpDW)PD] [WDWDWDWD]
[)WDpDWIW] [DWDWDWDW]
[W)W)WDWD] W--------W
[DWDWDWDW] 80 b5! 1-0
[WDWDWDWD] Now the a-pawn will decide. A
[DWDWDWDW] beautifully elegant ending by the great
W--------W Soviet endgame expert, and a highly
73...Íe8 instructive demonstration of some
73...Íc8 74 Êf4 is no better. typical endgame ideas.

45