Open Gambits - Botterill

George Botteri II
Open Gambits
Open Gambits:
Italian and Scotch
Gambit Play
GEORGE BOTTERILL
B.T.Batsford Ltd, London
First published 1 986

George Botterill 1986
ISBN 0 7134 5085 l (limp)

Photoset by Andek Printing, London
and printed in Great Britain by
Billing & Son Ltd,
London and Worcester,
for the publishers
B.T.Batsford Ltd, 4 Fitzhardinge Street,
London W I H OAH
A BA TSFORD CHESS BOOK
A dviser: R . D. Keene G M , OBE

Technical Editor: P . A . Lamford
Contents
Symbols and References

Introduction
8
9
Part 1: Greco Gambit
15
Section 1 : Moller Attack - Main Line
17
21
28
30
32
Section
Section
Section
Section
Section
Section
Section
2:
3:
4:
5:
6:
7:
8:
Moller Attack - 1 3 . . . 0-0

Moller Attack - 1 2 g4!?
Moller Attack - 1 0 . . . 0-0
Moller Attack - Black's 9th Move Alternatives
Steinitz's 9 b3 and 10 .ia3
8 . liJxc3
Has Anyone Got Any B etter Ideas?
..
35
38
42
Part 2: Evans Gambit
48
Section 1 : 6 0-0 - Lasker's Defence

Section 2: 6 0-0 - 7 . . . .ig4 and 7 . . . .id7
Section 3: 6 0-0 - Black's 6th Move Alternatives
Section 4: 6 d4 - Main Line
Section 5: Compromised Defence
Section 6: 6 d 4 d6
Section 7: 6 "t!Yb3
Section 8: 5 . . . .ic5 and the ' Normal Position'
Section 9 : 5 . . . .ie7
Section 10: The Gambit Declined
52
56
61
65
71
74
79
81
86
91
Part 3: Scotch Gambit
96
Section 1 : Main Line 5

de
Section 2: Declining with 5
d3
Section 3: 4 . . . li:lf6 (Two Knights Defence)
98
I 00
1 02
Part 4: Goring Gambit
106
Section
Section
Section
Section
Ill
. ..
. ..
1:
2:
3:
4:
Main Line with 8 . . . .ig4 !

Black's 8th M ove Alternatives
Black's 6th M ove Alternatives
5 . d6
Section 5 : 5 .ic4
Section 6: The Gambit Declined
1 13
1 18
1 20
1 26
1 30
Index of Complete Games

Index of Variations for I talian (Greco and Evans) Gambits
Index of Variations for Scotch (Scotch and Goring) Gambits
1 42
1 43
144
..
Preface
Throughout this book I have adopted an analytic , even clinical,
approach to these Open Gambits in an atte mpt to give a realistic
assessment of their prospects in the light of current skill and knowledge.
But we should not forget the commendable spirit of enterprise and
adventure which gambiteers have displayed. If chess is a game to be
enjoyed, then a sporting attitude must be at least as important as
technical accuracy. Whatever their success in future tournaments, t hese
gambits have left a rich legacy of charming variations and beautiful
games. I would like to acknowledge the encouragement supplied by
Peter Kemmis Betty and Paul Lamford and the many invaluable
suggestions and re minders that have come from my old friend Tim
Harding in the course of this attempt to record that legacy.
George Botterill
Aberystwyth
July 1 985
Symbols
+
=
;!; +
=F
H
00
!
!!
!?
?!
?
??
Ch
corres
check
equal position
slight advantage
clear advantage
winning advantage
unclear
good move
brilliant move
interesting move
dubious move
weak move
blunder
Championship
correspondence game
W or B at the side of a diagram indicates which player is to move. A

number in brackets after a move refers to the diagram of that number.
References to 'Botterill & Harding' or ' Harding & Botterill' are to the
books
G.S.Botterill & T.D. Harding, The Scotch ( Batsford 1 977)
T.D.Harding & G . S . Botterill, The Italian Game (Batsford 1 977)
References to 'Cafferty & Harding are to the comprehensive study
B. Cafferty & T.D. Harding, Play the Evans Gambit (Hale 1 976)
'
Other major sources:

Encyclopedia of Chess Openings (ECO ), Volume C
Batsford Chess Openings (BCO )
lnformator
New in Chess
Introduction
This book is devoted to four gambit lines in the open game ( I e4 e5
2 lbf3 lbc6) - the Greco Gambit (3 .i.c4 .ic5 4 c3 ltlf6 5 d4), the Evans
Gambit (3 .tc4 .i.c5 4 b4), the Scotch Gambit ( 3 d4 ed 4 .ic4) and the
Goring Gambit (3 d4 ed 4 c3). Aside from White's resolution to regard
the loss of a pawn as a mere trifle, the common factor is his early advance
d2-d4, which has two main objectives:
I. To eliminate Black's central pawn at e5.
2. To open lines as quickly as possible for the activation of White's pieces.
LAYOUT
The major innovation of this study is not in the moves, but in their
presentation . I n spite of subversive modernisms, many still insist that a
tale well told should have a beginning, a middle and an end, in that order.
The procedure adopted here inverts the traditional sequence. I start with
a straight and fairly long theoretical line and work backwards, retracting
main line m oves in the subsequent sections. 'Peel-back' analysis seems
an appropriate name.
Whether this experiment works is for the reader to decide. It has
certain advantages. Peel-back layout gives us u ncluttered access to
theoretically crucial positions, without side-tracking en route through a
maze of subsidiary variations. This should at least aid memorization of
the most important lines.
What appeals to me most about peel-back analysis is that it actually
reflects the thinking of someone who is trying to work out how an
opening should be handled. The one thing that you can rely on about
opening theory is that its current results will be unsatisfactory from
somebody's point of view. Perhaps White only gets equality when he was
hoping for more. Or a defence that you had favoured as Black is leading
to trouble. So we have the main sequence and the question is how to get
out of it. Let's consider retracting these moves and see what other
10 Introduction
possibilities are available atan earlier stage. That is exactly how peel-back
anai ysis operates.
To some extent selection of the main line is a matter of taste.
Sometimes there are rival candidates with equally good qualifications .
S o I d o n ot guarantee that the main lines (the first section o n each o f the
four gambits) constitute best play for both sides. Peel back to see
whether they do!
A MODERN ROLE FOR OLD-FASHIONED OPENINGS?
Do I come to praise these gambits or to bury them? Well, at the most

general level the verdicts are that the Greco G ambit has to be laid to rest,
the Scotch Gambit has little independent significance (since B lack can
and should switch out of it), and the Goring Gambit is a sporting chance
but ultimately unfavourable for White. Only the Evans remains fit for
active service.
So what is the use of these rather old-fashioned gambits? Hasn't
modern theory and practice renounced them for quieter and more
sophisticated methods? One motive for looking at them will appeal to
potential defenders. If you answer 1 e4 with 1 . . . e5, intending some
defence against the ever popular Ruy Lopez, you need to be prepared
against these gambits. But is there any reason why the rest of us should
bother with them ?
I think there is. F o r study and play of these gambits i s t h e ideal tactical
training ground. Training in tactics is important for everyone, and above
all for younger and less experienced players. Control of tactical
interchanges is essential to any degree of strength at chess . Without it all
the positional insight in the world will not raise you above the level of a
patzer. A common manifestation of this can be observed in matches
between teams of unequal strength. It often happens that three quarters
of the way through the session the lower-rated players have excellent
positions and alarm-bells are starting to ring in the head of the stronger
team's captain. But he need not have worried. In the last hour the weaker
players collapse and mess up all their nice positions. The tactical
weakness of the lower-rated players shows as soon as the pieces start to
clash in earnest. The pathetic spectacle of someone playing the first
twenty-five moves like a grandmaster and the next ten like a gorilla is
becoming quite common nowadays.
During a 'chat show' at the 1 983 British Championships one member
of the audience asked the panel what openings they would recommend
for beginners. There was general agreement that stodgy lines should be
Introduction 11
avoided: in chess you need to run before you can crawl with serpentine
nuance. Glenn Flear nominated the Goring Gambit as an excellent thing
to play in order to get acquainted with the power of the pieces.
In a way, this book is my considered reply to that question. My hope is
that it will be useful to coaches and that chess masters in schools will be
able to assist their pupils by setting them to play on from the diagram
positions at the beginning of the sections. Let the players then compare
the way their game went with lines given in the text. The results should be
interesting and instructive.
THEMES AND PRINCIP LES
From the positional point of view we are on the nursery slopes. There are
really no general positional themes that characterize these gambits. This
is because White's initial gambit approach opens up the centre of the
board, leaving little in the way of central pawn structure. A happy
hunting ground for those who dislike blocked positions with interlocking
pawn chains.
Everything depends on the relative activity of the pieces. Special atten..,
tion should be given to the role of White's queen's bishop, one of the
most important attackers, particularly when its dark-squared opponent
is exchanged off or driven back to a5 or b6. Watch out for the move
.tci-a3, which crops up quite often. By controlling the a3-f8 diagonal
White can sometimes prevent Black from castling, or pin a knight on e7,
or drive a rook away from defence of f7.
That point f7 is, of course, Black's traditional weak spot in open
games. In the following pages we repeatedly see White lining up on it,
most often with .tc4 and 'tWb3. It is worth noting, however, how often
Black can respond with ... lila5, allowing f7 to fall , but eliminating
White's light-squared bishop. For an example of this procedure look at
our main line in the Goring Gambit.
Is modern defensive technique killing off these gambit openings? If so,
it would be nice to know what the main principles of defensive play are;
In fact I think the increased strength of defence is not so much an
application of principle as a matter of'storage'. Once you have seen what
has gone wrong in the past you can avoid it in the future, and so defences
get toughened by a process of elimination. Perhaps the single most
important thing for the defender to bear in mind is that the timely return
of material is a vital resource. Examples abound - Lasker's Defence to
the Evans Gambit ('Evans Gambit with 6 : Lasker's Defence') being a
classic illustration.
12 Introduction
But it really is impossible to generalize about when material should be

returned and when it should be retained. Quite often Black gets the
opportunity to grab a second pawn. Sometimes this is suicidal, but on
other occasions it is the right thing to do (e.g. Goring Gambit: 5 .i.c4).
The ultimate lesson to be learned from these gambit openings is that
there is a balance between the elements of Material, Time, Space and
Co-ordination. A deficit in one factor can be compensated by advantage
in one or more of the others. But where exactly the balance lies has to be a
matter for on-site judgement.
THE EVIDENCE OF GAMES
We have to learn from results in practice, but games cannot always be

trusted. There is a strong temptation to infer that the side that won must
have exploited an advantage that was already there, especially when the
manner of the victory is striking and brilliant. Such inferences are risky.
Look out for the places where the annotator writes something like: 'This
loses, but it was hopeless anyway.' You should always pause to consider
whether that is really so.
The gambits in this book are offered quite frequently in postal play probably with a greater relative frequency than in over-the-board
tournaments. One ought to be wary of reading too much into the results
of correspondence games. Ideally correspondence play can be conducted
at the highest level without the time-trouble errors that mar so many
over-the-board encounters . But one of the charming features of
correspondence chess is that it gives weaker players a chance to debate
interesting theoretical lines . And if they are weaker, funny things are
liable to happen.
To be fair to the correspondence fraternity I will append a couple of
examples of odd goings-on at the board. Since these gambits are no
longer in fashion the theoretically important games may often be quite
ancient. What reliance can be placed upon the validity of 1 9th century
successes?
In general we should not be too sniffy about the quality of old games.
The old masters knew a lot about lines that their successors have
neglected. But I am worried about the quality of the Chigorin-Steinitz
encounters, which occupy an important position in the history of the
Evans Gambit. Consider this position, which arose after Black's 20th
move in the 1 7th game of the 1 889 Chigorin-Steinitz match (for the
opening moves see 'Evans Gambit with 6 0-0: Black's 6th move alter
natives', variation B):
Introduction 13
w
With queen, rook and bishop huddled together out of play Black
ought not to survive for long.
21
lt:lxeS+!
'it>g7
After 2 1 . . . fe 22 f4 White's major pieces would quickly kill off Black's

lonely king.
22
lt:lc4
bS
White was threatening not only 23 llJxb6 but also 23 e5 . So Steinitz

makes a desperate try to bring his queen at least into the game. 23 lt:lb6
is good enough now, but the move Chigorin actually played is j ust as
strong.
23
24
ab
b6
'it'a7
'it'a4
What a bishop on c8! For all practical purposes White is a piece up

and should win easily. Incredibly Chigorin not o nly failed to exploit his
advantage, but dropped the pawn at b6 and allowed Black's bishop to
come into play. In the end he was lucky to escape with a draw.
Three years later this position occurred after White's 23rd move in the
7th game of their second match (see ' Evans Gambit with 6 0-0: 7 -*.g4,
7 . . . i.d7' , variation B l for the opening):
. ..
ii
B
14 Introduction
White is somewhat better because Black owns weak pawns at c5 and
e5 and the bishop on b6 is a problem piece. But Steinitz has some
counterplay on the kingside - which he botched horribly with:
23
24
25
xg2
'tlfxf3
lD xg2??
.i. xf3+
'ti'gS+
With the 'point' that he wins the rook on d2.

26
27
<ot>hl
'ifxf7+
'itxd2
h7
Perhaps Steinitz thought there was nothing worse to fear than 28 'itxe8
'1Vxb2 - though even then he is mated in three!
28
ll:gl
1 -0
GRECO
GAMBIT
1
I
2
3
4
5
e4
ti:lf3
c4
c3
d4 (1)
eS
tt:lc6
cS
lDf6
The name 'Greco Gambit' is not a standard label for this early central
rupture. But I think it ought to be as Greco supplied some of the analysis
(given here in section 7) that provided the initial motivation for the
gambit.
The analytical biography of the gambit is a sad story, with no hope of a
happy ending for the gambiteer. In a way that is in keeping with the fate
of Gioachino Greco ( c. l 600-c. l 635). There is a tale that, having won
some 5,000 crowns by overcoming France's leading players in 1 62 1 , he
was robbed of all his prize money by outlaws while visiting England in
the following year. During his lifetime he peddled manuscripts on chess
openings to wealthy patrons. When these were collected and published
after his death they became enormously influential for more than a
century. Yet in the long run his pioneering analyses were received with
little gratitude, since masters who came after him - like Stamma and
Philidor - were more eager to stress their own superiority than to accord
credit that was due.
Unfortunately, there is n o way of retrieving this gambit from the pil e
of opening discards. Section 7, including Greco's contributions, is
16 Greco Gambit: Introduction

redundant now because after 5 ... ed 6 cd .i.b4+ 7 ll:lc3 ll:lxe4 8 0-0 Black
plays not 8 .. . ll:lxc3 (though even that need not lose - see section 7,
variation B) but 8 . .. .txc3! . Any hopes White might then have enter
tained after 9 be were squashed by Lasker in his 1 896 match against
Steinitz (see section 6).
After this attention turned to 9 dS - the Moller Attack (sections 1 -5).
I t seemed that this clever idea might keep the gambit alive. Detailed
analysis finally stabilized towards the opinion that it should all end in a
draw after 9 dS .i.f6 10 I!el l'De7 I I l he4 d6 I2 .i.gS .txgS 1 3 lL:lxgS 0-0
14 ll:lxh7 (as in section 2).
That in itself puts off the fi rst player who has healthy ambitions. But
there is worse to come. After all, 9 d5 is a very artificial move,
obstructing White's own bishop on c4. Black need not rest content with
1 3 . . . 0-0. 13 ... h6, as in s ection I, gives the second player excellent
winning chances.
As the Greco Gambit is practically moribu nd an illustrative game
would be out of place here. For model play by Black look at Barczay
Portisch in variation C of section I. For White there is nothing to
recommend. Section 8 surveys some ki ndred lines in Italian gambit style.
But they do not inspire confidence.
So I am forced to conclude that White ought to peel away the gambit
move 5 d4 and, if he is to play the Italian G ame at all, do so in its 'very
quiet' form ('Giuoco Pianissimo') with 5 d3 . This may not suit people
who like gambit openings, but there is life in the idea, favoured by the
idiosyncratic 1 9th century master H . E . Bird, of early expansion on the
queenside with b2-b4 and a2-a4. For example, 5 d3 d6 (Some prefer 5 . . .
a6 precisely because it does not encourage b4 and a4, e.g. 5 . . . a6 6 0-0
d6 7 I!e l .i.a7 8 .i.b3 0-0 9 lLlbd2 .i.e6 has been seen in several games.)
6 b4 .i.b6 7 a4 aS 8 bS ll:l e7 9 0-0 0-0 10 l'Dbd2 ll:l g6 11 .ta2 (2)
2
B
Here 'quietness' has been acco mpanied by avoidance of exchanges
17
Greco Gambit: Introduction
and a steady build-up of tension . A complicated struggle will develop

after II . . c6 (Miles-Nikolac, Dortmund 1 979) or 1 1 . . . .ig4 (Pinter
Mestel, Las Palmas IZ 1982) .
.
Section 1
Moller Attack:
Main Line
The position that is the focal point
of our investigation into the

Moller A ttack arises after the
moves:
1
2
3
4
s
6
7
8
9
10
11
12
13
e4
lt:Jf3
..tc4
c3
d4
cd
c3
0-0
dS
lle1
ll xe4
..tgS
li:J xgS
analyses have firmly established

the conclusion that after Black's
alternative 13 . . 0-0 a draw is
de monstrable. But can Black legi
timately try for more ? In an earlier
book (The Italian Game, 1 977)
Tim Harding and I boldly gave 1 3
. h6 an unqualified exclamation
mark. In so far as these symbols
have an objective value, I still
think that is probably the correct
selection. But here I more cautious
ly insert '!?' because in playing this
move Black does incur some risk
of losing. The resul tant variations
have not been so exhaustively
researched as those ste mming
from 13 . . . 0-0 and White has some
messy tries, which may not be
convi n cing and yet are certainly
confusing.
Obviously, retreating the attack
ed knight will not appeal to White.
After 14 li:Jf3 0-0 Black is a solid
pawn to the good with no prob
lems. So I analyse :
eS
lt:Jc6
..tcS
f6
ed
.ib4+
xe4
.ixc3
.if6
li:Je7
d6
.ixgS
h6!? (3)
3
w
A 14 1!t'e2
B 14 'ti'hS
C .14 ..tbS+
D 14 li:Jxf7
A
It is the assessment of the move

13 . . . h6!? that is really the
theoretically interesting question
about the Moller Attack. Previous
14
't!te2
Nothi n g could be more natural

than this attempt to smash
through along the open file.
18 Greco Gambit: Section 1

Black, how ever, has a resource,
blocking the file with a counter
sacrifice:
14
15
16
lle1
de
hg
i.e6!
f6 + (4)
4
w
Moller Attack
Main
Line
ordinated and the black pawns

rather vulnerable ( l-0, 3 1 ). In fact
the endgame is tenable with good
defence, but we need not go into
that. O ne commentator actually
claimed that 15 . . . i.d7? is the only
move. The truth is that only 15 . ..
1!fd7 is worse. Black could perfectly
well play 1 5 . . . 'it>f8 16 lle I .i.e6 1 7
d e f6 1 8 i.d7 c6 with the same
basic advantages as in the previous
diagram - although it may be a
little more difficult to unwind with
the bishop on d7 and the king on
f8. However, that need not
concern us either. 15 i.b5+ is
immediately refuted by 15
c6!
1 6 de (or 1 6 ll e 1 i.e6! ) 1 6 . . . 'it>f8 ! .
...
This is Zek's analysis, sub

sequently endorsed by Keres,
amongst others. The pawn at e6
may. be something of a thorn in
:Qlack's position, but as far as
White is concerned it is a total
road-block! The extra pawn should
tell in the end, in spite of Black's
backward development, thanks to
the fact that the solid black pawn
phalanx prevents any incursion by
the white pieces.
It might seem that this verdict
on 14 't!fe2 has been overturned by
the correspondence game Girod
Multala, 3rd European Team Ch
Preliminaries. Here White came
up with 15 i.b5+ and after 15 . . .
i.d7 1 6 ll e 1 i.xb5 1 7 lilxe7+ eMS
18 1Wxb5 'it'xe7 19 ll xe7 xe7 20
\i'xb7 ll hc8 2 1 h3 emerged with a
queen v two rook endgame in
which the rooks were po01"1y co-
B
14
1!fh5
The obvious drawback to this

move is that it does not prevent
castling.
14
15
0-0
llae1
Where should the black knight

go? Zek gave 15
ltlf5, with the
point 16 lL!xf7 1!rf6 ! . We can add
that after 16 li:lf3 g6 1 7 1!rh3 h5!
takes out the g2-g4 threat, and
that the combinative attempt 1 6
li:le6? ! fe 1 7 d e i s refuted b y 1 7 . . .
lL!e7 or 1 7 . . . d5 (but not 1 7 . . .
'fke7? 1 8 llf4 c6 1 9 l1xf5. lle8? 20
llf7 1 -0 Bateman-Boisvert, corres
1 9 84). Yet White has better i n
1 6 ltlh3 !?, allowing the queen to
withdraw along the h5-d 1 diagonal.
The situation is not so clear then:
the kni ght on f5 and bishop on c8
...
Moller Attack Main Line 1 9
Greco Gambit: Section 1
do not combine well.

So I prefer 15 lt.Jg6. There are
some tactical shots, but nothing
that works for White, e.g. 16 liJ xf7
f6 + again. On 1 6 lt.Je6 fe 1 7
xg6 e 5 (but not 1 7 . llf6? 1 8 d e
e7 1 9 f7+) w e find that 1 8 llxe5
does not suffice: 1 8 . . . de 1 9 d6+
h8 20 i.d3 i.f5 ! . So White
has nothing better than 16 liJf3
f6 , with indisputable advantage
to Black.
...
. .
i.d7
14 i.b5+
A brilliant piece of analysis by
Sozin (Sovremenny Debyut, 1 940)
eliminates the more loosening 1 4
. . . c6?!: 1 5 liJxf7! (the obvious 1 5
de? 0-0! i s n o t s o good) 1 5 . . . xf7
1 6 'it'f3+ i.f5 17 de be ( 17 . g6
18 i.d3 !) 1 8 i.xc6 llc8 19 ll ae 1
lt.Jxc6 20 1!fxf5+ g8 2 1 f4 (5)
.
5
B
2 4 b4! also leaves Black helpless to

prevent a decisive penetration by
the white rook. The best defence
seems to be 2 1 . . . d5 when it would
be a mistake for White to allow
the further advance of the d-pawn,
e.g. 21 . .. d5 22 ll e3? d4 23 ll e4 d3
24 1!Vg6 d2 25 lle8+ xe8 26
ll xe8+ llxe8 27 1!t'xe8+ h7leaves
White struggling for the draw. But
after the simpler 22 lle8+! 1!t'xe8 23
lixe8+ ll xe8 24 1!fxd5+ h7 25
'it'xc6 White has all the winning
chances.
1 5 'it'e2
Since this is unsatisfactory
White might feel impelled to lash
out here with 15 lt.Jxf7!? xf7 1 6
't!t'f3 +, which gives Black oppor
tunities to err. 1 6 . . . g6?, for
example, loses neatly to 1 7 llxe7
'flxe7 1 8 i.d3+ g5 19 h4+ xh4
20 'irg3+ h5 2 1 i.g6 mate. Also
weak is 16 . .. lt.Jf5? 17 i.xd7 xd7
18 lle6 when White regains the
piece with advan tage. B est is 16 . . .
g8 1 7 llae 1 (6)
6
B
It is hard to be sure about such

complex tactics, but I have not
been able to find any improvements
for Black on moves 1 5-20 and the
final position is certainly awkward
for him. White's main threat is
22 g6 (intending lie8) and 2 1 ...
'irf8 does nothing to stop this. 2 1
. . . 11rb6+ 22 h 1 'trb8 23 9g6 f8
This is one of the messy

possibilities I mentioned at the
beginning of the section . No
doubt White's sacrifice is inade

quate, a mere swindling try.
'Seeing' that is easy enough:
proving it is tougher. 17 . . . lLl xd5??
is a blunder because of 1 8 i.c4 and
19 lle7. But 17 . . . tbf5 ( H arding&
Botterill, 1 977), 1 7 . . . lLlg6 and 1 7
. . . i.xb5 are all plausible. They
are probably all good for Black,
but I cannot hope to give exact
demonstrations here. For example,
after 17 . . . i.xb5 1 8 ll xe7 'fi'f8 1 9
'fi'e4 '@f6 20 a4 llf8 2 1 f3 i.a6 22
b4 White still has rather awkward
pressure. 1 7 . . . lLlf5 is less obscure.
It is quite important that after
17 . . . lLlf5 18 i.xd7 '@xd7 19 lU4
Black breaks the bind with 1 9 . . .
lle8 ! , whilst 19 g4 lLl h4 gives
Black time to parry lle4-e7. Least
murky seems 1 7 .. . tbg6, after
which I cannot see any compen
sation for White's piece ( 1 8 1Wg3
i.xb5 1 9 1!fxg6 'irf6 H; 1 8 i.xd7
1!rxd7 19 't!rg3 'ti'f7) .
15
i.xb 5 !
This is t h e move that makes 1 3

.. . h 6 the crucial test o f the Mciller
Attack. After 15 . .. f8 1 6 ll e 1
Black's developmental problems
give White compensation for the
pawn:
a) 1 6 . .. lLl xdS? 1 7 lLlxf7! xf7 1 8
.tc4 c6 1 9 lle7+ g8 20 'irf3 Unzicker. (Co mpare 1 7 . . . tb xd5??
from the previous diagram.)
b) 1 6 . . . tbg8 17 lLlf3 lLl f6 18 lie7
and if now 18 . .. i.e6 White
smashes through with 19 lixe6 fe
20 tbd4 ! , when Black i s in big
Moller A ttack Main Line
trouble (20 . . . 'ti'e7 2 1 lLlf5).

c) 16 ... lLl g6 ! ? 17 lLlf3 ( 17 lLle6+
seems to be a near miss after 17 . . .
fe 1 8 d e i.xb5) 1 7 . . . c 6 1 8 i.c4
with continuing complicatio ns.
16
'ii'x bS+
17
'ti'e2
"t!fd7 (7)
7
w
White lacks convincing alter

natives:
a) 1 7 'it'xd7+? xd7 18 lLl xf7 ll hf8
H - Portisch.
b) The exchange sacrifice 17 llxe7+
xe7 1 8 lie!+ d8 1 9 't!rxb7 llc8
is a tempo short: White needs his
knight to be back on f3 already for
tbd4-c6 (20 lLlf3 c5).
c) 17 "@'d3 hg 18 llael 0-0 19 lixe7
llfe8 20 ll7e3 (Wolf-Laue, l Oth
German Junior Corres Ch 197 1 )
and now 2 0 . . . c5 ! (Wade) leaves
Black a sound pawn up.
d) 17 'it'xb 7 0-0 18 ll ae l lLlg6 1 9
lLlf3 lifb8 2 0 "ii' a6 ll xb2 i s the line
that gives White the best chance of
holding the draw, although both
2 1 tb d4 lbe5 22 f4 tbg6 (Laue) and
2 1 h4 'fi'b5 leave Black clearly on
top.
17
f8!
Greco Gambit: Section

18
lbxf7
White has little choice, since

Black is just two pawns up after
1 8 lDf3 lDxd5 .
18
19
lle1
xf7
lOgS!
Estrin calls this 'an unclear and

complicated position' ( Compre
hensive Chess Openings, Vol . 1 ),
but so far as I can see White is
simply lost. The king is going to f8
and the knight is coming to f6,
after which . . . li eS will simplify.
The game B arczay-Portisch, Hun
garian Ch 1 968 concluded: 20 lie6
f8 21 f4 lDf6 22 lie7 lieS 23
lixe8+ \txe8 24 'ii'f2 'tlt'h5 0- l .
All White's thrusts deftly parried,
an efficient defensive performance.
D
14
lbxf7
xf7
15 'it'f3+
lDf5
There are obvious objections to
other defences:
a) 15 ... g6? 16 lixe7 wins.
b) 15 ... g8 /e8 16 liae l .
c) 15 . . . i.f5 1 6 liae l lie8 1 7 i.b5 ! .
16
lie6!? (8)
8
B
Moller Attack Main Line 21

1 6 g4 is none too good because
Black can calmly reply 16 . . . lil_f8
1 7 gf g8 =t=. The exchange
sacrifice 1 6 lle6 gives White
assorted tactical chances, but it
can hardly qualify as adequate.
Black can choose between 16 . ..
i.xe6 1 7 de+ e7 1 8 't!Yxf5 \te8 !
and 1 6 . . . g 6 1 7 g 4 i.xe6 1 8 de+
e7 1 9 gf gf 20 'it'xf5 \tg8+ 2 1
h 1 't!fh7 ! .
Conclusion: 1 3 . . . h6 gives Black
excellent winning chances against
the Moller Attack.
Peel-bac k:- To determine whether
this invalidates the whole gambit

approach with 5 d4 we must peel
back to see if any earlier deviations
offer better prospects. This is
rather depressing from White's
point of view since it is pretty clear
that he will have to retract every
thing at least as far back as 9 d5,
the move that characterizes the
Moller Attack. By contrast Black
does have options on moves 9- 1 3 .
I n particular there is 1 3 . . . 0-0
which, as I said earlier, leads to a
draw. The next section checks this
out.
Section 2
Moller Attack:
13 ... 0-0
Simply regaining the piece with
1
2
3
e4
lijf3
.tc4
e5
lbc6
i.c5
ll GJYco Gtlmblt: Stction 2
Moller Attack 13 . . . 0-0
10f6
ed
.tb4+
lb xe4
.i x c 3
16
lith4
rs (I OJ
16 . .. f6 appears to lead even
4
5
6
7
8
9
10
11
12
13
c3
d4
cd
li)c3
0-0
dS
liel
li xe4
.igS
li)xg5
.if6
lbe7
d6
.ixg5
0-0 (9)
9
w
We can remark en passant that

13 . . . .if5? would be a fatal
mistake, losing to 1 4 'it'f3 ! ( 1 4 . . .
1!fd7 1 5 .ib5 ! 1!Vxb5 1 6 1i'xf5 ++I
14 . . . .ixe4 1 5 1!Vxf7+ 'it>d7 1 6
'ii'e 6+ 'it>e8 1 7 'ti"xe4 'it'd? 1 8 li e 1
a6 1 9 ltl xh7 'it>d8 20 lb g5 lieS 21
lbe6+ 'it>c8 22 ltl xg7 1-0 E mery
Menchik, Biarritz 1939).
14
lbxh7!
Nothing else is going to trouble

Black. Now he has an option
between:
A 14 ... xh7
B 14 . .ifS
.
A
14
IS
1!Vh5+
ct>xh7
ct>g8
more directly to a draw in view of

the variation 1 7 g4 lieS lS .id3
'it>f8 19 'ii'h S + lDgS 20 .ih7 'it>f7 21
.ig6 + ! 'it>f8 and White has a
repetition with 22 .ih7, but
nothing more.
10
w
The reason why White does not

deliver mate is that the bishop on
c4 is not properly participating in
the attack - blame it on 9 d5, if you
wish. White can try to work the
bishop into , a sniping position
with the manoeuvre .ie2-h5 , for
were the bishop already on e2
1 7 'ifh7+ 'it>f7 IS Ith6 followed by
.ih5+ would be decisive. As it is,
the .ie2-h5 switch devours precious
time and when the bishop stands
on e2 it annoyingly obstructs a
rook on e l .
17
'ii'h 7+
There are several other attempts,

but they have been heavily analysed
and all except 1 7 lih3 (and actually
I think that fails too: see below)
have been found wanting. Given
that this position has been subject-
Moller Attack 13 . 0-0 23

ed to independent scrutiny by the
likes of Maroczy, Bogoljubow,
Keres, Euwe, Vu kovic, Evans and
Estrin, one might think that the
probability of someone's coming
up with a novel coup that will
enable White to drive home his
attack must be vanishingly small.
Well, if you really work away at a
complex tactical position like this,
it is surprising and revealing to
find how many errors are commit
ted and perpetuated by famous
names.
a) 1 7 i.e2 l::l e 8! 1 8 l::le l (or 1 8
'iih 7+ f8 ! and both 1 9 l:th6 lbg8
H and 19 i.h5 g8 + are White
failures) 1 8 . . . f8 19 i.b5 (11)
11
B
<t>e7 26 1!t'e6+ and 27 l::l x g8 :::!) 24

i.b5 25 l::l h 8+ ltlg8 and
White has at least a draw in hand
(also at most?) with 26 1!t'f5+ <:J/;e7
27 1!t'g5+ 'it>f7 28 't!t'h5+ etc.
a2) 19
c6! 20 l:te6 (inadequate,
but 20 de be is no better) 20 . . .
i.xe6 2 1 d e ltlg6 ! 2 2 'iixg6 1!t'f6
H (von Feilitzsch) .
b) 1 7 l::l e 1 ltl g6 1 8 l:th3 l:tf6! (18 . . .
f4 i s regularly queried because of
19 l::le 6!?, though even this is not
particularly convincing after 19 . ..
l::lf6) 1 9 't!Vh 7 + <J;; f7 20 n e6 is a line
that Keres experimented with in
some 1 930s correspondence games.
But Vu kovi{: refutes it with 20 . . .
ltlf8 ! 2 1 't!Vh5+ g6, since White just
runs out of pieces after 22 "@h8
i.xe6 23 de+ l:txe6.
c) 1 7 g3 cuts out . . . ltlg6 but is
painfully slow. O ne good response
is 1 7 . . i.d7 1 8 l:t e l l::lf6! 19 nxe7
1!fxe7 20 1!t'h8+ <Ji;f7 2 1 1Wxa8
1!t'e l + 22 g2 f4! (Evans) with a
crushing counter-attack.
d) 17 l::l h3 (Keres) was intended as
a variation leading to a draw after
1 7 .. . f4 ! 1 8 g4 fg 1 9 1!Vh7+ <J;;f7 20
1!t'h5+ etc. However, this is
brought into question by 18
i.xg4! 19 't!Vxg4 1!t'c8 (Campbell
Mendoza, corres 1 946) when
Black has an extra pawn and no
great defensive problems. I n fact it
is not so clear that there is any
drawing continuation available
for White after 1 7 l::l h 3 f4 ! . Euwe,
for example, carelessly gives 18
1!t'h7+ <:J;;f7 1 9 1!Yh5+ <Ji;g8 20
i.fl
...
a I) The variation sometimes given

(embarrassing to confess, in H ar
ding & Botterill l 977, for example !):
19 . . i.d7 20 l::le 6 ltlg8 (Awarded ! ,
but ? is right. Black ought to
submit to 20 . . . i.xb5 21 l::l f6+ gf
22 't!fh6+ and perpetual check.) 2 1
'i!t'xf5+ 1!t'f6 +' is wrong. The
truth is that White responds with
22 l::l xf6+ lb xf6 23 1!t'g6 ! lite I+ (23
. . . i.xb5? 24 l::l h 8+ lDg8 25 1i'f5+
.
'
...
Moller Auack 13 .
..
0-0
Wh7+ with perpetual, on the

grounds that 19 . . . ltlg6 fails to 20
18 lith6!
18 Il e1 once again allows 1 8 . . .
.td3 'tWf6 2 1 g5. He has overlooked

that the pawn has not got as far as
g4 yet ! Let us look at the position
after 17 Il h3 f4 1 8 't!rh7+ f7 1 9
1!t'h 5 + (1 2) for a moment:
llJg6! 1 9 lih6 'it'g5 and Black

defends comfortably. But now
there is a serious threat of .ie2-h5+.
So Black's next is the only
acceptable move, defending g7
and vacating f8 .
12
B
18
19
lig8
li[e1 (1 3)
There is nothing m 19 .tel

lLlxd5 20 .ih5+ f8 2 1 li[e 1
.id7 H.
13
B
1 9 . g8 20 '@h7+ just repeats,

so if Black is going for a win he
must try either 19 . . . f6 or 19 . . .
ltlg6. 1 9 . . . f6 20 g4 !? (not 20
'fi'h4+ g5 2 1 '@h6+ ltlg6 22 Ilh5
Ilh8 ! ) looks exceedingly risky and
exceedingly messy . A clearer line
would be desirable. So what of 19
. . . ltl g6? After 1 9 . . . ltlg6 20 .id3
1i'f6 White has 2 1 g4 ! when 22 g5
really does become a very awkward
threat. 21 . . . Ile8!? 22 g5 Ile5 23 gf
Ilxh5 24 Ilxh5 ltle5 (intending . . .
.ig4) might just hold, but 25 .ie2!
f3 26 Ilxe5 de 27 .ixf3 is at least!.
The clean defensive solution lies in
2 1 . . . .ixg4 ! 22 'it'xg4 ltle5 =f.
So we can see that, although a
few theoretical variations need to
be amended, tries other than 1 7
'@h7+ are tricky but unimpressive.
17
f7
..
Here Black has three moves

worth considering:
A1 19
A2 1 9
A3 19
.id 7?!
. 'iff'S
...
..
...
f8
A1
19
.id7! ?
A natural move, but dubious

because it allows a beautiful com
bination. However, the outcome
is not so clear as previous theory
has made out.
20
21
22
liee6!
.txe6
de+
e8
litg6! (14)
Moller A ttack 13 . .. 0-0 25
Greco Gambil: Section 2

14
B m &
-
.
" \1111
EJ'
B [IABn
,.
Bi..B B R

p,q
og
WA
o
-m"
' White wins' says Estrin. Might

be true, but one would like to be
told how . Evans is just a little
more helpful, adding '22 . . . c6 23
lii xg7 wins' . True- but is 22 ... c6
best? Well, 22 ... 'it>f8 is no
improvement because of 23 lii f6+
'i!?e8 24 lii f7 with the same basic
threat of lii xg7 followed by decisive
checks on f7 or the back ran k . But
try 22 ... d5!. This threatens 23 .. .
de and more significantly 23 . . .
lt:lxg6 24 'it'xg6/g8+ <t>e7 followed
by escape to d6. Therefore White
must play 23 lii xg7. Now 23 . . .
'it'd6 ! (only move) and instead of
some instant mate we find that a
queen ending is in prospect. Now
24 lii x g8+ ltl xg8 25 't!rf7+ d8 26
1i'xg8+ 'i!?e7 27 'it'xa8 (note that 27
'it'f7+ 'i!;>d8 28 1!Vg8+ with perpetual
is available a nd that here 28 b5 !
forces 28 . . . c6) 27 . . . de results in
an ending in which Black's chances
do not seem too bad . So is ' White
wins' false? No. From the diagram
best play is: 22 . . . d5 23 l hg7 "t!fd6
24 i.b5+! (but note that the same
position can be reached by 24
Iixg8+ lbxg8 23 .t b5+ or even, as
hinted above, 24 Ii xg8+ lbxg8 25

1!f7+ 'i!?d8 26 Wxg8+ 'i!?e7 27
't!t't7+ d8 28 .tb5 ! ) 24 . . . c6 25
llxg8+ lt:lxg8 26 t!t'f7+ 'i!?d8 27
'it'xg8+ 'it>c7 (27 . . . 'i!?e7 28 1!Vxa8
cb 29 Wxb7 ) 28 Wxa8 cb and
now I think we really can s ee that
after either 29 'W'e8 or 29 1!Vxa7
White does win - eventually .
A2
19
This m ove i s the simplest,

generating a forced variation in
which White has to take a draw by
perpetual check.
20 .tb5
The gallant winning attempt 20
i.e2?! 'i!?e8 21 .tb5+ d8 (not 21
. . . 'i!?f7? 2 2 llee6 ) 22 Ii f6 gf
(and certainly not 22 . . . Wxf6 23
1!fxg8+ ltl xg8 24 lieS mate) 23
llxe7 just falls short because after
23 . . . c6! 24 de be (not 24 . . . 't!fxe7
25 c7+! ) 25 i.xc6 1!Vxe7 26
1i'xg8+ c7 White cannot recap
ture the second rook he has
sacrificed, owing to the back ra nk
mate. I thi n k this is rather a
shame.
15
w
20
llh8! (15)
Moller A ttack 13

Necessary, since White was
threatening to win with 2 I l:Iee6
(intending lU6+) .
2I
22
23
24
'fi'xh8
'fi'h7+
nxe7
'fi'xh6+
..
. 0-0
16
B
gh
f6
1Wxe7
and White has perpetual check

(24 ... e5 25 'fi'e3+).
A3
f8
I9
20
lith3
.id7
20 ... f4 21 llh4 g5 is prettily
refuted by 22 llh6 lD f5 23 l:Ie8+ !
forcing mate in a few (Est rin).
2I
22
llhe3
.id3
lDc8
'fi'f6!?
The obstinate try to p reserve

some winning chances. In practice
the most sensible way to play
would be 22 ... g6 23 h4 l:Ig7 24
"W'h8+ llg8 and neither side can
escape the drawing mechanism
(e.g. 25 "t!rh6+ f7 26 li[e6 fails to
26 . . . "t!ff8!). By contrast, definitely
to be avoided is 22 . lDb6 23 .ixf5
.ixf5 24 'irxf5+ "t!ff6 25 no! and
because of the picturesque mate in
the f- and e-files Black loses his
queen .
..
23
24
25
26
27
.ixfS!
liteS+
"t!rxfS+
"t!fgS
lle3 (16)
WxfS
.i xe8
.if7
g6
'And White wins' says Estrin.

This time one can understand why
he thinks so: 27 . . lDb6 28 't!Ve7+
g7 29 li[f3 lita/gf8 30 Wf6+ and
31 llh3 : or 27 ... g7 28 llf3
.
::. Meanwhile White threatens

28 'iff6. Nonetheless there is a
defence of sorts in 27
l:Ih8!?,
although one would have to be
pretty pig-headed to choose this
in preference to the draw in A2 or
22 . .. g6 in the present line. For
after 27 . . . l:Ih8!? 28 't!Vf6 ll h7 29
1!Vd8+ r$Jg7 30 1txc7 :White, with
queen and two pawns versus a
disorganized combination of rook,
bishop and knight, is at any rate in
little danger of losing.
B
..
I4
IS
.ifS
l:Ih4
The interesting move. IS l:Ixe7

"t!rxe7 16 lD xf8 llxf8 looks quite
equal, although the position is by
no means dead - given the
asym metrical pawn structures,
some difficult endgames could
ensue.
IS
I6
lle8
'ifhS
Keres suggested I6 lDgS lDg6 I 7

llh5 , when White hopes for 1 7 . . .
lDf4?? 1 8 ll h8+ o r 1 7 . . . tOeS?? 1 8
.lt:\xf7. But against sensible moves

tangling up the white pieces like
this is highly suspect. A simple
and good reply is 17 .. . 1Wd7 ,
which I am inclined t o prefer t o 1 7
. . . li e 5 1 8 f4 li e7 (again not 1 8 ...
lt:\xf4?? 1 9 li h8+) 1 9 g4?! .ie4 20
't!t'fl c6! + Weissleder-Ungnand,
corres 1966-7, on the grounds that
this is primarily an example of
White suffering from self-inflicted
wounds.
16
lt:\g6
17
li d4
After 1 7 lt:l gS 1Wf6! Black wins
material. There is a great danger
in this line that White will be
tempted into concentrating (per
haps I should say 'de-concentrating')
too much fire-power in a futile
demonstration down the h-file,
whilst Black calmly takes over the
centre and thus obtains dominating
control.
17
18
:es
f4(1 7)
17
B
Moller A ttack 13 . .. 0-0 27

because of 20 ... 't!Vxf6 and 20 1tf3
\t;xh7 2 1 ..id3 .i xd3 22 't!t'xd3+
g8 is a position in which it is
questionable whether White has
enough for the pawn. However,
White can hold the balance with
20 't!t'h3 't!t'c8 (Vukovic) 2 1 lt:lf6+ gf
and now:
a) 22 't!rh4 f5 (intending . . . 1!Yd8)
retains tension, though I think
Black is better.
b) Simply 22 1Wxc8+ llxc8 23 llxf6
is = .
Less satisfactory, from Black's
point of view, is 18 ... lie4 in view
of 19 lixe4 ..txe4 20 ll:lg5 'tt f6 21
't!rh7+ f8 22 ll:l xe4 't!rd4+ 23 lt::J f2
't!t'xc4 24 f5 lt::J e 7 25 f6 ! gf 26 1Wh6+
g8 27 "t!Vxf6 a s in Andersson
Johansson , Swedish Ch 1 969.
This is t or , though the position
remains complicated with both
sides having difficult moves to
find. The game continued 27 . . .
lt::J g6 2 8 lie! f8 (but maybe 28 . . .
't!Vxd 5 ! ? - i t is n o t clear why Black
should 'trust' White) 29 lt::J e 4 lle8
30 li[fl ll:l e5 3 l lil g3 Wd4+ 32 'i!lh 1
't!Vd3 33 li c l 'it'h7 34 h3 lt::J d 7 35
'ti'd4 lt:\c5 36 b4 and White wins a
pawn, also retaining positional
advantage. Unfortunately the game
was disfigured by errors from
both sides before 1 -0, 57.
Conclusion: l 3 . . . 0-0 14 lt::Jx h7
The standard theoretical re

commendation here is the neat
little co mb ina t ion 18 ... ll:lxf4 1 9
llxf4 ..tg6! when 20 lt::Jf6+? is bad
xh7 is a draw by force (asin A2).

The only possible way for the
defender to avoid this outcome,
against best play by White, is by
0Hff Otmblt: Stet/on 2

ln122
Moller A ttack 1 3 .
Wf61? in A3- and that

11 hardly an appealing option.
13
0-0 14 ltlxh7 f5 leads to
equality, though perhaps not a
completely dead draw. So, to be
4
s
6
7
8
9
10
11
12
...
...
strictly accurate, one might insist

that I ought to qualify the claim
made in the previous section that
the draw is demonstrable after
1 3 . .. 0-0. However, if Black is at
all interested in the full point this
is the merest quibble: 1 3 . . . h6 is
definitely a better choice than
1 3 . . 0-0. Of course, if White has
any ambitions, 1 3 . 0-0 is almost
as serious an objection to the
Moller Attack as 13 . . . h6.
c3
d4
cd
ltlc3
0-0
dS
llel
lhe4
g4!? (18)
..
0-0
ll:l f6
ed
b4 +
ll:l xe4
xc3
f6
ll:le7
d6
18
B
. .
Peel-back: To proceed with our

peel-back analysis, we can rapidly
strip away the two half- moves 1 2
. . . xg5 1 3 ltlxg5 , there being n o
respectable alternatives. Black did
once play 1 2 . . 0-0?! in a tour
nament game (Spielmann-Duras,
Carlsbad 1907), but it is not
'
surprising that he suffered for it
after 1 3 xf6 gf 14 d2 ltlg6 1 5
llae l . So w e go back t o White's
1 2th move, where there is an
alternative of sorts - the frenetic
1 2 g4! ?.
.
Section 3
Moller Attack:
1 2 g41?
l
2
3
e4
ll:lf3
J.c4
eS
ll:lc6
J.cS
General principles positively

bellow that this must be a rotten
move . Leaving his queen, queen's
bishop and queen's roo k yet to be
activated White makes a loosening
advance that weakens a whole
array of squares in his own king's
field. Such a procedure can only
be j ustified by special circum
stances - even if you were winning
a piece by force you would have to
think twice about it. Aside from
such generalities, however, I really
don' t know what to say about 1 2
g4 i n this concrete situation. There
is no published analysis on the
move that is worth anything, there
is no tournament experience to
learn from. The theoreticians
Moller Attack 12

mention 1 2 g4, attribute it to
Schlechter (though a master of
such sober s tyle surely did not
intend it for use in serious games),
tack on a short string of plausible
moves, and then give it the
thumbs-down because it is weaken
ing. It is indeed very difficult to
come up with any convincing
sequence that helpfully clarifies
the situation.
Two replies come into con
sideration:
g4!? 29
his black-squared bishop to good

use on the long diagonal. But it is
not apparent why White should
improve his chances by parting so
readily with the g-pawn.
/9
B
A 1 2 . .. 0-0
B 12
h6
...
0-0
12
Black proceeds on the reasonable

assumption that after 1 2 g4 there
is no need to bother about
hanging on to the extra pawn.
Simple developing moves should
work against a weakened structure.
13
14
15
gS
ll:lxeS
li xeS
i.eS
de
ll:l g6
suppose the ideal formation

for Black would be knight on d6,
bishop on f5 and queen on d7,
followed by rooks to the e-file.
H owever, . White only needs to
attack f5 twice (e.g. answering . . .
li:ld6 with i.d3) i n order t o upset
this scheme.
16
lie1 (1 9)
An old ( Tidskrift for Schack

1 9 1 3) but untested variation is
1 6 lile3!? 1!i'xg5+ 1 7 lig3 'it'e5 1 8
i.d2. White has hopes of putting
In Sovremenny Debyut Sozin

appended the continuation 16
't!Vd7 1 7 i.e3 b5 1 8 i.d3 (a trap:
1 8 . . . 't!fxd5?? 19 i.xg6 t!fxd l 20
liaxd l and 2 1 i.c5 wins) 1 8 . . .
't!fh3 1 9 'i!fc2 ll:lh4 and Black is
winning. These moves are hardly
best. White can obviously improve
with 1 8 .i.fl . More rational seems
16 ... t!fd6 (as given by Estrin). I
can only say that the resultant
position is unclear: White's position
is very loose, but on the other
hand his pieces are active and he
has two bishops on an open
board. The main headache for
White is that Black threatens to
open the f-file with . . . f6. For
exa mple, after 16 . . . t!fd6 17 1Wd4
f6 18 i.d2 ll:le5 19 .ie2 c5 Black
parries the threat of i.b4 and
starts to open up the white
king.
...
MOller A ttack 12 g4!?
H tJHH Otmblt: Stet/on 3

8
12
h6
A neglected alternative which

may be even stronger than I2 . . .
0-0.
f8
13
h4
14
h5 (20)
20
B
In The Italian Game Harding

and I gave a crushing White win
fro m this position: 1 4 ... g5 I 5
lbd4 c6 I 6 "itf3 lb xd5 I 7 .id2 lbc7
1 8 :ae i d5 I 9 i.b4+ ot>g7 20 lit e7 !
de 2 1 ltJxc6! 'i!Vd3 (2 1 . . . b e 22
Wxf6+!) 22 llxf7+ <t/xf7 23 lle7+
ot>g8 24 'i!Vxf6 llh7 25 lle8+ 1-0
Lazard-Gibaud, Paris 1 909. Very
pretty, but Black would surely do
better to play . . . c7-c6 a move
earlier. After 1 4 ... c6 he threatens
15 . . . cd 1 6 i.xd5 ltJ xd5 17 'it"xd5
i.e6 1 8 'it'xb7 'ti'c8 , when the
bishop pair and White's vulnerable
pawns at g4 and b2 give Black a
definite advantage . The pressure
against d5 and g4 after 14 . . . c6
makes it hard to find a reasonable
1 5th move for White - 1 5 g5 just
loses a second pawn to 1 5 . . . hg
16 lbl .txg5 lbxd5.
Conclusion: 1 2 g4!? has never

really been given serious consi
deration because, understandably,
nobody believes in it. Actually, I
would say that the variations
deriving from 12 g4 0-0 are not
obviously any worse for White
than the main line ( 1 2 .ig5 i.xg5
13 lb xg5 h6) - though that is not
much of a recommendation. In
any case, 12 . . . h6 seems to stop
White dead in his tracks, provided
that Black plays . . . c7-c6 quickly
enough.
In the next phase of our peel-back

we can retract the moves 1 1 llxe4
d6. . The former requires no
com ment , and although Black
might well play I I . . 0-0 instead of
I I . . . d6 this will simply transpose
into the next section , in which
Black plays I 0 .. . 0-0 instead of
10 . . . ltJe7.
.
Section 4
Moller Attack:
10
0-0
. . .
1
2
3
4
5
6
7
8
9
10
e4
ltjf3
.t c4
c3
d4
cd
ltJc3
0-0
d5
lle1
e5
lbc6
.ic5
lbf6
ed
.ib4+
lbxe4
.ixc3
.if6
0-0 (21)
Moller A ttack 10 . . . 0-0 3 1
better chances.
21
w
12
cd
Now comes a quite problematic

choice between:
A 1 3 't!Vxd6
B 13 i.. g S!?
A
13
14
11
lhe4
li:Je7
One can hardly approve of 1 1 ...

li:Ja5!?, putting the knight out of
play and leaving the kingside
short of defenders . Two game
examples : 12 i.d3 d6 and now:
a) 13 i.d2 c5 14 Ilf4 i.e7 1 5 'i!Va4
b 6 16 ll e 1 f5 17 g4 i.. d 7 (after 1 7 . . .
g5 1 8 llxf5 i.xf5 1 9 i.. xf5 White
has more than enough for the
exchange) 18 'i!Vc2 fg 19 i.xh7+
<i;>h8 20 't!Vg6 ! :i Eisinger
Degenhardt, Bad Aibling 1 965 .
b) 1 3 g4 c5 1 4 g5 (or 1 4 't!Ve2) 1 4 . . .
i.e7 1 5 i.d2 b6 1 6 'ire2 i.. f5 1 7
li e 1 c4 1 8 i.c2 ll e 8 1 9 i.a4 i.d7
20 i.xd7 'fxd7 (Black is almost
totally paralysed) 2 1 li:Jh4 <i;>f8 22
f3 <t>g8 23 e3 <i;>f8 24 i.c3 ltl b7
25 d4 f6 26 lle6 li:Jc5 27 gf i.xf6
28 ll xf6+ 1 -0 Thomas-Markwell,
corres 1 964-5 .
d6!
12
1 2 g 4 d 6 transposes into A of
section 3, Opening the diagonal for

the bishop on c4 and at the same
time blocking the development of
Black's queenside surely offers
'fxd6
'fd5 (22)
li:Jf5
22
B
A safe and easy way out for

Black here is 1 4 ... ltle7 1 5 'i!Vd6
(otherwise 1 5 . . . d5) 1 5 . . . ltlf5 and
draws by repetition. This option
indicates that, unless White can
come up with something special in
B ( 1 3 i.g5), 10 . . . 0-0 is not really a
bad move, but just less challenging
than 1 0 . . . ltl e7 . But can Black try
for more than the draw from this .
diagram? The test of this is:
14
15
d6
ltlg5!
A very dangerous tactical idea

which makes the draw option at
move 1 4 look the prudent course.
The variations are:
a) 1 5 ... i..x g5 16 i.. x g5 and now:
. 32 Grtco Gambit: Section 4

16
'irxg5?? 1 7 'ti'xf7+ l1xf7
1 8 :S:e2 mate.
a2) 16
i.e6 1 7 E:.xe6 ''xg5 1 8
E:.xd6 .
a3) 1 6
'ti'c7 1 7 1t'd3 i.d7 1 8 g4
h6 1 9 i.f4 (Keres ) .
b) Since ' a ' is unsatisfactory and f7
must be defended somehow this
only leaves 15
ll:\ h6, ugly as it is.
Estrin then suggests 16 h4 i.f5
without assessmen t, i.e. oo. Black
can perhaps hold out in the event
of 1 7 ll:\xf7 ! ? E:. xf7 (not 1 7 . . .
i.xe4?? 1 8 ll:\xh6++ 'it>h8 1 9 ''g8+
and mates) 1 8 i.xh6 i.xe4 1 9
1t'xf7+ 'it>h8. Stronger i s 1 6 ll:e l ! ,
which is good for White because
he threatens both ll:\ e4 and i.f4
and is ready to meet 1 6 . . . i.f5
with 1 7 ll:\ xf7.
al)
...
...
Moller A ttack 10 . . . 0-0

23
B
. . .
...
13
...
Conclusion: l O ... 0-0 is playable,

with a quick draw in prospect (in
line A). But nobody is going to
play it, since 10 . . . lb e7 is clearly
stronger.
i.g5 !?
An en terprising attempt to
avoid the possible draw in line A.
13
ll:\ g6
It might s eem that Black can
break out at once with 13
i.xg5
1 4 lL\ xg5 d5 1 5 i.xd5 lL\ xd5 1 6
'irxd 5 h6, but this provokes the
forcing line 1 7 lbxf7! E:.xf7 1 8
E:.ae l @f6 1 9 l:i:e7 'ifxf2+ (this is
necessary: after 19 . . . g5 20 E:.xf7
1t'xf7 21 E:.e8+ 'it>g7 22 'ife5+ 'it>g6
23 't!t'e4+ 'it>f6 24 1i'd4+ 'it>g6 25
\i'd6+ Black loses his queen) 20
'it>h l g5 2 1 E:.xf7 'ti'xf7 22 E:.e8+
'it>g7 23 't!t'e5+ 'it>g6 24 'ife4+ 'it>h5
25 @e2+ 'it>g6 , when White can
take a draw by perpetual check.
...
14
This is Estrin's recommendation.

However, Black counters with 14
ll:\e5. White is struggling to show
compensation for his gambit.
'ttd5 (23)
Peel-back proceeds by flashing

past 10 li:e l (which can be taken
for granted) to consider what
alternatives to 9 . . . i.f6 Black has.
The analysis so far makes it plain
that Black does not need alter
natives to 9 . . . i.f6. So prospective
defenders could skip the next
section. For gambiteers it may
provide a little light relief.
Section 5
Moller Attack: Black's
9th M ove Alternatives
1
2
3
e4
li:lf3
i.c4
e5
ll:\c6
.tc5
Moller A ttck - Black 9ths 33

4
5
6
7
8
9
c3
d4
cd
lbc3
0-0
d5 (24)
lbf6
ed
.ib4+
lb xe4
.txc3
u
8
longer, though also in White's

favour) 14 nxe5 lbf5 15 "ti'h5 + g6
1 6 ll xf5! - Unzicker.
These variations are rather
more illustrative than analytical,
but the basic message that Black
has a hard time keeping on the
board comes through loud and
clear. Rather more problematic,
though also considered ropey, is:
d) 9 .. .ia5 10 de, and
d l ) 10 . . de 1 1 "ti'a4 .tb6 1 2 .ixf7+
1Jxf1 13 "ti'xe4 with a strong attack.
Euwe gives the plausible follow-up
1 3 . . . 'i!fd5 14 lbg5+ ct18 1 5
lbxh7+ g8 1 6 lbf6+! gf 1 7 9g6+
1Jf8 1 8 't!fxf6+ 'it.'g8 19 't!fg6+ 'it>f8
20 .ih6+.
d2) 10 ... 0-0 1 1 't!td5 lbd6 12 .id3
.tb6 13 .i xh7+ ! 'it>xh7 14 Wh5+
'it.'g8 15 lbg5 lle8 1 6 .if4 and
Black is defenceless.
d3) 10 . . b e 1 1 lbe5 (an improve
ment on 1 1 Wa4 .tb6 12 .i xf7+
1Jxf1 13 't!txe4 d5 1 4 1!t'a4 lle8 +
Maroczy-Janowski, Carlsbad 1 907)
1 1 . . lbd6 1 2 't!fg4 't!ff6 ( 12 . . . 0-0?
1 3 .i.g5 and 14 .if6) 1 3 b4 (25)
d3 1) 13 ... 't!rxe5 14 .if4 't!ff6 1 5
nae l + 'it>f8 1 6 .ig5 't!rg6 1 7 .ie7+
1;g8 18 .ixd6 .ib7 1 9 1!t'xd7 h5 20
.tc5 was given in Schlechter's
Handbuch.
d32) 13 . .ixb4 14 .ib2 and it
seems the attack will win through.
For example, 14 . . . lbxc4 15 lb xc4
't!rg6 1 6 llfe l +! 'it.'f8 (or 16 . . .
.txe 1 1 7 nxe l + 'it>d8 1 8 .txg7) 1 7
1!t'h4 f6 1 8 lb e 5 fe 1 9 l:txe5
Euwe.
.
lbe5
In the end this proves inferior to

9 . . . .if6, but it should not be
dismissed too casually, if o nly
because it was once recommended
by Emanuel Lasker.
Weak are:
a) 9
lba5 10 .id3 lbc5 1 1 be
lbxd3 12 "ti'xd3 0-0 1 3 lbg5 f5 1 4
d6! c6 1 5 ll e 1 .
b) 9 .. lbe1 1 0 be ltJd6 (on 1 0 . . . 0-0
I I ll e 1 lb f6 1 2 d6l is regularly
given, though the consequences
are hardly very clear) 1 1 .ib3 0-0
1 2 c4 b6 1 3 .ib2 lbe8 1 4 .ic2 lbg6
15 "t!fd3 d6 16 lbd4 Moller.
c) 9 . lbd6 1 0 de ( 10 be is also
good) 10 . . .if6 (or 10 . . . lb xc4
1 1 'i!fe2+ 'i!fe7 12 Wxc4 .ia5 1 3
.ig5 ) 1 1 lle l + .ie7 ( 1 1 . . . <t>f8
12 cb .ixb7 1 3 .ib3 ) 12 .ig5 f6
1 3 lbe5 ! fe ( 1 3 . . . 'it.'f8 1 4 cd .ixd7
15 lbxd7+ 11rxd7 16 .ih4 resists
...
..
Moller A ttack - Black 9ths
34 Grtco Gambit: Section 5

13
B
26
w
1 3 b4 was one of Moller's

original ideas and you can see
from this var iation, and also 'b'
above, that after 9 d5 White would
like to put his queen's bishop on
the long diagonal. H ence the
virtue of 9 . . .i.f6.
.
10
11
be
'it'd4
lt:lxc4
f5
I foll o w the tradition of treating
this as the main line , although the

simpler and less compromising
11
0-0 1 2 1!t'xd4 ll:ld6 is pr obably
a better choice, e.g. :
a) 13 1!t'f4 lt:le8 1 4 d6!? cd 1 5 i.a3
b6 1 6 Iife 1 i.b7 17 lt:ld4 't!t'f6 1 8
1i'xf6 li:l xf6 1 9 lt:lf5 lUeS 20 ll:l xd6
Iixe l + 2 1 llxe 1 i.d5 of Kopylov
Levenfish, USSR Ch . l 949 is + or
...
=f.
b) 1 3 1!t'd3 ll:le8 1 4 c4 d6 15 .i.b2

f5 (26)
This is Mieses-Suchting, Vienna
1 908. Structurally the position is
identical to variation 'a' in the
note to Black's 9th, but her e Black
has an extra tempo ( . . . d7-d6) and
his knight is on a more relevant
square. After 1 6 litfe 1 White's
more active pieces offer fair
compensation for the pawn .

By contrast, 1 1
lLl cd6? is bad:
12 't!t'xg7 Wf6 13 "t!kxf6 lt:lxf6 1 4
Ii e l + lt:l fe4 ( 1 4 . . . f8 1 5 .i.h6+
<t>g8 16 Iie5! lt:l fe4 1 7 I::t e l f5 1 8
lie7 is catastrophic) 1 5 lt:ld2 f5
1 6 f3 0-0 1 7 fe lt:lxe4 1 8 lt:lxe4 fe 1 9
Iixe4 d 6 2 0 .i.h6 .
...
12
1!t'xc4
Definitely not t he miscombina

tion 1 2 i.g5? lt:lxg5 1 3 'tWxg7 I::t f8 !
1 4 ll:lxg5 't!t'f6 H Schlechter
Lasker, London 1 899 .
12
13
14
d6
0-0
lt:l d4
f3
lt:lc5
This is Lasker's recommenda
tion, covering the sensitive spot
at e6. If instead 14
lbf6, then
1 5 i.g5 h6 16 .ih4 (Schlechter
later tho ught that 16 .ixf6 followed
by doubling on the e-file was even
more convincing) 1 6 . . . g5 1 7 .i.f2
g7 1 8 llfe 1 Schlechter-Meitner,
London 1 899 .
...
15
lie1
Also good is 15 .ia3 b6 16 .ixc5

be 1 7 ltl c6 1!t'f6 1 8 llfe 1 i.d 7 1 9
I::t e 7 ll f7 20 liae 1 Karaklajic
Alexander, Belgrade 1 952.
Moller A ttack
h8
15
Black is under severe pressure

11 nd cannot, for example, contem
plate 1 5 . . . lle8? 1 6 i.. a 3 llxe l + I 7
Zlxc 1 lt:la6 1 8 'ii'b 5 Schlechter
M c:it ner, Vienna I 899.
16 i. a3
b6 (2 7)
Bladk 9ths 35
llae 1 ). 1 I . . . 0-0 may be less

precarious, but even here 12 'ii' xe4
li:ld6 I 3 't!t'd3 lLle8 I 4 c4 d6 I 5 i.b2
looks to be pleasant for White
(splendid bishop plus lead in
development to compensate for
the pawn).
Peel-back: Our examination of the
Moller Attack ends here, since we

now proceed to retract 9 d5. That
move has introduced us to a
wealth of sharp play and exciting
possibilities. But it is not quite
good enough against the sternest
defence.
Romanov-Kotkov, corres I 963-4
continued 1 7 lt:l c6 i.. a 6 1 8 'ii' d4
g5 I 9 i.xc5 de?! ( I 9 . be!?
usually the better recapture) 20
e5 i.. c4 2 I llad i . However, as
Estrin points out, Black should
have played 18 . . . 'ii'f6! I9 'ii' x f6
l:Ixf6 20 lle7 i.c4 with chances for
both sides.
For that reason one should
surely prefer the immediate 17 i.xc5
be I 8 lLlc6, when White has more
than enough for the pawn, e.g. I 8
. . . 1Wg5 I 9 lle7 i.b7 20 llae i 'ii' d 2
2 1 ll7e2! 'it'g5 22 lLla5 i.. c 8 23 f4
fol lowed by 24 1!rb5 :t.
. .
Conclusion: After 9
lt:le5 Lasker's
intended defence with I I . . . f5 and
1 4 . . . lt:lc5 gives White a strong
bind for his pawn , and it is not
difficult to find moves that intensify
the pressure ( i.a3xc5, lLl c6, :e7,
...
Sectio n 6
Stei n itz ' s 9 be and
1 0 .ia3
28
B
1
2
3
4
5
6
7
8
9
e4
e5
ll:lf3
i. c4
lLl c6
c3
d4
cd
lLl c3
0-0
be (28)
lLl f6
.ic5
ed
.i b4+
lLlxe4
.ixc3
Steinitz's 9 be and 10 .ta3

dS
9
10
.ta3
Nowadays everyone knows that
this is insufficient because Black
can counter with 10 . . . de 1 1 lle I
.ie6 ! 1 2 l'lxe4 't!rd5 followed by
. . . 0-0-0. I have found it interesting
to trace the path by which this
conclusion was reached. It is
surprising how long it took those
two great players Steinitz and
Lasker to get there !
de
10
Black is not obliged to accept

the sacrifice. Schlechter played
10
.te6 against Steinitz at
Hastings 1 895 and after I I .tb5
lbd6 12 .txc6+ be 1 3 e$ 0-0 14
lbxc6 "t!lf6 Steinitz blundered
a way the exchange with 15 ll b l
.if5 1 6 llb3? .id7 1 7 b4 .ta4
(eventually Y2-Y2. 49). Lasker was
prompted to write in the tourna
ment book: 'Black declines the
acceptance of the sacrifice with
doubtful judgment'.
...
11
lle1 (29)
From this position we will

examine two games played in the
1 896 Lasker-Steinitz world cham

pionship match.
In the very first game of the
match Lasker followed the analysis
that he had given when annotating
Steinitz-Schlechter:
11
rs 12 d2 f7 13 xe4 fe
1 4 ll xe4 "t!lf6 1 5 We2 .tf5 (not 1 5
. . . .ie6? 1 6 lle l ll he8 1 7 d5 : )
1 6 1Wxc4+.
Lasker's note in the Hastings
tournament book ended: ' 1 6 llf4
h5 1 7 1Wxc4+ 'it?g6 1 8 d5 e5 1 9
1Wxc7 llhe8 and although Black is
two pawns behind for the piece,
and may lose a third, his attack is
excellent'. It is difficult to imagine
that 16 1Wxc4+ immediately can be
anything more than a marginal
i mprovement .
1 6 . . . Wg6 1 7 lle3 llae8 1 8 llae l
lit xe3 19 llxe3 h5 20 h3 h4 2 1 d5
lbe5 22 't!rxc7 lbd3 23 1Wxb7 .tc8 ! .
This forces a simplification after
which two of the three pawns
White has for the piece are
doubled and isolated. White's
chances are gone.
24 't!rc6 "t!Vxc6 25 de lbf4 26 lle7
a6 27 c4 c;Pf6 28 lia7 lbd3 29 .te7+
c;Pe6 30 llc7 e5 3 1 .ib4 llg8 32
.ie7 g5 33 c5 f7 34 f3 lle8 35
>f2 llxe7 36 lil:xc8 c;Pd5 37 lla8
lbe5 38 c;Pe3 lb xc6+ 39 c;Pd2 a5 40
litf8 lle5 4 1 f4 gf 42 ll xf4 llh5 43
c;Pe3 lbe5 44 lla4 lbc4+ 45 c;P4
c;Pxc5 0- 1 .
I n spite of the one-sided char
acter of this first game of the match
Steinitz must have believed that
...

t here was a way for White to
prosper, for he repeated the first
eleven moves with his next White
i n the third game. I wonder what
i m provement he had in mind.
1 7 d5!? or 18 lig3+!? or some new
try at move 23? Lasker knocked
o u t all t he se possibilities. He had
discovered the definitive solution,
u l i ne in which Black runs no risks
nt all:
ll
12
13
14
15
lixe4
1We2
lbeS
lb xc6
Steinitz's 9 be an d 10 .tal 37
21 f4 .td5 22 g3 b7 23 h3 'it'b5 24
h2 llg6 25 't!tc2 f6 26 .ih4 .ic6
27 g4 'it'd5 28 'it'f2 h5 29 g5 fg 30
.txg5 h4! 3 1 llfl lig8 3 2 'it'd2 a5
33 a4 lie8 34 f5 lig8 ! (31)
.ie6 !
1Wd5
0-0-0
l:r.he8
1!Vxc6 (30)
A classic zugzwang.
35 l:r.e I 'it'xf5 36 l:r.e5 1!t'f3 37 d5
'it'g3+ 38 h I 1!t'xe5 39 de+ xc6
0- 1 .
Black's king is completely safe,

h e has no weaknesses and an extra
pawn. I f simplification occurs, the

bishops of opposite colour would
make a draw likely. Although
B l ack is not in the slightest danger
of losing, a win is no easy task. A
constructive plan for Black m ust
be based upon attack on the white
king - a plan that Lasker carries
out with marvellous skill.
1 6 lie l l:r.g8 ! 17 l:r.e5 b6 1 8 .tel
g5 ! 1 9 l:r.xg5 llxg5 20 .ixg5 llg8
Conclusion: O ne might be able to

come up with some slight improve
ments on Steinitz's play in this
game. But searching for them
would be a thankless task. For all
practical purposes Lasker refuted
9 be.
Peel-back: 8
.. .txc3 has served

Black so well in the preceding
sections that retracting it borders
on ingratitude . But we ought to
see what happens if Black takes
with the knight instead, if only
because White's subsequnt attack
ing possibilities provided the
original rationale for the Greco
Gambit.
.
8 . . lbxc3
Section 7
. . .
lb xc3
l
2
3
4
5
6
7
8
e4
e5
ll::l f3
ll::l c6
J.c4
c3
d4
cd
ll::l c 3
0-0
J.c5
ltJ f6
ed
-'.b4+
ll::l x e4
ll::l x c3 (32)
Black certainly ought to play

9 . . . d5! here, but since the analysis
of the apparently suicidal alter
native is very interesting (and has
not been done properly in the
past!) we shall sub-divide into:
A 9 . . . J.x c3
B 9
d5!
Weaker than either is 9 -'.e7?
1 0 d5 lba5 (or 1 0 . . . ltJb& l l d6! cd
1 2 ll::l g5 J.xg5 l 3 1!fd5 0-0 14 J.xg5
t) I I d6! with:
a) l l
J.xd6 12 lie ! + J.e7 1 3
-'.xf7+ '.txf7 14 \i'd5+ '.tf& 1 5
ltJg5 \i'e8 1 6 'it'xa5 .
b) ll
cd 1 2 -'. xf7+ '.txf7 1 3
't!fd5+ 'it>f& 14 ll::l g5 1!fe8 1 5 't!fxa5 .
...
...
...
...
A
9
J.xc3
We need to sub-divide again:

This was the only move that
Greco analysed. To modern eyes
it looks dangerously co-operative,
allowing White to recapture with
a tempo-gaining attack on the
bishop. Prospective Black defenders
who, like me, see no defect in 8 . . .
-'.xc3 could afford to bypass this
section . It is curious, however,
that in spite of three and a half
centuries of buffeting 8
ll::l xc3
has not been conclusively refuted.
Black could also play 8
0-0
(intending 9 ll::l xe4 d5), but then he
must reckon with 9 d5, transposing
into a line mentioned in section 5
after 9 . . . J.xc3 1 0 be ll::l e7 I I li e !
ltJf6 1 2 d6.
. . .
...
be
AI 1 0 1!t'b3
A2 10 J.a3
1 0 't!fb3 is the old move, indeed

Greco's move, whereas 10 .ta3 is
supposed to be a complete crush.
As we'll see, it isn't necessarily so . . .
AI
10
't!fb3
d5
This move, suggested by Bern

stein in the 1 930s, is Black's only
hope. Greco already showed how
to demolish the alternatives:
a) 10
.txal is incautious and
greedy and, not surprisingly, gets
blasted after I I -'.xf7+ '.tf& 1 2
.ig5 ll::l e7 1 3 ll::l e 5! (Steinitz's 1 3
liha l i s simple and strong, but
Greco's coup is more decorative).
...
Various pretty mates may now

ensue. For example, one of
G reco's composed gamelets rattles
o n 1 3 . . . d5 1 4 't!kf3 .tf5 1 5 .te6 g6
1 6 .th6+ 'it>e8 1 7 .tfl mate. 1 3 . . .
..t xd4 1 4 .tg6! d 5 1 5 't!kf3+ .tf5 1 6
..txf5 .txe5 1 7 .te6+ is another
< i reco variation. Of course , it
would be amazing if such things
were to occur in a modern tour
na ment game, but studying these
possibilities can help in sharpening
y o u r tactical teeth.
h ) 10
.txd4 1 1 .t xfl+ f8 1 2
.ig5 .tf6 1 3 li ae 1 li:Je7 1 4 .i.h5
h I ) 14
li:Jg6 15 li:Je5 ! li:Jxe5 16
li xe5 g6 17 .i.h6+ .tg7 18 lif5+ gf
19 'irfl mate - Greco.
h2) 14
dS 1 5 lixe7 't!kxe7 1 6 li e !
i.e6 ( but here 1 6 . . . i.xg5 1 7 li xe7
.txe7 1 8 't!i'xd5 g6 also fails to 19
e5 'it>e8 20 li:J xg6!) 17 li:Jd4
G reco again.
What we have here relegated to
a footnote was the 1 7th century
main line. If Black had to enter it,
the Greco Gambit would be amply
justified. But he doesn't and it isn't.
11
.txdS
0-0
12
.txf7+ (33)
...
...
...
. . .
li:Jxc3 39
It might appear that 12

li:xti
loses at once to 1 3 li:Jg5 , but for
what it is worth Black then has
1 3 . . . .i.e6 , meeting 14 'ixe6? with
14 . . . 'ird7 and 14 li:Jxe6 with 14 . . .
li:Jxd4. H owever, i t isn't worth
much since 1 4 't!t'xc3 ! lif6 1 5 li:Jxe6
lixe6 1 6 'irc4 leaves Black in
serious trouble - the threat of
d4-d5 is terrible. On the other
hand, 12
\t>h8 1 3 1!fxc3 lixfl 1 4
1/tb3 ! i s also very pleasant for
White, e.g. 1 4 . . . lif5 1 5 li e 1 .td7
16 d5 li:Ja5 ( 1 6 . . . li:Je7? 17 i.g5
Euwe-van den Kar, Amsterdam
1 928) 1 7 ti'c3 - Euwe.
White even has another route to
advantage in AI via 1 1 'it'x c3 de
12 dS! li:J e7 1 3 i.a3 0-0 14 li fel as
in Rajaram-Averbakh, Calcutta
1 965. After 14 . . . li:Jxd5 1 5 1!t'xc4 c6
16 .txf8 'ffx f8 1 7 liab 1 Black's
two pawns do not adequately
compensate for the exchange
because of the activity of the white
rooks.
...
...
A2
10
.t a3 (34)
This is universally adorned with

a ' ! ', and -is often used as a teKt-
8 . . lilxc3

book example of how a lead in
development can lead to a decisive
attack. Aitken first suggested the
idea in the British Chess Magazine
in 1 937 and he supplied a very
attractive variation.
However, I cannot agree that it
is really superior to 10 1!t'b3 .
Before seeing why, let u s deal with
the defensive attempts that have
made 10 .ta3 see m so appealing:
a) 10 d5 1 1 .ibS .t xa l l 2 l:!. e l +
.te6 1 3 "i!Va4 1!t'b8 (horrid, but
even more so is 13 . . . li b8 14 ltleS
Corte-Bolbochan, Parana 1 946)
1 4 li:leS a6 I S "i!Vb4! \!i'd8 and now
best is 16 .txc6+ be 1 7 1!t'b7 (Wade).
b) 10
li:le7 I I 't!fb3 dS 1 2 't!t'xc3
de 13 life l .ie6 1 4 .i xe7 Wxe7
and now both I S dS 1!t'xdS 1 6
l:!.ad l and I S li:lgS h6 1 6 li:lxe6 fe
1 7 dS are winning.
c) 10
d6 1 1 lic l .taS 12 't!i'a4 a6
(otherwise he drops a piece to dS)
13 .idS .ib6 1 4 li xc6 ! .td7 l S
l:!.e l + (or I S li xd6 when I S . . . cd
1 6 l:!.e l + is a transposition and
1S . .txa4 1 6 l:!.e l + "it'e7 1 7
l:!.xe7+ xe7 1 8 li xb6+ i s )
I S . . . Wf8 1 6 lixd6 ! (35)
...
...
...
. .
This is the main line of Aitken's

analysis. A fter 16 . cd 17 .txd6+
g8 White can probably close the
show in various ways. But simplest
is best, so I like the unfussy 1 8 "it'b3!
( 1 8 . . . .te8 1 9 lixe8+ 1rxe8 20
ltleS; 1 8 . . . .ie6 19 lixe6). Aitken,
going for an analytical beauty
prize, thought up 1 8 li:lgS g6 1 9
.txf7+ Wg7 20 .ie5+ h6 2 1 li:le6
.txa4 22 .ig7+ and White delivers
mate in six, which is indeed pretty
but is longer and allows more
distractions (e.g. 21 . . . .txe6 22
.txe6 should win, but is not exactly
crystal clear).
Variations a-c are all you get in
most manuals, which regularly
and culpably ignore the best
defence, proposed by Bernstein:
. .
10
'ti'f6!?
Keres now gives 1 1 l:!. c l .ib4 1 2

.txb4 li:l xb4 1 3 li e l +, claiming
advantage for White - a sequence
and assessment repeated by Euwe.
White undoubtedly has good com
pensation for the two pawns
sacrificed, but after 13 . .. d8
there is nothing particularly clear.
A plausible continuation is 14 'ti'd2
aS (not 14 . . . li:lc6 1 5 li:lgS, nor 14 . . .
d S I S 'ti'xb4 dc l 6 't!i'xc4) 1 S a3 dS
Black's position remains highly
precarious and I would not like to
have to defend it. But it is
questionable whether White has
i mproved on the solid he gets in
AI.
B
9
d5!

A sensible move, after which
White has more trouble proving
his case.
10
11
cb
llel+
de
li:J e7 (36)
36
w
. . .
(}JxcJ 41
a) Everyone says that 13

fg 1 4
"t!t'xc4 i s too risky for Black. T o be
honest, I am not at all sure about
this, but it would be foolhardy to
test the verdict in a serious game.
b) 13 o-o 14 'it'xe7 fg 15 1Wc5 1Wf6
worked in Schwarz-Teschner, Ber
lin 1 949 : 1 6 li:Jxg5 'it'xf2+ 1 7 h l
.i.g4 1 8 1Wxc4+ h8 1 9 lUI .te2!
20 lt:lf7+ g8 21 lt:lh6++ <t>h8 22
lt:lf7+ and draws. Euwe, however,
has a significant improvement on
this in 16 dS ! , defending f2 along
the diagonal and ready to meet
. . . g4 or . . . .i.g4 with li:Jf3-e5. White
has a definite advantage then: the
black bishop is awkward to develop
and the pawns at c7 and g5, as well
as the 'spare' at c4, are vulnerable.
...
...
12
.tgS !?
don't think that White can

extract more than a draw from
12 'fe2 J.e6 1 3 .i.g5 ( 1 3 li:J g5!?
might be tried) 1 3 ... 1Wd5 1 4 .txe7
xe7 1 5 1!t'c2 f6 1 6 li:Jg5 ! (Bogolju
bow's idea) 16 . . . fg 17 l:t e5 'it'xd4
18 llae 1 llae8 1 9 l:txe6+ <t>d7 20
l:i:d1 (not 20 lii: xe8? llxe8 2 1 ll d l
c8 ! ) 2 0 . . . 'ffx d l + 2 1 "it'xd l <t>xe6
22 "t!fg4+ f6 23 h4 etc.
12
13
.tf4 (37)
37
B
f6
"t!fe2!?
White offers a piece. The

quieter 13 .tf4 0-0 14 'ife2 li:Jd5 1 5
1!t'xc4 h8 (Leonhardt-Perlis, 1 908)
causes no problems and is probably
+ in view of Black's better pawn
structure.
13
14
i. g4
A grey move - neither White

nor Black will be fully satisfied
with the outcome. But it is hard to
find better:
Can Black slip out of t he pin

and establish his knight on d5?
a) 14
'it'd 7 15 'it'xc4 .txf3 16 gf
c6 (after 16 0-0-0 1 7 .txc7 Wxc7
1 8 'ffx c7+ xc7 19 :xe7+ Black
loses a pawn, but whether White
wins the ending after 19 . . ll d7 is
another matter) 1 7 :litac l ! <t>fB 1 8
b5 cb 1 9 9b4 aS 20 'tlrd6 l:[d8 2 1
..
. . .
42 Greco Gambit. Section 7
8 . . . lbxc3
llc7! Estrin-Klaman, Leningrad

195 1 . The power of the doubled
rooks on the seventh is decisive
after 2 1 . . . 'i!xd6 22 .i.xd6 llxd6 23
llexe7.
b) 1 4
'it>f7 15 'ifxc4+ li:Jd5 1 6
lld2! (Sakharov's move; i t i s far
better than 1 6 .i.xc7? Ie c8 and
White is losing a piece - 1 7 .i.xd8
l::t x c4 18 i.a5 b6 etc) 1 6 . . . .i.e6 1 7
.ig3 l::t e8 1 8 lt:Je4. You will find
these moves in lots of books, with
the conclusion that White has the
better chances. So far as I know,
this has not been tested in practice.
The pressure on the c-file plus the
irritating threat of li:Jc5 probably
does count for more than the
strong knight on d5.
analysis of the Greco Gambit. For

the next peel-back will remove
7 . . . lt:Jxe4 8 0-0 (both clearly best) ,
thus taking us back to consider
alternatives to 7 li:Jc3.
...
Conclusion: 8 ... ltlxc3 is just about

playable so long as Black continues
(after 9 be) with 9 . . . d5 ! . In that
case it seems that, with accurate
play, White can preserve just a
slight advantage (line 'b' from the
last diagram). 9 . . . i.xc3? is really
just an old error from the glory
days of the Greco Gambit when
defences were feeble. It does seem
to me that Greco's original rejoin
der 1 0 Wb3 has been unfairly
slighted. It is just as effective as the
20th century's 1 0 i.a3 . However,
White players nowadays are not
likely to have the pleasure of
testing the relative merits of these
two moves.
This
section
concludes
our
Section 8
H a s Anyone Got Any
Better I deas 1
We have seen that the Greco
Gambit proper with 7 ltl c3 does
not work out well for White. He
can no longer indulge in the
Moller Attack with the comfort of
the drawing line of section 2 in
reserve - our main line with 1 3 . . .
h 6 (section 1 ) i s the mausoleum of
that hope. But perhaps one can
play in the spirit of Greco without
sticking to the letter of his pioneer
ing quill. The general objective is
to play an early d4 in the I talian
Game, to get an open position
with early contact in which White
stands ready to regard the survival
of one of his pawns as a mere
trifle. Let us see if there is any
promising way of doing this from
the position after
1
2
3
4
5
e4
e5
ltlc6
li:Jf3
.i.c4
.i c5
c3
d4
ltlf6
ed (38)
We need to look at White's 7th

move alternatives after 6 cd .i.b4+ ,
but there are a couple of other 6th
move options too. So let's do
some fission:
A ny Better Ideas? 43

38
w
A 6 e5
8 6 0-0
C 6 cd
the aid of . . . .tg4, whereas it is not

easy to suggest an active plan for
White. Some examples:
a) 9 lt:l c3 0-0 1 0 .te3 and Black can
happily choose between
a 1 ) 10 . . . f6 1 1 ef lb xc3 1 2 be 't!Vxf6
1 3 't!Vb3 lbe7 1 4 0-0 h6 1 5 lbe5 c6
1 6 .te2 lbf5 + Steinitz-Schiffers,
Vienna 1 89 8 .
a2) 1 0 . . . .tg4 l l h3 i.h5 1 2 't!Vb3
.txf3 13 gf lt:lg5 =t= Stanciu-Urzica,
Rumanian Ch 1 974.
b) 9 i. e3 0-0 10 1Wc2 lt:le7 1 1 lt:lc3
ll:lxc3 12 be i.f5 1 3 i.d3 is the .
line given in ECO.
This is a grey variation, hardly
in the Greco spirit.
=
e5
White does not lose a pawn in

this line, but he does lose control
of d5 and must beware of Black's
potential f6 break.
d5
After 6 . . lt:le4? 7 i.d5 Black is

forced to sacrifice, insufficiently,
with 7 . . . lbxf2 8 <ot>xf2 de+ 9 <ot>g3
cb 1 0 i.xb2 . White's develop
ment is too good for that to work .
lt:le4
7 .tb5
.
cd
Compare this with the Two

Knights Defence with 5 e5 (Scotch
Gambit, section 3). 8 lt:lxd4 is a
transposition of sorts in which
White has substituted c3 for the
more useful i.e3 : 8 . . . i.d7 9 .txc6
be 10 0-0 f6! + Boutteville-Larsen,
Le Havre 1 966.
8
i.b6
Black has, if anything, the

better prospects. He can arrange
. . . f6 or think of attack on d4 with
0-0!?
Here is something more worthy

of the gambiteer's favours. Unfor
tunately, there is one line ( Black's
best ! ) in which the queens are
soon exchanged, leaving too little
muscle for the attack to carry
clout.
6
lt:lxe4!
Other m oves have poor creden
tials :
a) 6 ... de 7 lbxc3 is a Scotch/Goring
Gambit that Black has already
misplayed.
b) 6 d6 7 cd .tb6 gives White his
ideal centre, and yet it cannot be
said that the position is entirely
clear after 8 lbc3!? .tg4!?.
c) 6 . . . d3 7 b4 .tb6 8 a4 a6 9 e5 d5
10 .txd3 lb e4 1 1 't!Vc2 .tf5 1 2 .ta3
Estrin-Korelov, Leningrad 1 970.
Estrin says that 7 . . . i.e7 is better,
..
44
Greco
Gambit: Section 8
although it see ms that White can

also co unt on a certain advan tage
after 7 . . . i.e7 8 e5 lt:\g4 9 l:l e l d6
10 ed cd I I i.x d3 ::!; Estrin Bily ap,
Varna 1 969.
d) 6 ... d5 7 ed lLixd5 8 l:le l + ..ie6
9 lt:\g5 0-0 lO 't!ld3 g6 I I Il xe6! fe
1 2 1!fh3 'ii'e 7 1 3 't!lxe6+ 'ti'xe6 1 4
lt:\xe6 Estrin Let ic, corres 1967-9.
-
cd
d5!
At first sight 7 . ..ie7 looks a

safe move, but th e ga mbiteer gets
just the sort of position he wants
after 8 d5 lt:lb8 9 lil e l and now:
a) 9 . lt:lf6 10 d 6 ! cd 1 1 c3 and
Black is unlikel y to survive the
savage attack, th e natural ! ! . . . 0-0
running into 1 2 lil: xe7! 'ii'x e7 1 3
.i.g5 .
b) 9 d6 lO i. d3 0-0 1 1 c3 (39)
.
39
B
...
An ideal posit ion for Italian

gambit-play. Wh i t e s central pawn
pushes have given his pieces full
freedom of action , whereas Black's
cluttered forces are unable to
intercede on thei r king's behalf:
a6 1 2 .t f4 .i.f6 1 3 l:l e3
bl) ll
.i.xd6
cd 1 5 .i.c2 b6 1 6
lt:\c5 1 4
9d2 .i.b 1 1 7 lilae 1 .i.xc3 1 8 be f5
19 ltld4 g6 20 J.xf5! gf 2 1 lile7 llf6
'
...
Any Better Ideas?

22 'ii'g 5+ lilg6 23 'ti'h5 llg7 24
xf5 1 -0 Corden-Perkins, British
Ch 1 969.
b2) 1 1 . lt:\e8 1 2 'ife2 (quite
possibly 12 d6!? xd6 1 3 i.xh7+
xh7 14 ll xe7 't!lxe7 15 d5 'ti'd8
16 g5+ might force a win - but I
will let someone else prove it, if it
does ! ) 12 . . . i.f6 1 3 lt:\e4 Estrin
Konovaltsev, Moscow 1 969 .
..
de
8
9 't!lxd8+
10
ll:dl+
10 . 'it>e8 1 1 ll e l
de
xd8
.i.d 7
f5 12 lL!c3
recovers the pawn with the better
game, for 12 . . . b4? fails to 1 3
xe4 c2 1 4 d6++! .
11
.i.e3 (40)
.
40
B
Does White have compensation

for his pawn? In a number of
games Black played the natural
1 1 . . 'it>e7, stepping out of the pin,
and after 1 2 a3:
a) 1 2 ... e3 13 be lt:lxc3 14 Ji[d3 lt:la4
1 5 lile 1 J.e6
a 1 ) 1 6 lt:\b5? J.c4 1 7 .i.g5+ 'it>f8 H
Honfi-Sax, Budapest 1 970.
a2) 16 lDd4 lbxd4 1 7 .i.xd4 is quite
good for White according to Estrin
( 1 7 . . . lithd8 1 8 b5 c6 1 9 lDd6).
.
Any Better Ideas? 45
b) 1 2 ... i. e6
b l ) 13 lilb5 Ithd8 1 4 li:Jxc7 nxd l +
1 5 lixd l It d8 1 6 lixd8 lil xd8 +
Ravinsky.
b2) Estrin suggests 13 lidc1 , though
it hardly looks more than
However, in Kopylov-Jezek,
Potter Memorial corres 1 974-7,
Black innovated with 11 . lie8!?
and after 1 2 li:Ja3 c3 13 be li:J xc3
14 lld3 lile4 1 5 llad l If.e7 White's
16 lild4? lile5 just lost a second
pawn. Messere suggests that 1 6
lil e l intending f3 would have been
' more prudent' , but things have
already gone wrong when the
gambiteer has to grovel like that.
Black's rook holds everything
when it gets to e7 and the king on
d8 defends c7. The gambit pawn
starts to look rather more than a
mere trifle !
=.
c
i.b4+
6
cd
Everybody knows that 6 ... i.b6?
is bad but you need to understand
the reason why: 7 d5! lile7 (7 . . .
lilb8 8 e 5 li:Jg4 9 0-0 d 6 1 0 e6! , or
7 . lila5 8 i.d3 c5 - to stop b4
9 d6 ) 8 e5 lilg4 9 d6! (41)
..
4/
B
a) 9 . .. cd 1 0 ed
a l ) 10 ... ..txf2+ I I e2 lilg6 1 2 h3
lil4e5 13 lil xe5 lilxe5 14 'it>xf2
lilxc4 1 5 't!Ve2+ : .
a2) 1 0 . . li:Jg6 1 1 0-0 0-0 1 2 ..ig5
lilf6 1 3 lilc3 h6 1 4 't!fd3 ! 'it>h7 1 5
i.xf7 ll xf7 1 6 lil e 5 : .
b) 9
lilg6 1 0 i.g5 ! f6 1 1 e f g f 1 2
't!Vd5 li:J4e5 1 3 li:Jxe5 li:Jxe5 14 ..th6
intending i.g7xh8 .
c) 9 . . . lil f5 1 0 i.g5
c l ) 10 . . ..txf2+ 1 1 e2 li:Jd4+ 1 2
fl ! and White wins o n material.
(Check it out ! )
:2) 1 0 .. f6 I I 't!id5 li:J fh 6 1 2 ef
lilxf6 13 't!ke5+ ct>f8 14 i.xh6 gh 1 5
0-0 cd 1 6 't!t'f4 g7 1 5 lilh4 .
d ) 9 .. lil x f2 1 0 't!fb3 0-0 1 1 .i g5
.
...
These variations were put to

gether by Zak in 1 936. They would
not be easy to find at the board.
Now the reader is primed. But to
get the chance to play them? You
should be so lucky !
7 'it>fl!? (42)
This introduces the Krakow
Variation, which takes its name
from some analyses published in
1 909 by members of the Krakow
chess club (though in fact it had
been known and played long
before). Even if Black does not
take the pawn on offer at e4, we
can treat it as an honorary gambit.
To renounce castling rights like
this ought to be reckless enough to
qualify.
There is a respectable move that
does not make a genuine gambit
Any Better Ideas?

out of the opening, namely 7 i.d2.
It used to be possible to recommend
this as a viable way of playing (as
Harding
and I did in The Italian
.
Game), but in the light of the
following game it emerges that
White must then be prepared to
accept an early draw: 7 . . . i.xd2+
8 lLlbxd2 d5 9 ed lLlxd5 10 'it'b3
lLla5 ! (previously 1 0 . . . li:Jce7 1 1
0-0 0-0 1 2 Ilfe l or 1 2 li:Je5 led to
positions that were either ;!; or
and which White could at any rate
keep working on) 1 1 1i'a4+ lLlc6
(Black's point would now be
revealed by 1 2 li:Je5?! 0-0 13 li:J xc6
1!re8+! +. Meanwhile, there is ...
lLlb6 to worry about , e.g. after
1 2 0-0 0-0. So:) 1 2 "ti'b3 li:Ja5 1 3
1Wa4+ li:Jc6 \-1 - \-1 Miles-Korchnoi,
South Africa 1 979.
1 0 i.g5 i.e7 1 1 d6! cd 1 2 li:Jc3 0-0

1 3 Il e 1 li:Jc6 1 4 'ti'd2? lLle5 ! and
0- 1 , 25. But 14 'it'h4! gives a
tremendous attack, e.g. 14 . . . h6
1 5 Ilxe7 li:Jxe7 1 6 i.xf6 gf 1 7
'it'xh6 with the end i n sight) 9 1Wd4
li:Jf6 1 0 i.g5 (43)
43
B
=,
42
B
d5!
Spoilsport! White would enjoy

the game much more if Black
would only take the pawn: 7 . ..
lLlxe4 8 d5 li:Je7 (Znosko-Borovsky
thought 8 . . . li:Jb8 was better and
played it against Schwarzmann,
Paris 1927: 8 . . . lLlb8 9 1i'd4 li:Jf6
White has intense pressure and

Black's chances of prolonged
survival are slim. Some examples:
a) 1 0 . . . c5 1 1 1i'e3 d6 12 i.xf6 gf 1 3
li:Jbd2 0-0 1 4 li:Je4 lLl g6 1 5 'it'h6
Ile8 16 lLlg3 ct>h8 1 7 h4 Vatcek.
b) 1 0 . . . lLlg6 1 1 lLl bd2
b 1 ) 11 . . . h6 was Marshall-Burn,
Ostend 1 905, a brilliancy prize
game: 1 2 lite l + ct>f8 1 3 i.d3 ! i.e7
(better 1 3 . . . i.xd2 14 i.xd2 d6
Marshall) 1 4 i.xg6 hg ( 1 4 . . . fg 1 5
li:Je5 'ti'e8 1 6 'it'd3 ) 1 5 li:Je5 ! fg
1 6 li:Jxg6+ 'llf7 1 7 Il xe7+ <t>xg6 1 8
'ti'd3+ 'll h 6 1 9 h4! g4 ( 1 9 . . . 1Wxe7
20 hg++ 'l;xg5 2 1 lLlf3+ <t>g4 22
'it'g6+ <t>f4 23 g3+ and mate next
move) 20 h5 li:Jxh5 21 'it'f5 1 -0. If .
2 1 . . . g6, 22 Ilxh5+ gh 23 "t!i'f6
mate . Remembering the famous
story of Marshall-Burn, Paris 1 900
- 1 -0 so swiftly that Amos never
got to light his pipe - one wonders

how many puffs he enjoyed in this
game !
b2) 1 1 . j_e7 1 2 ll e 1 0-0 1 3 h4 d6
1 4 h5 li)e5 1 5 h6 li)xc4 1 6 li) xc4
lii: e 8 1 7 hg .if5 ( 1 7 . . . wxg7 1 8
llxe7 ! 1Vxe7 1 9 .ixf6+ 'ft'xf6 20
lii: h7+) 1 8 lt:l h4 .ie4 19 li)d2 .i c2
20 li) hf3 wxg7 2 1 .ih6+ g8 22
li)e4 .ixe4 23 ll xe4 .if8 24 ll f4
.ixh6 25 ll xh6 : Bartmanski
Broch, corres 1 9 1 0.
..
8
9
ed
lt:lc3
li:lxdS
.ie6 (44)
44
w
If this is the best the Krakow

Variation has to offer against
sensible defence, is there any hope
that it can supply the life-giving
injection that Italian gambit-play
needs? The most striking feature
of the position is that, whereas
Black is achieving a normal
development, the same cannot be
said for White, thanks to 7 Wfl .
Possible continuations are:
a) 10 .ixdS .i xd5 1 1 1Ve2+ .ie7 1 2
"iWb5 ( 1 2 li) xd5 'ft'xd5 l 3 .ig5 does
not stop Black from castling) 1 2 . . .
.ixf3 1 3 gf 'i!Vxd4 + Lenz-Schlage,
A ny Better Ideas? 47
1937.
b) 1 0 'it'b3 lt:la5 1 1 'it'a4+ .id7 12
.ib5 li) xc3 13 bc ( 1 3 .ixd7+ "t!txd7
14 'it'xb4 "t!tb5+) 1 3 . . . .ixc3 1 4
lil b 1 c 6 1 5 'it'c2 cb 1 6 'it'xc3 0-0 +
Levenfish.
c) 10 'it'e2
c l ) 1 0 .ixc3 1 1 be li)xc3 1 2 1!t'e l
li)d5 1 3 ..ta3 a6 ( 1 3 . . . "iWd7 1 4
.ib5 ! ) 1 4 lil c 1 "@'d7 1 5 'ire2 d8 (a
strange move , but at least it is
better than 1 5 . . 0-0-0? 16 j.xa6)
16 lt:le5 lt:lxe5 1 7 de li)f4 1 8 'it'f3
.ixc4+ 1 9 lhc4 li)e6. Bartmanski
Batik, corres 1 9 10, continued 20
e2!? We8 2 1 lld l 'it'b5 +.
However, the suggested improve
ment 20 g3 !? (Scacco! 1982) may
make the position playable for
White, in view of the awkward
spot the black king is in.
c2) 1 0 ... 0-0 looks sensible, but is
it good for Black? White can reply
1 1 .ig5 with the idea that 1 1 . . .
li)xc3 1 2 b e ..txc4 1 3 'it'xc4 j.e7
14 h4 is t. So 1 1 .ig5 1i'd7 and I
think Black is better.
...
I find it difficult to feel very

enthusiastic about the Krakow
Variation, but this may just reflect
a conventional p rejudice. The
correct assessment of the 1 0 "t!te2
line (variation 'c' from our last
diagram) is in doubt. It would
bear further investigation. There
is nothing else in this section that
White has any hope of making
much of against an adequately
prepared defender.
2
1
2
3
E VA NS
GA MBIT
e4
lt:Jf3
.tc4
b4 (45)
45
8
e5
lt:Jc6
.tcs
Captain William Davies Evans ( 1 790- 1 8 72) invented this ga mbit about
the year 1 824 whilst skipperi ng a postal steam packet in service between
Milford Haven and Waterford. Its first public appearance was a brilliant
success against Alexander McDonnell (given on p . 5 6), played in the
Subscription Rooms, St Martin's Lane, London in either 1 826 or 1 827
(exact date unknown). I t rapidly achieved popularity and was acclaimed
as 'a gift from the gods to a languishing chess world'. The gambit
appealed as a way of rej uvenating the open Italian style of play, which
had been in danger of becoming extinct through the influence of
Philidor's more positional, pawn-structural methods. The Evans was a
favourite of most of the great attackers of the 1 9th century - e.g.
McDonnell, La Bourdonnais, Morphy, Anderssen and Chigorin - and
was regularly played at the highest level until the turn of the century.
However, those staunch masters of defence, Steinitz and Lasker,
even tually turned the tide of opinion against the gambit, and its
appearances in 20th century tournaments have been sporadic.
The main area of debate is acceptance with 4
.ixb4 5 c3 .i a 5
(sections 1 -7). The big theoretical question is: should Whife choose 6 0-0
..
49
Evans Gambit: Introduction
or 6 d4 ? (The rare alternative 6 't!Vb3 is the subject of section 7.)

Chigorin used to favour 6 0-0 , but the usual verdict of theory has been
that it is not so good as 6 d4 because it runs up against the defence
recommended by Lasker in Common Sense in Chess (6 . . . d6 and 7 .
.ib6, offering to return the pawn for an ending in which Black has the
bishop pair and sounder pawns). Yet if you look at the end result of the
Lasker Defence (section 1), you will see that it is not at all clear who
stands better. In fact White seems to have rather a good plan of
campaign (in volving h I and f4-f5 ) , so that I am tempted to recommend
this as the appropriate contemporary handling of the Evans Gambit.
However, I discovered a coun ter-plan for Black ( 1 7 . . f5 !?). All I can say
i s that further investigation and practical tests are needed to resolve the
relative chances.
6 d4 enables White to sidestep Lasker's Defence. But in so doing it
encounters other defensive lines not available against 6 0-0. In particular
6 . ed 7 0-0 tt:lge7 (section 4) is a sound and reliable response. Black
achieves the advance . . . d5 and thus gets freedom for his pieces.
So which is the better option 6 0-0 or 6 d4? In the words of some
anonymous author of graffiti, ' I used to think I was indecisive . . . but
now I am not so sure' . What I am confident of is that the Evans is a fully
viable gambit that still poses open questions. It should be noted that in
practice White's results are rather good. Even against strong defence
White should retain fair compensation, one of the chief reasons for t his
being that the pawn offered is a wing pawn , the absence of which is not
such a telling factor in the middlegame.
No introduction to the Evans would be complete without the score of
that most famous of games, the ' Evergreen':
.
An de rssen-Dufresne
B erlin 1 85 1
1
e4
lt:l f3
3
4
5
6
7
8
9
10
.ic4
b4
c3
d4
0-0
't!V b3
e5
lit e1
e5
tt:lc6
.i c5
.ixb4
.ia5
ed
d3?!
't!Vf6
\!Vg6
50 Evans Gambit: Introduction
1 0 lld l intending .txd3 is a good alternative.

10
11
12
13
14
15
.ta3
't!Yxb5
't!Ya4
lbbd2
tbe4
tbge7
b5!?
llb8
.ib6
.tb7
t!rf5?
Safer was 1 5 . . . d2 1 6 tbexd2 0-0 - Lasker.
16
17
.ixd3
tbf6+!?
t!rh5
Simply 17 tbg3 ! was strong.

17
18
19
ef
llad1 !?
gf
llg8
Although this leads to a brilliant finale it has been suggested that

19 .ie4 is really a better move .
19
'fi'xf3 (46)
1 9 . . . llg4! would have prevented the beautiful mating finish by
ma king g8 available to the black king.
46
w
20
21
22
23
24
tbxe7
llxe7+ !
xd7
't!Yxd7+ !
We8
.if5+
d8
.td7+
.txe7 mate
Pleasing as this game is, it i s not of much use as a guide to how the
gambit is li kely to fare i n mod er n practice. So I offer a more recent
example of successful gambiteering.
Evans Gambit: In troduction 51
Wedberg-Kaiszauri
Stockholm 1 981
1
2
3
4
5
6
7
e5
lLlc6
i.c5
i.xb4
i.a5
e4
lLlf3
i.c4
b4
c3
d4
't!t'b3
d6
lLlxd4
7 . . . "@'d7 is the main line of section 6.
8
9
lLlxd4
i.xf7+
ed
f8
de
0-0
10
For 10 . . . "@'e7 see note 'c' to Black's 7th in section 6.
11
12
e5
i.xg8?!
't!t'e7
Wedberg (whose notes in Informator 33 I am drawing on) pointed out

that he missed the chance of 12 ed! 't!t'xf7 13 1i'a3 .
12
13
ed
Ilxg8
"@e5!
Black goes for cou nterplay. 1 3

"@xd6 was dubious because of
14 i.a3 c5 1 5 Il d 1 and if 1 5 . . . 1!t'b6? 1 6 Ild8+ "ii'x d8 1 7 i.xc5+ .
13 . . . cd 14 lLl xc3 is also promising for White - the extra pawn at d6 is a
weakling.
14
15
i.e6
lLl a3
1i'xb1 (4 7)
47
8
IS
i.dS?
An aggressive move, but B lack's position proves to be too loose.
52 Evans Gambit: Introduction

Wedberg gives 1 5 . . . 'iPf7! 1 6 '4!t'f3+ '4!t'f5 1 7 '4!t'xf5+ i.xf5 1 8 llJc4 i.b6
19 i.f4 =
.
16
17
18
'ii'b 5
etJc4!
lDe3
i.b6
1!t'e4
Since it would be suicide for Black to open the f-file to a discovered

check this holds everything together for White. M eanwhile Black's game
falls apart as he lacks defensive cover on the a3-f8 and b 3-g8 diagonals.
i.c6
cd
Against 19 . . . g5 White has 20 i.a3 g7 2 1 l:I ac l i.d4 22 :ii xc3 ! .
g6
20
i.a3
21 :ii ad1
.,Pg7
'iPh6
22 'ii'x c3+
h5
23 i.cl
24 etJd5!
1 -0
18
19
'ii'b3
There is a neat mate after 24 . . . i.xd5 25 "t!t'h3+ '4!t'h4 26 g4.

Sectio n
6 0-0:
lasker's Defence
1
48
w
2
3
4
5
6
7
e4
lDf3
i.c4
b4
c3
0-0
d4
e5
lDc6
i.c5
.i.xb4
i.a5
d6
i.b6 (48)
Advocated by Lasker in his

lectures of 1 895 (from which
stemmed the book Common Sense
in Chess) , this sequence is almost
universally regarded as a reason
for preferring 6 d4 to 6 0-0 . As we
shall see, it is not quite as simple as
that. The moves 6 . . . d6 and 7 . . .
i.b6 were n ot i n themselves novel.
Strictly speaking, the only new
move that Lasker came up with is
an untested l Oth move innovation
in a line that I am going to treat as
a side-variation. However, Lasker
deserves the credit for the basic
idea behind the defence , which is
to offer White the chance of
recovering his gambit pawn, but
at the cost of exchanging queens
and going into an ending in which
Black has every prospect of
6 0-0: Lasker's Defence 53
Evans Gambit: Section 1

e xploiting the weakness of White's
queenside (split a- and c-pawns).
8
de
It is interesting to note that

Ch igorin's final try (in the year
before his death) against the
Lasker line was the non-committal
K i.e3. I doubt whether this is
active enough to justify the
gambit . After 8 i.e3 ll:lf6 9 ll:l bd2
0-0 1 0 't!fc2 ed I I cd d5 Black
emerges with a satisfactory develop
ment - though the consultation
ga me Chigorin and Protoklitov v
Znosko-Borosvsky and Levin, St.
Petersburg I 907, ended in a draw
in 32.
8
de
't!t'b3!
This manoeuvre enables White
t o avoid the main obj ective of the
Lasker Defence, which is achieved

in the variation:
a) 9 't!fxd8+ li:lxd8 10 ll:lxe5 and
now:
a I ) 10 i.e6 1 1 ltld2 (significantly
weaker is 1 I i.a3 f6 ! 12 i.xe6
ll:l xe6 1 3 ll:lc4 i.c5 ! =!= JohnerZauer, corres l 9 I 2) l l . . . ll:le7 1 2
i.a3 ( 1 2 a4! ? - Levenfish) 12 . . . f6
13 ll:ld3 li:lg6 1 4 llab 1 f7
(Chigorin himself thought that
1 4 . . i. xc4 1 5 ll:lxc4 ll:le6 was
more accurate) I5 i.d5 lle8. This
was Chigorin-Pillsbury, London
1 899 (0- l , 47), regularly given as
=F . However, I am inclined to
agree with Romanovsky, who
claimed that 1 6 would be
quite accepta ble for White and
...
that all the damage was done by

miscalculation. Chigorin played
16 c4? c6 ! 17 i.xe6+ li:lxe6,
overlooking that the intended I 8
c 5 would fail to I 8 . . . ll ed8 .
a2) 10 ... ltlf6 was Lasker's
innovation. It isn't clear to me
that this is much stronger than
1 0 . . . i.e6. But the resulting
endgame is unlikely to appeal to
the gambiteer, and no body has
cared to put Lasker's judgement
that it favours Black to the test.
One should note the sacrifice
b) 9 i.xf7+ xf7 10 ll:l xe5+ is
unsound. It led to a quick win in
Ecke-Schonewald, corres I968:
10 ... f8? I I i.a3+ ll:lge7 I2
"it'f3 + I -0. But this was a bit silly
since it had been known for a long
time that Black beats off the
attack with IO . . . 'it>e8! 1 I "it'h5+ g6
I2 li:l xg6 ltlf6 1 3 't!t'h6 llg8 and if
I 4 ltlh4 ltl e5 ! .
Finally, there is nothing to
excite White in:
c) 9 ll:lbd2 "it'f6 10 i.d5 ll:lge7 1 1 h3
h6 12 ll:lc4 0-0 13 ll:lh2 lld8 1 4
ll:le3 i.e6 =t= Marshall-Blackbume,
Ostend 1 905.
9
10
11
12
13
14
,'' "' J tl\11( .,,41
i.g5
d5
xe7
i.xc6
li:lxe5
"it'f6
't!fg6
ll:lge7
xe7
't!fxc6
't!fe6
So far as I know, the only move

played or analysed. But 14 . .
not obstructing the bishop on c8,
seems a reasonable alternative.
ICA l
54 Evans Gambit: Section
1 5 llJc4 (49)
In his newspaper column Chigorin
analysed 1 5 'fi'a3+, with the
conclusion that White has good
chances after either 1 5 . . . 'ct>f6 1 6
lbf3 or 1 5 . . . c 5 1 6 f4. However,
Levenfish's 15
'ird6! spoils all
that. After 16 'it'xd6+ cd 1 7 lbc4
..te6 or 1 6 l:id 1 'fi'xa3 1 7 llJ xa3
..te6 Black has a very favourable
ending. White needs to keep the
queens on the board.
...
How to assess this position?

The fundamental directive for
White is: avoid exchanges ! H is
correct plan is to play 'ct>h 1 and
f4, aiming to use a spatial plus to
blot out the black bishops. In
particular White should try to get
in f5 in order to hinder the
development of Black's queen's
bishop.
To illustrate the chances I shall
give two games in full. In the first
White adopts the correct plan and
scores a stirring success - against a
rather ineffective B lack defence.
In the second White permits
6 0-0: Lasker's Defence

exchanges and finds himself in the
sort of ending that Black hopes to
enjoy in this variation .
ft. Skotorenko-Ahman, corres 1976:
1 5 . . . .ic5? (a waste of time)

1 6 llJbd2 li[d8 1 7 'ct>h l ! 'ct>f8 1 8 f4
'ti'e8 1 9 llJf3 b6 20 f5 ! h6 2 1 li[fe l
a5 22 'ti'c2 .ia6 2 3 lb ce5 'ct>g8 24
lbg4 .if8 (24 . . . .id3 25 'ti'c l ) 25
'it'f2 'ti'd7 26 'irg3 (White has
worked his queen and two knights
into attacking positions before
B lack's long-range bishops have
got any target in their sights) 26 . . .
'ct>h7 2 7 f6 .ic8 2 8 h3 c 5 29 'trh4
't!Vc7 30 llJg5+ 'i!?g6 (50)
3 1 lL!xf7 ! 'i!?xf7 32 fg ..txg4 (32 . . .

..txg7 33 l:tf l + 'ct>g6 3 4 lil6+! )
33 gf'it'+ lhf8 34 't!txg4 'it'e5 35
lii: a b l litab8 36 lit fl + 'ct>e7 37 litfd l
'ti'e6 3 8 "it'g3 1 -0.
2. Botterill-Williams, Pontypridd
1 978: 1 5 . . . :Sd8 ! l 6 1i'b4+ 'ct>e8 27
1 7 llJba3 't!t'e7 1 8 't!kb2 .ie6 1 9
llJxb6 a b 2 0 lbb5 i.c4 2 1 litfe 1
..txb5 ! 2 2 "ti'xb5+ 'ct> f8 2 3 li[e2 'tlfc5
24 't!t'xc5+ be (51)
6 0-0: Lasker's Defence 55

51
w
(No doubt I could have made a

better job of defending this
endga me, but weak pawns at a2
and c3 make it pretty horrific)
25 f3 l:la3 26 llc l c4 27 llb2 b6 28
ll l c2 lld3 29 Itb4 lldxc3 30 ll xc4
llxc4 3 1 li xc4 c5 32 ll c2 e7
(White has contrived to eliminate
one of his weaknesses, but the
other still remains and he faces a
protected passed pawn and the
more aggressive rook. H ere I
really should have seen that
33 lld2! was essential) 33 f2?
l!i>d6 34 llb2 l!i>c6 35 l!;>g3 c4
36 l!i>f4 'tt>c 5 37 h4 c3 38 Ite2
'it>d4 39 e5 d3 40 Ite3+ d2
0- l .
Since I have learned about this
position the hard way I ought to
be able to draw some useful
conclusions. The moral seems to
be that with the right plan White
may prosper, but without it he will
certainly flounder. IS
lld8 is
definitely Black's best move. But
has he any way of countering
...
White's plan? As White's main

threat is to play f5 it must
be rational to consider fore
stalling this with . . . f5. Thus from
diagram 49 we arrive at the
sequence:
lld8
IS
f8
I6
bd2
f5!?
I7
hi
This might look rather risky,
but I have not found any serious
objection to it.
As a last point, you should note
that a variation sometimes given
as leading to equality after 1 5 . .
lld8, namely 1 6 't!t'a3+ e8 1 7
lt:lxb6 cb, is in fact bad for White.
Black continues with . . . 't!t'e7 and
. . . i.e6 with easy play against the
weak White pawns.
.
Conclusion: Against the Lasker

Defence White should as a general
rule avoid the exchange of queens,
which usually leads to a favour
able ending for Black. Our main
line introduces a complicated two
knights v two bishops middlegame.
With chances for both sides?
That depends on what happens after
1 5 . . . lld8 1 6 bd2 1 7 h i
and here I cannot sensibly
say much more than that further
investigation is required . ..J'
-
Our next section takes up the

issue of whether 7 . . . i.b6 is really
superior to the alternatives that it
superseded.
6 0-0: 7 . . .tg4/ i.d7
56 Evans Gambit: Section 2
Section 2
The major alternatives are:
6 0-0:
7
i: g 4 a n d 7
. . .
l
2
3
4
5
6
7
e4
lt:lf3
.i. c4
b4
c3
0-0
d4 (52)
A 7 ... i. g4
. . .
B 7 . . . .t d7
d7
These moves were the subject of

some sharp theoretical exchanges
in the 1 892 Chigorin-Steinitz
world championship match. The
20th century, perhaps overly
impressed by Lasker's level-headed
pragmatism, has completely neg
lected them.
e5
lt:l c6
.tc5
.i.xb4
i.a5
d6
7
i. g4
W hite has tried three moves
here:
52
B
Al 8 \!fb3
A2 8 .i.b5
A3 8 \!fa4
Al
8
1!t'b3
.i.xf3!
A distinct improvement on 8
'W'd7, which was played in Evans
McDonnell, London 1 826 (or 7?).
This game is of great historical
interest, as it is the first Evans
Gambit on record, possibly the
first public outing for the Captain's
inspiration: 8 . . . \!fd7 9 lb g5 lt:ld8
10 de de 1 1 .i.a3 lbh6 12 f3 .i.b6+ 1 3
ct>h l i.h5 1 4 lld l 1!t'c8 1 5 ll xd8+!
'W'xd8 1 6 lt:lxf7 ! 'W'h4 ( 1 6 . . . .txf7
17 .txf7+ lt:lxf7 1 8 \!fe6+ and 16 . . .
lt:lxf7 1 7 .i.b5+ - everything ends
in mate ! ) 1 7 t!t'b5+ c6 1 8 1!t'xe5+
ct>d7 19 1We6+ ct>c7 20 .td6 mate.
It is understandable that this
game should have made a deep
impression on the loser, who was
to convince La Bourdonnais of
...
Black has other moves apart

fro m the 7 . . . .i.b6 of the Lasker
Defence. 7
ed 8 cd .i.b6, for
example, would transpose into the
old 'Normal Position' of the 5 .
i.c5 defence, examined on p. 8 2.
7 . lt:lf6 is a natural developing
move which here turns out to be
bad because of a fa miliar tactical
theme in the Evans: 8 \!fa4 ed
(White threatened to win a piece
with 9 d5, and if 8 . . . a6 9 .i.d5 i.b6
10 de lbg4 1 1 ed ) 9 e 5 ! lt:ld7 1 0
i.g5 f6 1 1 ef lt:lxf6 1 2 lt:lxd4 (a
convincing piece of analysis that
Cafferty and Harding ascribe to
Richter and Teschner).
...
..
6 0-0: 7 . . j_g41 i.d7 57
the value of the gambit in the

next decade.
9
gf
White is well advised to steer
clear of 9 x f7+ 'it>f8 (compare
our main line in the Goring
Gambit) 10 gf i.b6 1 1 i. xg8 ll xg8
12 d5 li:Ja5 13 't!Vc2 g5 +
(Levenfish and Sokolsky).
ed
9
li:Je5
1 0 "ti'xb7
1 2 'it>h1
This sets the little trap I I
li:Jxf3? 1 2 "ti'd5 . But in view of
what follows 1 1 cd!? may be a
better winning try.
...
11
12
13
14
15
litb8
't!Va6
"ti'xc4
li:Jxc3
't!Vxc3
llJ xc4
de
i.xc3
"ti'f6 (53)
l:lxe7+! ..t>xe7 20 "ti'xc7+ 'it>e8 21

'it'c8+ 1i'd8 22 1!1c6+ "t!fd7 23
WaS+ V:! - \.'2 . Such a sweetly
reasonable, delicately balanced
little game !
A2
8
b5
This was Chigorin's choice in

the l st, 3rd, 5th and 1 3th games of
his 1 892 match w i th S teinitz.
ed
9
cd
i.d7
li:J f6
10 b2
In the 1 s t game Steinitz played
the passive 10 . . . llJce7? ! and got
into trouble: 1 1 xd7+ 't!Vxd7 1 2
li:Ja3 li:Jh6 1 3 llJ c4 i.b6 1 4 a4 c 6 1 5
e5 ! d5 1 6 llJ d6+ ..t>f8 1 7 a3 'it>g8
1 8 l:l b l li:Jhf5 (54)
53
w
This was Santasiere-Marshall,

New York 1 926. Obviously 1 6
1t'xc7 allows a draw by 1 6 . . .
Wxf3+ 1 7 ..t>g l "t!g4+ etc, s o
Santasiere tried 1 6 e5!?. However,
it led, in a very pleasant way, to
the same outcome: 16 e5 de 1 7
lite l li:Je7 1 8 l:lxe5 l:lb5! 1 9
A famous and controversial

position. Chigorin sacrificed with
19 li:Jxfi!? ..t>xfi 20 e6+ 'it>xe6 2 1
li:Je5 - a procedure which, though
it led to success ( 1 -0, 3 1 ), might
well be viewed as a piece of
culpable over-exuberance when
the simple and powerful 1 9 a5!
was available. Lasker's comment
was delicious: " Even the most
6 0-0: 7

sympathetic critic would have to say
that Chigorin fought with a
corpse, gave him new life, and
then killed him again." Actually,
Chigorin's sacrifice can probably
be justified by analysis (2 1 . . . 't!fe8
22 lite ! Wf6 23 .ixe7+! seems to
be the critical line). But fortunately
we don' t have to prove that, as
Steinitz's moves are hardly likely
to be repeated.
11
. . .
.ig41 .id7
14 d5! 0-0 1 5 .ib2 is .
9
10
gf
cd (55)
ed
Chigorin's later suggestion of

1 0 .ib5 does not seem adequate to
me after 1 0 . . . .ixc3 I I .ixc6+ be
1 2 1!Vxc6+ f8 1 3 tl:l xc3 tl:le7 1 4
't!Vc4 d e 1 5 'ifxc3 tl:lg6 .
55
B
tl:la3?!
The 3rd and 5th games both

continued 11
0-0 1 2 d5 and
eventually ended in draws. By the
time the 1 3th game was played
Steinitz had realised that he could
j ust take the e-pawn: 1 1
lbxe4
1 2 d5 lbe7 1 3 'ira4 .i.c3! + (0- 1 ,
3 8). I find this pretty mystifying.
Why did it take so long for the
penny - or rather the pawn - to
drop? And why 1 1 lba3? There are
better moves, but I doubt whether
any will suffice to revive 8 .ib5.
..
...
A3
8
'W'a4
I n this position Black needs an

antidote to White's threat of d5.
But even if White does win a piece
he may not win the game, as his
broken kingside provides Black
with the resource . . . 't!fxf3-g4 with
perpetual check. The neatest
resolution would be 10 'irf6! I I
.ib5 'irxf3 1 2 .ixc6+ be after
which the only point of dispute is
who should give perpetual check
Black ( 1 3 'ii'x a5 't!fg4+) or White
( 1 3 tl:ld2 .ixd2 1 4 1i'xc6+ e7 1 5
'irxc7+ <li'e8 etc) This i s analysis
by Chigorin to the 1 7th game of
the 1 892 match, in which Steinitz
chose the more interesting 10
a6 !? 1 1 .id5 tl:lge7 12 .i.xc6+
tl:lxc6 13 d5 b5 14 Wa3 lb d4
(forced: 14 . . . .ib4? 1 5 Wb2 or
14 . . . 'iff6 1 5 de Wxa l l 6 .ib2 )
...
This was the improvement

Chigorin produced the next time
the Evans was on the agenda, in
the 1 5th game of the match.
.i.xf3
8
Steinitz's first reaction was 8
ed 9 cd (9 ti:lxd4!?) 9 ... a6 1 0 .id5
.ib6 1 1 .ixc6+ be 12 't!fxc6+ .i.d7
13 't!fc3 tl:le7. The actual con
tinuation now was 14 lba3 0-0 1 5
tl:lc4 d 5 ( 1 -0, 4 7), sometimes given
as =, sometimes as ;!;. But of course
White should not permit . . . d5:
...
...
6 0-0: 7

15 'ti'xa5 'ti'f6 1 6 'ila3 ! (56)
S teinitz now played the feeble

ll:\c2? and had no chance
after 1 7 'ti'd3 ll:\xa l 1 8 'ti'e2 ! 0-0 1 9
i.b2 ( 1 -0, 39). But what happens
after 16
b4!? White could reply
17 't!b2, but then 17 . . . 'ti'xf3 1 8
"tifxd4 'i!Vg4+ i s that perpetual
again. Chigorin gave a very
curious piece of analysis with 16 . . .
b4! 1 7 't!d3 ! ? ll:\xf3+ 1 8 'it>g2
ll:\h4+ 1 9 'it;h3 , claiming an
advantage for White after both
19 . . . lL!g6 20 1!t'b3 and 19 . . . 1!fxa l
20 'it>xh4 1!fxa2 2 1 ll:\d2 0-0 22
'ti'd4. At first I thought this must
be potty. How could White be
better when Black has rook and
three pawns v bishop and knight
and the white king is roaming on
h4? But Chigorin has a point in
that the black queen is right out of
play on a2 in the last variation. So
play . .. 0-0 instead of 20 . . .
!t'xa2 - with a totally unclear
position .
B
i. d7
7
Advocated and analysed by
16
...
...
. . .
i.g41 i.d7 59
Alapin. It does at least guard

against l!t'a4, though one might
well suspect the move of being
overly passive. Given such a quiet
defence, White has many options.
He could, for example, even try
the extragavant 8 ll:\g5 ll:\h6 9 f4!?,
with the idea that after 9 . . . ed 1 0
ll:\ xd4! ? 1 1 ll:\xf7 ll:\xf7 1 2 'it'h5
0-0 13 'it'xa5 ll:\c2 14 i.b2 ll:\xa l l 5
i.xa l i t i s not at all easy for Black
to combat White's diagonal pres
sure. The moves usually considered
- the moves actually played on the
few occasions on which this
variation has got on the board are:
Bl 8 !t'b3
B2 8 de
Bl
8
'ifb3
This is what Chigorin played in

the 7th and 9th games of the 1 892
match, when Steinitz introduced
7 . . . i.d7 as an interlude in the 7 .. .
i.g4 campaigns. It seems that
Alapin had a hand in putting the
world champion up to this, having
supplied him with some private
analysis. It would be interesting to
know what Alapin's secret package,
intended to dish his compatriot
Chigorin's title bid, contained. I
suppose the contents were not
very explosive since the ' mine'
turned out to be a damp squib Steinitz lost the 7th game and
drew the 9th.
Wf6
60 /:.' vans Gambit: Section 2

8 .
'it'e7 is rather weaker
because it gives White a later i.a3
with tempo: 9 de de 1 0 ll d 1 and
now:
a) 10
lidS 1 1 lld5 i. b6 1 2 i.b5
llJf6? ( 1 2 . . . f6 is necessary, but
probably ) 13 i.a3 Chigorin
and Saburov v Alapin and Schiffers,
St Petersburg I 897.
b) 10 ... 0-0-0 I I lt:l bd2 lt:lh6 1 2
i.a3 1Wf6 I 3 i.d5 was theory
according to Keres, put to the test
in Sko torenko-Timejer, corres
I 976: 1 3 . . . i.b6 I4 lt:l c4 1 5
liab i i.g4 1 6 lld3 lt:l e7 I 7 i.xe7
'lrxe7 I 8 1!t'a4 b8 I 9 llJfxe5 i.e6?
( I9 . . . i.c8 was called for) 20
i.xb7 ! ll xd3 (if 20 . . . xb7 2 I
llJa5+ Wc8 2 2 lbac6 ) 2 l llJc6+
xb7 22 llJxe7 llxe7 23 llJa5+
c8 24 lt:lc6 lled7 25 1i'a6 mate. A
nice example of sustained pressure
in the Evans: Black, although
formally developed, never got
properly organized.
..
...
9
de
de
10 lld1
1 0 'it'g6 I I :adt looks an
alternative worth investigating, if
ever this variation comes back
into favour.
10
11
h6
i.a3
The liquidation 11 i.xti+ 1!rxf7
1 2 1i'xf7+ xf7 1 3 llxd7+ llJge7 1 4
fl e6 1 5 lld3 llad8 was pl ayed
in Chigorin-Steinitz, 9th game,
Havana 1 892, and is usually
condemned because the position
after Black's 1 5th is at least +, even
6 0-0: 7
...
k,g41 i.d7
though Chigorin hung on for a

draw. White has problems with his
queenside development.
Of course, it would be far worse
for White to try to recoup the
pawn with 1 1 1!t'xb7?? since this
loses the queen to I I . . . llb8 12 "t!Va6
llb6.
11
lld8
i.b6
1 2 lt:lbd2
13 i.dS (57)
All so far as in the 7th game of

the I 892 match. Steinitz now
played the inferior 13 . . . lt:laS?!
and after 1 4 1i'b4! c5 1 5 'ft'b2 had
been forced to make a most
undesirable pawn move. He lost in
28 moves, but only after a truly
vile blunder which is discussed in
our section on 'The Evidence of
Games' in the Introduction. A
natural and better move, as
Steinitz himself indicated after
wards, was 13 . . . llJge7. The
position remains difficult and
complicated, though I must say
that I cannot see what White has
for his pawn, e.g. 1 4 llJ c4 0-0 - and
6 0-0: 7 . . .ig41 i.d7 61
now what?
82
me what is wrong with 15 )

14 . . llJa5 15 llJxa5 .txa5 1 6 ..txe 7
1Wxe7 1 7 1Wxb7 a6 1 8 I[fd 1 I[fb8 1 9
"i!t'xa8 lha8 20 .ixa8 .ixc3 2 1
ll:ab 1 oo. Yet surely only Black
can hope to win legitimately from
this position.
.
9
10
de
liJbd2
.idS (58)
de
'fi'f6
58
B
Conclusion: 7 . . . .ig4 is a forcing
There is a second pawn on offer,

but it would be foolhardy to grab
it:
a) 1 0 . .ixc3 1 1 I[b l .ixd2 1 2
'fi'xd2 (Hardin's original 1 893
analysis of this line gave 12 llJ xd2
lbd8 1 3 llJc4 c6 14 .ia3 . But
that too nearly resembles a help
mate) 1 2 . . . I[b8 1 3 lbg5 lbh6 1 4
.ia3 'it'f4 1 5 'fi'xf4 ef 1 6 I[fc l f6 1 7
nxc6! Ljundqvist-Kjellander,
2nd World Corres Ch, 1 956-9.
A more prudent evolution
would be:
b) 10 ... llJ g e 7 I I 'fi'b3 0-0 12 lbc4
h6, and now White can try 1 3 a4,
which is indeed the sole feature
that distinguishes this position
from the one we reached in B I . An
untested piece of analysis by
Levenfish ploughs on: 13 . .. .ib6
14 .ta3 ( 1 4 aS .ic5 15 .t xc6 llJ xc6
1 6 "i!t'xb7 I[ fc8! + is the staildard
note here. But it isn't obvious to
..
move, after which both players

will be swept along on a tide of
threats and counters. Whether
White chooses 8 'it'b3 or 8 'it'a4 the
appropriate outcome seems to be
a draw. 7 . . . .td7 is much more
difficult to assess and there is
too little practical experience with
the move to support any firm
conclusion. It is not at all clear
that these moves are inferior to 7
. . . .tb6. If Black knows the
analysis and is content with a
draw, 7 . . . .ig4 is probably the
safest choice. For winning pur
poses 7 . . . i.b6 or 7 . . . .id7 would
have to be preferred.
Peeling back, we can take 7 d4 to
be the only continuation that
merits consideration after 6 . . . d6.
So Black's 6th move alternatives
are next on the agenda.
Section 3
6 0 - 0 : Black's 6th M ove
Alternatives
e4
e5
2
3
lbf3
i.c4
lbc6
.ic5
6 0-0: Black's 6ths
4
5
6
b4
c3
0-0 (59)
59
B
Given the high esteem in which

the L asker Defence is held,
alternatives to 6 . . . d6 have
vanished from the board and from
the mind. So this section is a sort
of lost property office with a few
abandoned variations waiting dis
consolately on the shelves for
someone to come along and pick
them up.
What have we got in store? The
records show:
A 6 ... .liJge7
B 6 ... 1!Vf6
c 6 ... .liJf6
A
.llJ ge7
7
8
9
.i.xb4
.i.aS
.llJ gS
dS
.liJxdS
A s this ends oo , is it possible to

suggest something stronger?
a) 9 .llJ xf7!? xt7 1 0 't!Vf3+ e6 1 1
.i.a3 .i.b6 1 2 li e ! .llJ a5 1 3 ll xe5+
xe5 14 d4+ e6 1 5 g4 (Morphy
Ford, New Orleans 1 858) fails to
Maroczy's 1 5 . . . c5 !
b) 9 .WhS!? seems strong to me. If
9 . . . g6 1 0 't!Vh6 threatening both
1Wg7 and d4, and if 9 .. . 1!Vd7
White has a choice between 1 0
't!fxt7 + ( = , I think), 1 0 .i.a3, 1 0 d4
and 10 .liJxh7 ( oo, but I think
probably good for White: there
is a threat of .liJ f6+ and Black
will have to renounce castling
rights).
9
h6
O therwise White is winning:

a) 9 . . ed? 10 .llJ x t7 <ct.>xt7 1 1 1f3 +
with a devastating attack.
b) 9 . 0-0 10 de .i.e6 1 1 't!Vh5 h6 1 2
.llJ x e6 fe 1 3 .i.xh6 ! .
.
10
de
The move that has been analysed,

but 10 .llJ x f7!? xt7 1 1 1!t'f3+ e6
12 .i.a3 must surely be pretty
strong now that Black has played
the irrelevant . .. h6 rather than
the useful . . . .i.b6 of Morphy
Ford.
-This ought to be the try, since
10
7 d4 ed would transpose directly
11
into the next section. If that were

the best, it would undermine any
argument for preferring 6 0-0 to
6 d4.
ed
d4
i.xdS
hg
.i.e6 !
I f 1 1 . . . .llJ xe5 1 2 lle l f6 1 3 .i.xg5

( Unzicker).
12
13
i.xe6
lhd1
1xd1
fe
14
.t xgS
6 0-0: Black's 6ths 63
lbxeS (60)
60
w
xe7 1 4 d6+ <t>f8 1 5 'i!Vb4!

( 1 5 lbxe5 f6 1 6 liJf3 .tc5 ! had been
played in the 1 5th game, which
Steinitz won) 1 5 . . . f6 1 6 .tb3 g6 1 6
lbc4 with a totally winning
position for White. However, see
the section in the general intro
duction on 'The Evidence of
Games' .
Against 7
h6 Chigorin
proposed to play 8 .te3 with the
idea 8 . . . .tb6 9 de liJxe5 10 liJ xe5
1!Vxe5 1 1 .td4! . More incisive is
8 de lbxe5 9 liJxe5 'i!fxe5 1 0
1!t'b3 1!t'h5 1 1 e S ! (Ciocaltea
Brzozka, Polanica Zdroj 1958) Black cannot get his pieces out.
Compare the very similar situation
in Goring Gambit Section 3,
variation A .
...
Harding and Botterill ( 1 977)

sa id 1 5 ;!;, but this assessment
is probably wrong because of
1 5 . . lbc4 when White has serious
problems with the dominated
knight on b l and roo k on a l .
However, I think that the gam
biteer can rejoice in the improve
ments suggested at moves 9 and
10.
B
'ii'f6
6
Steinitz showed a stubborn
attachment to this move, which
cost him dearly.
.
d4
lbh6
This - would you believe it? was intended as an improvement

on 7
liJge7, which Steinitz had
played in no less than eight games
of his 1 889 match against Chigorin.
Chigorin won the eight-game
match-within-a-match +4, 1 , -3.
The last of these eight and 1 7th of
the match went 7 ... lbge7 8 d5
lbd8 9 \ta4 .tb6 I 0 .tg5 'ii'd 6 I I
lba3 c6 1 2 ll ad l '@'b8 1 3 i.. x e7
...
.tgS
1!t'd6
Less greedy is Bogoljubow's

suggestion 8
'itg6 9 d5 liJ b8 1 0
.txh6 \txh6 1 1 lbxe5 0-0 1 2 d6
lbc6. Vasyukov and Nikitin then
give 13 liJg4 'i!Vg6 14 llel " and
White stands bette r" (all repeated
in CO). But this is rubbish, since
now that the bishop on aS is
protected Black will just play
13 . . . 'i!Vxd6. There is no need for
White to play 1 2 d6. Simply
U .i.d3 intending either liJc4
or lbg4 (depending on whether
Black plays 12 . . . d6 or 12 . . . c6)
is good.
...
10
dS
\ta4
11
12
liJ a3
.te2 (61)
lbd8
.tb6
c6
6 0-0: Black's 6ths

6/
B
62
B
It is astonishing that Steinitz

persisted in defending such hideous
positions. But he was unable to
improve on 12 . . . i.c7 1 3 c4 1!Vf8
14 d6! i.xd6 1 5 b6 llb8 1 6 'it'xa7
when Black is in a bad way, as
sh own by:
a) 16 ... lbg4 17 h4 lbe6 1 8 i.xg4
xg5 19 f5 e6 20 llfd 1 i.c7 2 1
l2J a8 xa8 2 2 \!fxa 8 'it>d8 23
Itxd7+! 'it>xd7 24 I!d l + l -0
Gunsberg-Steinitz, 1 2th match
game, New York 1 890- l .
b ) 1 6 . l2Je6 1 7 i.c l ! g8 1 8 i.a3
c5 1 9 I!ad l f6 20 i.c4 i.c7
21 l2Jd5 Chigorin-Steinitz,
tele graph match 1 890- 1 .
..
l2Jf6
6
This looks considerably more
sensible than A or B. Indeed, if it
had not been for Lasker, it would
probably have become main line
theory in the Evans.
7
d4 (62)
xe4
Others:
a) 7 ... ed allows White to launch a
full-scale assault with 8 i.a3
(be tter than 8 e 5? d 5 9 i.b5 as
Morphy played against Anderssen

in their 1 85 8 match) 8 . . . d6 9 e5 !
and now:
a l ) 9 . . . de 10 't!Vb3 "@d7 I I lilei
'it'f5 1 2 i.b5 d7 1 3 \!fd5 ! i.b6 14
l2J xe5 l2Je7 15 l2Jxd7 ! 'it'xd5 1 6
l2J f6+ ++ Steinitz-Plihal, Vienna
1 862. From such play as this who
could foresee the gritty old master
who would defend variation B?
a2) 9
l2Jxe5 10 l2Jxe5 de I I 'i!t'b3
1!Vd7 1 2 lle l e4 1 3 l2Jd2 i.xc3 I 4
l2Jxe4 i.xe i I S xf6+ g f I 6
I! xe l + 'it>d8 1 7 't!Vf3 f5 1 8 1!Vf4 f6
I 9 \!fh6 Levenfish .
a3 ) 9 . . . l2Je4 1 0 e d l2J xd6 (perhaps
10 . . . cd!? is better) I I lle I+ e7
1 2 l2Jg5 0-0 1 3 't!Vh5 i.f5 I 4 l2Jxf7!
l2Jxf7 15 llxe7 Levenfish.
b) 7 ... 0-0 8 i.a3 and now:
b l ) 8 ... d6 9 de l2Jxe4 1 0 i.d3 l2Jc5
1 1 i.xc5 de I2 \lfc2 h6 1 3 bd2 J;
Botterill and Harding.
b2) 8 ... ll e8 9 d5 b8 10 d6! cd
( l 0 . . . c6? 1 1 g5 ) I l i.xd6
l2Jxe4 I2 xeS l2Jxd6 1 3 1!Vxd6
because f7 caves in: I 3 . . . lle7 1 4
l2J xf7 llxf7 1 5 "ildS , Kolisch
Winawer, friendly game, London
..
6 0-0: Black's 6ths 65

1 8 83.
c) 7 ... d6 8 1!fa4 a6 9 .idS .ib6 10
de ll.'lg4 1 1 ed Unzicker.
8 lt:l xe5
8 de 0-0 9 .idS lLlcS is probably
not good enough for White. But if

now 8 . . . ll.'l xe S 9 de 0-0 1 0 1!fg4
ll.'lcS 1 1 i.h6 .
8
9
10
11
.t a3
ll.'l xc6
1!1a4 (63)
0-0
d6
be
.ixh7+. This is old but untested

analysis usually given as = or !. In
view of the bishop pair and
Black's exposed king White must
have the advantage.
Conclusion: 6 . . . liJge7 and 6 . . .
't!Vf6 had better stay o n the shelves

gathering dust. But won' t some
body come and claim 6 . . . lt:l f6?
Chigorin's idea at move 1 1 seems
to make it as good a defence as any
against the Evans.
Peel-back: The overall conclusion
on 6 0-0 is interestingly problematic
in all of sections 1-3. But does 6 d4

offer White better chances? This is
taken up by the next three
sections. I will start with what I
consider to be Black's most
reliable defence.
Chigorin had a very interesting
idea in this position: 11 .. 1!1g5!?
and if 12 f4 g6 ! ( 1 2 . . . 'ii'h S ?
1 3 fS ) 1 3 WxaS i. h 3 1 4 g3 lt:l xg3
l S gS xg5 1 6 fg .ixfl 1 7 i.xfl
ll.'le4 18 .te l f6 ! and White's
minor pieces are disorganized
whereas the black rooks are
active. I have not been able to find
anything much against this. Perhaps
White's best is just 1 2 .te l 'tWhS 1 3
xc6 .if5 , though Black seems to
have the edge.
The less enterprising 1 1 i.xc3
1 2 lt:lxc3 lt:l xc3 is more pleasant
for White: 1 2 1!1xc6 .ie6 13 .id3
ll.'ld5 14 .te4 lt:lb6 15 1!1c2 litb8 1 6
.
...
Sectio n
6 d4:
M a i n Lin e
1
2
3
4
5
6
7
e4
ltJ f3
i. c4
b4
c3
d4
0-0
e5
lLl c6
.ic5
.ixb4
.tas
ed
liJ ge7! (64)
As I explain in the introductory

survey, the variation introduced .
by this move constitutes my main
reason for thinking that 6 d4 i s not
superior to 6 0-0. Black sensibly
prepares to hit back in the centre
6 d4 Main Line
with . . . d5.
situation that can be compared

with the old 'Normal Position' of
Anderssen's heyday (see Evans
Gambit 8). White has pressure in
the long black dia gona l and a
cramping central wedge. But if
Black reacts quickly with . . .td7,
. . . c5 and a general queenside
advance, I feel that the extra
pawn ought to tell.
64
w
8
9
8
cd
Other moves do not convince,

although White would dearly like
to st rike immediately:
a) 8 ltlg5 d5!
a l ) 9 ed lbe5! and Black usurps the
initiative. For example, 10 'ffxd4
f6 1 1 ll e l i.b6 and White is in
trouble.
a 2) 9 .txd5? lbxd5 10 Wh5 g6 1 1
'ti'h6 .te6! =F Estrin Kondali
corres 1 97 1 . As a reminder that it
is n ot always Black who gets
quickly rubbed out in the Evans
note the drastic finish: 1 2 ltlxe6 fe
1 3 ed Wxd5 1 4 i.g5 ll f8 1 5 cd
lbxd4 1 6 ltld2 llf5 ! 1 7 .te3 lbe2+
1 8 h 1 lith5 0- l . The point is
19 1!fg7 lhh2+ 20 xh2 Wh5
mate.
b) 8 't!rb3!? (well-motivated: to
stop . d5) 8 . 0-0 9 cd .tb6 10 d5
lba5 (as White's bishop on c4 is no
longer a great piece, it may be
better to play 10 . . . ltld4!?: 1 1 Wd3
. ltlxf3+ 1 2 'ifxf3 ro) 1 1 Wc3 ltlxc4
1 2 'ttxc4 d6 1 3 .tb2 Bloch-Foley,
Islington 1 97 1 . This is an interesting
-
. .
ed
d5
lb xd5 (65)
65
w
H ere White has an important

choice to make between:
A 1 0 1!fb3
B 1 0 i.a3!
Keres once played 10 .tgS, but
after 1 0 . . . Wd6 1 1 1!rb3 i.e6 Black
is achieving harmonious develop
ment and can meet 1 2 ltlbd2 i.xd2
1 3 lbxd2 ( i ntend i ng ltle4) with
13 . . . ltlxd4! 14 Wxb7 ( 14 1!rb2
Wb6; 1 4 'ti'a4+ Wc6) 14 . 0-0 1 5
lbe4 We5 ! 1 6 llae 1 litfb8 1 7 @'a6
ltl b4 +.
..
A
10
Wb3
This forcing move has been the

usual choice in actual ga mes. But I
6 d4 Main Line 67

do not trust it. Surely the
gambiteer is not making the best
use of his potential by capturing
on b7!
10
.te6
10 . . ll:lce7?! looks suspiciously
.
passive and was refuted in great

style in Erlandsson-Demiden ko,
corres 1 980-3: 1 1 .ta3 c6 1 2 .txe7
xe7 ( 1 2 . . . 1Wxe7 13 .i.xd5 cd 1 4
b5+ costs Black a piece) 1 3
.i.xd5 'fi'xd5 1 4 1!t'a3+ Wd8 1 5 It c 1
b6 (to avoid what follows he
would have to try 1 5 . . . .tc7 , but
then 16 ll:lc3 1!t'd6 17 1!t'b3 seems
strong) 16 Itc 5 ! (66)
66
B
1 6 . . . be 1 7 1Wxa5+ 'it>e8 1 8 ltlc3

1M8 19 'fi'xc5 .te6 20 d5 ! .txd5
(20 . . . cd 2 1 ltlb5 !) 21 Ite l + .te6
22 ll:ld4 1!t'd7 23 ll:lf5 lilg8 24 It d l
.txf5 (24 . . . 't!t'c7 25 ll:lb5; 2 4 . . .
.i.d5 2 5 ltlxd5 cd 26 Itxd5 ) 25
llxd7 .txd7 26 ll:le4 f6 27 h3 Wd8
28 't!Vd6 ct>c8 29 ll:lc5 lld8 30 ltl xd7
1 -0. 1!t'xc6+ will win one of the
rooks.
11
1!t'xb 7?!
This has met with success, but I

think that the mission should
really prove fatal. Better 1 1 .taJ!?

although B lack has a good reply in
1 1 . . . 1!t'd7 ! , e.g:
a) 1 2 1!t'xb7?? lilb8 and . . . llb6 H.
b) 1 2 .i.b5 ltl de7 !
c) 12 Itdl 0-0-0 and White has
nothing for his pawn (Selfe
Harding, corres 1 974-5).
d) 1 2 ll:lbd2 is, as Cafferty and
Harding suggest, relatively best.
11
12
ltldb4
.tb5 (67)
Forced, since he does not get

nearly enough for his queen after
12 .txe6 Itb8.
67
B
With three moves to look at for

Black, though I would say only
one that really matters:
At 12 ... .i.d7
A2 12 . 0-0
AJ 12 ... .i.d5 !
..
Al
12
.td7
White gets plenty of tactical

chances, although it is far from
clear what result perfect play
would produce. With such over
whelming complications perfect
6 d4 Main Line

play is unlikely to be found in a
game - and not all that likely to be
revealed by the following analysis,
I fear, in spite of the aid of
some penetrating investigations
by Cafferty.
13
lael+
'it;f8 (68)
68
w
Black is threatening to win the

queen with . . . l:tb8, so White must
take active measures. Two moves
come into consideration - 14 i.a3
and 14 e5 . The first was tried in
a recent game:
Davis-Peters, USA 1 98 3 : 1 4 i.a3
ItbS ( Peters shows that unpinning
loses material: 1 4 . . . 'i!?g8 1 5 i.xb4
Itb8 16 i.xc6 and now both 16 . . .
Itxb7 1 7 i.xdH Wxd7 1 8 i.xa5 and
1 6 . . . i.xb4 1 7 i.xd7 i.xe 1 1 8 11Vd5
should win for White) 1 5 'it'a6
Itb6 ? (but here 1 5 . . . g8 ! was
right. Peters gives the following
variations after 1 6 i.xb4: a) 1 6 . . .
Itb6? 1 7 1!rxa5 ! as i n the game;
b) 16 . . . xb4 17 'it'xa5 Uxb5 1 8
'it'xa7 c2 1 9 c3 xe 1 20 Itxe 1
and White has compensation for
the exchange; c) 1 6 . . . i.xb4 1 7
l:tc l ! e5! oo ) 1 6 'it'x a 5! xa5 1 7
.txb4+ rt>gS ( 17 . . c 5 ? 1 8 .txa5 !
.
i.xb5 1 9 de 'it'f6 20 cb Peters)

1S i. xd7 1Ixb4 (or 1 8 . . . 'it'xd7 1 9
.txa5 ) 1 9 lieS+ 'it'xeS 2 0 .txeS
g5 (the position has s tabilised to
an ending with bishop and knight
v rook which we can safely say is
. But it would be a shame to
omit the finish of this rousing
game) 2 1 h3 'i!?g7 22 a3!? lbb3
( Peters gives 22 . . . 1Ic4! 23 t'D bd2
n xe8 24 t'D xc4 lbxc4 25 xg5
t'D d2 as the last chance, but here 25
a4 and 25 lacl lbxa3 26 Uxc7 are
both ) 23 ab xa1 24 i.a4 la bS
25 b5 c6 26 c3 cb 2 7 xb5 b3
2S xa7 c1 29 c6 la b2 30 d5
e2+ 31 'i!'h2 rt>f6 32 d6 'i!?e6 33
fe5 d4 34 xd4+ 'i!?xe5
35 d7 1 -0.
Now let us turn to 14 t'De5. This

was analysed at some length by
Cafferty and Harding in their
1976 book and then , i n response
to the annotations to the Davis
Peters game, some further refine
ments were added by Cafferty and
Nunn in the British Chess Magazine
( 1 984). Here are the main con
clusions:
14 e5 and now:
a) 1 4 ... labS? 1 5 xd7+ 'it'xd7
1 6 'it'xb8 .
b) 1 4
xeS 15 lixe5:
bl) 1 5 ... libS 16 'it'xb8 ! ( 1 6 "@xa7
i.xb5 1 7 'it'xa5 1Wxd4 +) 1 6 . . .
'it'xb8 1 7 i.xd7 and the threat of
lie8+ recoups the queen . But who
stands better? Approximate equality
results from 1 7 . . . "@d8 1 8 lle8+
1t'x e8 1 9 .txe8 'i!?xe8 20 lba3.
..
6 d4 Main Line 69

Attempts to win the rook on a l
seem unsatisfactory, e.g. 1 7 . .
lt:Jc2 I S i.a3+ 'i!?gS (better l S . .
lt:J xa3 1 9 ltJ xa3 'it'dS 20 lieS+
'i!t'xeS 21 i.xeS 'i!?xeS +) 1 9 lieS+
'i!t'xeS 20 i.xeS lil xa l 21 i.a4
intending i.b2xa l .
b2) 1 5 ... f6? 1 6 i.xd7 ! is s trong
because 16 . . . fe loses to 1 7 'it'f3+
b3) 15 . . . c6 1 6 i.c4 (69) is critical .
.
So I think that in the last

analysis 14 ltJe5 proves inadequate
and hence White should take his
chances with 14 i.a3 .
A2
12
14
15
Now Cafferty and Nunn con

cocted the following spectacular
line against 16 . . . i.b6 (Peters):
1 7 lif5 ! ( 1 7 i.f4 1!fcS ! ) 17 . . . libS
18 i.f4! lixb7 19 i.d6+ 'f!/e7 20
lixf7+ <ct>e8 21 llxe7+ 'i!?d8 , with
best play appearing to be 22 i. xb4
i.xd4 23 li xd7+ <ct>xd7 24 i.c3
lixb 1 + 25 llxb 1 i.xc3 and
although White can win a pawn
with 26 llb7+ it must be doubtful
whether he can win the game
(opposite-coloured bishops).
That is a very decorative
variation, but the initial move
is not the best. Black should play
16 ... ll b8 ! 1 7 'it'xa 7 i.b6 lS 't!Va3
( Cafferty and Harding) when he
can win by 1 8 .. i.xd4 19 .ib2
.ie6! !
.
0-0
I ronically, this sensible and, to

all appearances, 'modem' defensive
treatment predates the Davis
Peters game by well over a
century!
lib8
1 3 i.xc6
'f!/xa7
'it'c5
lt:Jxc6
i.d5
Anderssen-S.Mieses, Breslau 1 867,

continued 1 6 i.a3 lle8 1 7 ltJ bd2
lle2 lS li fd I 'f!/d7 19 ltJfl .
Anderssen eventually won ( 1 -0,
7S), but that cannot be put down
to the opening. Black's active
bishop pair fully compensate for
the pawn.
A3
12
i.d5 ! (70)
In my opinion this is the

strongest. As Cafferty and Harding
acknowledge in their book, I first
suggested it some ten years ago,
but it still continues to escape the
attention of other anal ysts.
70
w
6 d4 Main Line

13
lt':le5
Since 12 . . . .id5 does not break
the pin on the knight on c6 one
would expect this to be critical.
Besides, with an angry swarm of
black pieces encircling the queen,
nothing else will do:
a) 13 tt:lc3 lib8 14 .ixc6+
lt':lxc6 =t=F because the knight on
c3 hangs.
b) 13 liel + f8 1 4 .ia3 g8 1 5
i.xb4 lt':lxb4 (here i t makes all the
diffe rence that the bishop is on d5
instead of d7) 16 't!'xa8+ 'ti'xa8
1 7 lie8+ '@xe8 18 i. xe8 .ixf3
intending . . . lt':lc2 H.
lib8
.ixc6+
Weaker is 14 lt':lxc6 lii: x b7 1 5
13
14
lt':lxd8+ lhb5 (the knight on d 8 is

trapped) 16 lt':lc3! lt':lc2!? 17 lt':lxb5
tt:lxa l 18 .if4 xd8 19 l ha l a6 +.
14
15
tt:lxc6
71
w
'ti'd3
10
l:tb6
0-0 (71)
.ia3!
i.e6
Note that 10 . . . i. b4?? loses to

l l i.xd5 '@xd5 12 i. xb4 tt:lxb4 1 3
'ti'e l +.
11
.ib5
Black has no problems after

1 1 tt:l bd2 .ib4 now.
11
.tb4
Black might even have time to

play 1 1 ... f6 !? intending . . . fl.
For example, 12 1!Va4 .ib6 1 3
.t xc6+ b e 1 4 1!t'xc6+ fl. and
Black is certainly not worse.
This possibility gives Black more
chances of advantage than 1 1 . . .
i.b4.
12
1Wa6
Black's advantage is obvious

after 15 lt':lxc6 l:lxb7 16 tt:lxd8
xd8 .
15
16
So White did not lose his queen

after all! But there is nothing
for him to rejoice about in this
position. Black has the two bishops
and much better development
( +).
B
.ixc6+
Others:
a) 12 .ixb4 tt:lxb4 1 3 i. xc6+ might
tranpose, but that move-order
gives Black the option 13 . . . ll:lxc6
1 4 tt:lc3 lt':le7 oo.
b) 12 'ffa4 'ti'd6 13 tt:le5 ? ! 0-0! 1 4
tt:l xc6 b e 1 5 .ixc6 lii: b 8 +
Muravlev-Tanin , corres 1 967-8.
12
13
14
be
tt:l x b4
"t!Vd6 (72)
As the queen is rather awkwardly
exposed on this square Black
might also consider 14 ... l:tb8!?
1 5 a3 ll:ld5 16 'ffx c6+ 'ftd7 1 7
"t!Va6 oo.
.ixb4
"t!Va4
6 d4 Main Line 71
but as the motto has it 'Be

prepared ! ' . So our next section
deals with the notorious 'Com.,
promised Defence .
72
w
'
Sectio n 5
Compro m ised
Defence
We have been following some

analysis that originated from
Levenfish, who now gave 1 5
lbbd2 0-0 1 6 lbe4 f4 1 7 lbeg5 =
or 1 6 . . . 't!te7 1 7 lbc5 =. The reader
should be warned, however, that
although these assessments have
been frequently repeated they
have never been properly tested in
practice.
Conclusion:The 7 . . . lbge7 line
seems to be an excellent defence
against 6 d4. Variation A with
10 'ifb3 is tactically complex, but
Black forces a significant advantage
with 1 2 . . . .td5! (A3). 1 0 .ta3 !
seems more satisfactory for White
(probably =) but the possibility
of 1 1 . . . f6 !? for Black deserves
investigation . The activity of
Black's minor pieces is a striking
feature of the whole line.
-
Peel-back: So 7 . . . li:Jge7 is good .
But what if Black just keeps on

eating pawns with 7 . . . de? Few
people will play like that nowadays,
1
2
3
4
5
6
7
e4
lt:\3
.t c4
b4
c3
d4
0-0
e5
li:Jc6
.tc5
.txb4
.ta5
ed
de (73)
73
w
If gambit play ever has any

justification then surely this
ought to get smashed! The view
that the Compromised Defence is
not only too risky, but also
objectively inferior, is practically
universal. Few will disagree with
ECO's a ssessment of the
resultant. In consequence it is a
fairly safe bet for Whites who
embark on the Evans Gambit that
Black will not play like this. I t
,
wasn't always so. A century ago

the Compromised Defence was
theoretically active. Chigorin could
not see a refutation, which was
why he preferred 6 0-0 to 6 d4.
Zukertort played it as Black on
several occasions, including four
games of an 1 87 1 match against
Anderssen . There is an interesting
qu estion as to whether modern
theorists are just right about the
objective assessment, or whether
their threshold for acceptable
risk is set at a lower level.
8 'it'b 3
''f6
The alternative is 8 . . . 'it'e7 9
lh x c 3 and now:
a) 9 ... lbf6? 10 lbd5! lh xd5 1 1 ed
lbe5 12 lbxe5 "t!fxe5 13 J.b2 "i!t'g5
1 4 h4! 't!Vxh4 1 5 J. xg7 llg8 1 6
llfe l + d8 1 7 'it'g3 ! 1 -0 Fischer
Fine, 'friendly game', New York
1 963.
b) 9 . . . J.xc3 1 0 "t!fxc3 lbf6 ( 1 0 . . .
f6!?) 1 1 J.a3 d 6 1 2 e 5 lhe4 1 3 1i'b2 !
lbxe5 1 4 lhxe5 't!t'xe5 1 5 ll fe 1
(Fischer).
c) 9 . . . "t!fb4!? 10 J.xf7+ d8
1 1 J.g5+ lbge7 1 2 lhd5 't!xb3
1 3 ab and now:
c 1 ) 13 . . . J.b6 14 llfc l h6 1 5 ll xc6
hg 16 lbxb6 cb 1 7 ll xb6 (Free
borough and Ranken) is said to
favour White.
c2) 13 ... J.b4 14 lla4 a5 1 5_4
lbxb4 1 6 llfa l (Unzicker) is
supposed to be ;!;.
I don't find variations c l and c2
very convincing. The assessments
could be questioned. And in c l
Compromised Defence
what about 1 6 ... aW? 1 7 lixa8
lbxc6 18 lb xg5 e7? White is
probably better, but it isn't an
easy position for either side. So
maybe the Compromised Defence
is playable after all, with 8 . . . tWe7
9 ltJ xc3 'ti'b4 1 0 J.xf7+ 'it>d8. But
White has another, quite recent
move JJ1 J.b2! After this Black
will have to improve upon 1 1 . . .
"t!fxb3 1 2 J.xb3 J.xc3 1 3 J.xc3
lbf6 14 lbg5 e7 15 e5 lbe8 16 f4
h6 1 7 lbf3 d8 1 8 lbh4 1 -0
Hartoch-Eslon, Netherlands-Sweden
1 976. It is doubtful if he can since
1 2 . . . lbf6 13 lbg5 llf8 14 lbd5 (
- EC (j} also puts him in a terrible
mess.
9
e5
't!Vg6
9 . . . lbxe5? loses to 1 0 lle l d6 l l
lbxe5 de 1 2 "fi'a4+.
10 lbxc3
lhge7
Others:
a) 10 ... b5 I I lb xb5 llb8 12 't!Ve3
ltJge7 1 3 'it'e2 'it'h5 1 4 J.a3
(Kolisch-Anderssen, London 1861)
and Black's position is very
difficult.
b) 1 0 . . . J.xc3 I I 'it'xc3 lbge7 1 2
lhg5 0-0 1 3 J.d3 (Larry Evans).
11
J.a3! (74)
The best way of maintaining the

pressure. 11 lbe2, as Anderssen
played against Zukertort ( 1 87 1
match), is none too convincing
after 1 1 . . . b5 12 J.d3 'i!Ve6.
It hasn't exactly been proved
that White wins from this position,
but nobody has been able to come
up with anything that looks
Compromised Defence 73
li.Jf5 1 6 li ae l + .i.xe 1 1 7 llxe l +
'i!?f8? 1 8 'irxc6! . However, he
subsequently queried the correct
ness of this explosive combination
lt.J5e7!
because of the move 17
'and it is difficult to see how White
can do more than draw' . A draw
could indeed result from 1 8 de
lib1 1 9 .i. xf7+ 'i!?xf7 20 't!rf4+ 1!rf6
2 1 1!t'c4+ 1!t'e6! 22 1!t'f4+ etc.
ll b 8 1 2 lt.Jd5 lt.J xd5 13
c) 1 1
.txd5 b5 14 e6 ! fe 1 5 .txc6 de 16
lt.Je5 1i'e4 17 'irg3 and Black's
defences are overstrained, e.g.
17 . . . lig8 18 't!fg5 b4 19 't!fh5 + g6
20 't!fxh7 lif8 2 1 lt.J xg6 .
d) 1 1 a6 1 2 li.Jd5 lt.Jxd5 1 3 i.xd5
b5 1 4 e6! fe 1 5 .txc6 de 16 ltJe5
@e4 17 't!Vg3 with the same attack
as in (c). Roikov-Orlov, Leningrad
1 968, concluded 1 7 . . . g6 18 llad 1
b4 1 9 life 1 't!ff5 20 ltJxc6 ba 2 1
1i'xa3 l-0 - because of 2 1 . . . 't!ff6 22
lidS+ 'it>f7 23 l hh8 'tlfxh8 24
1We7+ 'i!?g8 25 't!fd8+ 'i!?g7 26
't!fd4+ 'i!?g8 27 lt.Je7+ etc.
...
...
healthy for Black. Known variations

are:
0-0 (it has always seemed
a) 1 1
to me one of the wonders of chess
that Black manages to do this in
the main line of the Compromised
Defence!) 12 llad l lle8 1 3 .i.d3
@h5 14 lt.Je4 lt.Jxe5 15 lt.J xe5 'ti'xe5
1 6 .tb2 'ti'e6 1 7 @b5 is analysis by
Lasker. After 1 7 . . . i.b6 18 'irh5
Black's troubles should soon be
over, e.g. 1 8 . . h6 1 9 lt.J g5 '@xa2
20 .th7+ 'i!?f8 (20 . . . 'it>h8 2 1
'irxh6 ) 2 1 .i.b 1 @c4 (or 2 1 . . .
'irb3 22 lid3) 22 lt.Jh7+ 'it>g8 23
li.Jf6+ gf 24 '@xh6 and forces mate.
bS 1 2 ltJ xb5 llb8 is the
b) 11
least clear variation, but unlikely
to be much good since Black no
longer has much of a material
advantage to justify his suffering.
A simple and promising response
is 1 3 .txe7 xe7 ( 1 3 . . . lt.Jxe7? 1 4
lt.Jd6+) 1 4 'ira3+ .tb4 1 5 '@e3
with a strong attack against the
king in the centre ( lt.Jxc7 and ltJg5
are both threatened). Blackburne
won an 1 875 exhibition game with
13 't!ra4!? a6 14 ltJd6+!? cd 1 5 ed
...
...
...
Conclusion: The Compromised

Defence seems to be fatal for
Black.
Peel-back: Our next section will
cover 6 . . . d6, which is an
important alternative to 6 ... ed.
There are a few odds and ends to
be tidied up on the way. Alternatives
for Black at move 7 are dubious:
d3 8 't!fb3 't!ff6 9 e5 't!g6 10
a) 7
lie l (as in the ce1ebrated ' Ever
green' game A nderssen-Dufresne,
Berlin 1 85 1 ) or 8 . . . file? 9 i.a3 d6
...
Compromised Defence

10
e l intending e5 .
b) 7
lbf6 8 .i.a3 d6 9 e5! with a
strong attack.
c) 7 . . . d6 8 t!lb3 t!ff6 9 e5 ! de 10
e l with a strong attack again (as
in M orphy-Kipping, B irmingham
1 8 5 8).
However, in view ofthe adequacy
of the 7 . . . l:.ge7 line (Evans
Gambit 4) it is White who stands
in need of an alternative at move
7. I nteresting and little known is
7 1lb3!?, e.g. 7 . . . t!fe7 8 0-0 .i.b6
9 cd lb xd4 10 l:.xd4 .i.xd4 1 1 tt:'l c3
with a complicated position that
certainly offers the attacker prac
tical chances. If there is reason to
prefer 6 d4 to 6 0-0, this is a
possibility well worth considering.
..
Section 6
6 d4 d6
75
w
e4
1
2
3
4
5
lbf3
.i.c4
b4
c3
d4
e5
lb c6
.i.c5
.i. x b4
.i.a 5
d6 (75)
We already known about 7 0-0

in this position since it transposes
into Evans Gambit 1 or 2, with
Black choosing between 7 . . . .i.b6,
7 . . . .i.g4 and 7 . . . .i.d7. But the
whole point of playing 6 d4
instead of 6 0-0 is supposed to be
to avoid those Jines (and, in
particular, Lasker's 7 . . . .i.b6
defence). So the question is: Can
White do any better with 7 \lt'b3 ?
7
tlt'b3
First pl ayed in Morphy-Ayers,

Mobile 1 8 55 - though presu mably
Morphy did not play it in order to
avoid Las ker's Defence.
7
\lt'd 7
There are three other moves

that Black has tried. I don 't want
to get bogged down in the m, as
they are pretty clearly inferior but not simple. So in mere
summary form:
a) 7 . . . 'ti'e7?! appears to lose a
piece, but is a little tricky: 8 d5
liJd4 9 Wa4+ (9 .i.b5+, as Morphy
played , and 9 lb xd4 are both
weaker) and now:
a ! ) 9 . . .i.d7 1 0 'ti'xa5 l:.c2+ 1 1
d l lbxa l l 2 l:.a3 as White can
pick up the knight on a l at his
leisure.
a2) 9 . . \lt'd7! 10 Wxa5 b6 1 1
l:.xd4! ba 1 2 .i.b5 ed 1 3 .i.xd7+
i.xd7 14 cd Riello-del Pezzo,
Venice 1 950. That is not an easy
ending, but it ought to be better to
have doubled d-pawns than doubled
a-pawns.
b) 7 . . . l:. h6?! was tried in two
.
6 d4 d6 7_5

games by Bronstein (Sokolsky
Bronstein, Kiev 1 944 and Ragozin
Bronstein, USSR Ch 1 945), but
after 8 i.xh6 gh 9 i.xt7+ l&>f8 1 0
de 'f!/e7 1 1 i.d5 h e lost them both .
His own comment in 200 Open
Games is a p propriate: "Alas, the
open p<>Siti"n of the black king,
which had literally nowhere to
hide from the scorching rays of the
attack, meant that Black could
not take sufficient advantage of
the open f- and g-files."
c) 7
lbxd4 is less clear than 'a'
and 'b', although after 8 lbxd4 ed
9 i.xt7+ >f8 1 0 0-0 "@e7 1 1 i.c4
White has plenty of play. The
game A.R.B. Thomas- Unzicker,
Hastings 1 950- 1 , continued 1 1 . . .
lbf6 1 2 cd .!D xe4 1 3 'f!/f3+ .!Df6 1 4
lbc3 i.xc3 1 5 "@xc3 i.f5 1 6 l:Ie 1
't!fd7 1 7 i.g5 .!De4 1 8 nxe4 i.xe4
19 lii: e l d5 ( 1 9 . . . 't!tg4!? 20 f3 1rxg5
2 1 fe oo seems a better defence)
20 lii: xe4! de, when a win has been
claimed with 2 1 1i'b4+ instead of
2 1 'it'g3 , which led to a draw.
8
de (76)
...
76
B
Black has two lines of defence:
A 8
B 8
..
...
de
i.b6
A
8
9
de
0-0
An interesting possibility is
9 i.a3 i.b6 10 0-0 .!Da5 1 1 .!Dxe5 !?
lbxb3 1 2 ab 'f!/e6 1 3 i.xe6 i.xe6,

generally assessed as +.
9
i.b6
10
l:Id1
1!t'e7
And not 10
lba5? 1 1 i.xf7+!
'it>f8 12 't!tc2 and Black has only
the abject 12 . . . We7 since 12 . . .
't!t'xf7 1 3 l:Id8+ 'it>e7 1 4 i.g5+ 't>e6
15 1!t'd2 is curtains (Cafferty
and Harding). Black has to be
continually on his guard against
this sort of shot. For example,
in the last diagram, if 8 . . .
lbxe5? then 9 .!Dxe5 d e 1 0 i.xf7+
't!t'xf7 1 1 Wb5+.
11
a4 (77)
This move of Shaposhnikov's
has been taken to supersede:
a) 1 1 i.a3 t!Vf6 1 2 lbbd2 oo compare with variation B (7
i.d7) in Evans Gambit, section
2.
b) 1 1 li dS a6 1 2 .!Dbd2 1!t'f6 1 3 .!D fl
i.e6 1 4 i. g5 1!t'g6 =t= Lisitsin
Rabinovich, Leningrad 1 940. Here
Fine has contributed the murky
suggestion 14 lbg3 !? lbge7 1 5 i.g5
1Wg6 16 i.xe7! oo. But White's
attack does not look sufficient
after 14 . . . i.xd5 1 5 ed .!Da5 1 6
1!t'a4+ c6.
...
...
6 d4 d6
'76 Evans Gambit: Section 6

77
B
S epp, corres 1 959-60.

e) 1 1 . i.e6?! 1 2 a5 i.xc4 1 3 't!t'xc4
't!Vc5 ( 1 3 . . . i.c5 1 4 lidS i.d6 1 5
a6! ) 1 4 'W'fl ! lbxa5 1 5 lid5 'i!t'c4 1 6
liaxa5 Aronin. Note that
White picks the e-pawn up with
check.
f) 1 1 . lbh6 12 a5 i.xa5 1 3 i.a3
1!ff6 14 i.b5 i.d7 1 5 c4 (Keres) is
given as in many places, but is
far from clear after 1 5 . . 0-0-0!?,
e.g. 16 't!t'a4 a6 17 i. xc6 i.xc6 1 8
lixd8+ lixd8 1 9 'fi'xa5 lid I + 20
ltJ e l .txe4 + (Botterill and
Harding). Also after 16 .tel .i.b6
1 7 i.g5 1!t'g6 1 8 i.xd8 li xd8 Black
seems to have more than enough
for the exchange.
..
..
Black has several possibilities:

a) 11 ... aS 12 i.d5 i.g4?
(Shaposhnikov said 12 . . . i.c5 was
better, but Cafferty and Harding's
suggestion 1 3 1!fc4! threatening
i.xf7+ is strong) 13 11d3 i.e6 (if
13 . . 0-0-0 14 lbbd2 lb f6 15 lbc4
i.c5 1 6 li b l Larry Evans)
14 i.a3 1!f6 1 5 lbbd2 lbge7 1 6
lt) c4 Shaposhnikov-Veltmander,
RSFSR Ch 1 958 1 1 . a5 is an
undesirable concession because it
robs Black of . . . lba5 and . 0-0-0.
b) 11 ... ltJaS? 12 i.xf7+ 'fi'xf7 1 3
litd8+ r!Je7 1 4 i.g5+ lbf6 1 5
'W'xf7 -i- r!Jxf7 1 6 ll: xh8 or if 1 2
r!;f8 just 1 3 'fi'a2 lbf6 1 4 i.a3 c 5 1 5
i.d5 Shaposhnikov.
c) 1 1 ... i.c5 12 i.g5? ! lt) f6 1 3 i.d5
0-0 14 a5 a6 1 5 lbbd2 lbd8 +
Yurkov-Bykhovksy, Moscow Ch
1 963 . White's 1 2th had little point.
12 a5 or 12 i.a3 should be tried.
d) 1 1 . . a6 1 2 a5 i.c5 1 3 i.a3 i.xa3
14 lt)xa3 lt)f6 1 5 i.d5 lt)d8? (Euwe
suggested 1 5 . . . 0-0 1 6 lt)c4 i.d7 1 7
lbe3 lbd8, when White does not
appear to have much for his pawn)
16 lt)c4 0-0 1 7 lt)cxe5 Soko1ov.
..
...
B
8
i.b6
A move motivated by the same
sort of common-sensical defensive
considerations that inspired the
Lasker Defence . Black does not
mind too much if White wins
back the pawn, so long as he can
neutralize White's attacking for
mation with . . . lba5. The virtues
of this approach are shown by the
variation 9 0-0 lba5 10 'it'b4 lbxc4
1 1 't!Vxc4 de 12 lbxe5 'fi'e6 when
Black has a stable advantage in
view of his bishop pair and
sounder pawn structure.
9
lbbd2!?
An interesting and imaginative

piece of analysis by Cafferty
seemed to make this White's best
try, but a recent Black innovation
at move 1 1 forces us to think
6 d4 d6 77
again. The other options are:

a) 9 ed ll:l a5 10 "it'b4 ll:lxc4 1 1 't!fxc4
"it'xd6 1 2 JLa3 iLe6! and now:
a l ) 13 't!fb5+ 'i!fd7 14 1!t'b4 c5 1 5
't!fb2 'id3 ! 1 6 ll:l bd2 0-0-0 =F
Mnatsakanian-Korelov, USSR Ch
1 963.
a2) 13 'ie2 't!rc6 ( 1 3 . . . 'ii'd 7 ! +
Unzicker) 1 4 ll:ld4 'ii'c4 1 5 l0xe6
'iWxe2+ I 6 'it>xe2 fe = Pfleger
Unzicker, Bamberg 1 963.
b) 9 1Wc2 de 10 iLa3 ll:l ge 7 l l
ll:lbd2 0-0 I 2 iLb3 'it>h8 1 3 0-0
f6 + Ragozin-Mikenas, USSR
1956.
c) 9 iLb5 and now:
c l ) 9 . ll:l ge7 10 0-0 'it'g4 !? I I ed
cd 12 iLa3 JLe6 I 3 'ii'a4 0-0-0 oo
Cafferty-Marriott, corres 1 956-7.
c2) 9 . a6 10 iLa4 \!t'e6 l l
JLxc6+!? (why? l i 'ifxe6+ !? JLxe6
12 ed looks right to me) l i . . . be I 2
0-0 li b8 and now Chandon-Moet
v Bottlik, corres I 974, continued
13 'it>h i ?! de 14 'ii'x e6+ JLxe6 1 5
ll:lxe5 ll:le7 1 6 iLa3 f6 + . Bottlik
suggests that White would do
better with 13 't!Yxe6+ iLxe6 1 4
iLa3 / l 3 . . . fe 1 4 e d cd I 5 JLa3.
Of these alternatives only 'c'
holds out any prospects for White.
78
B
Cafferty-Clarke, British Ch 1 970,

though that is also quite good)
14 . . . JLd7 ( 1 4 . . . lt:ld7 15 lid5 ! ) 1 5
't!t'b3 (intending f4-f5 , and i f 1 5 . . .
0-0-0? 1 6 l0 xf7) 1 5 . . . lid8 (79)
..
..
ll:la5
ll:lxc4
1!Vc2
ll:l xc4 (78)
9
10
11
Fro m here we move

Cafferty's analysis with:
into
1 l ... de 12 l0fxe5 'ii'e6 13 JLa3!
and now:
a) 13 . l0f6 14 0-0-0 ! (an
improvement on the 1 4 f4 of
.
1 6 lid6 ! cd 1 7 lt:lxd6+ 'it;>f8 1 8

ll:ldxf7+ and White either wins
material through 1 8 . . . 'it;>g8 1 9
l0xd8 JLxd8 2 0 litd2 or drives the
attack home after 1 8 . . . 'it>e8 1 9
lt:ld6+ 'it;>f8 2 0 lt:l xd7+ 'ii'x d7 2 1
ll:lf5+ 'it;>e8 22 ll:lxg7+ 1Wxg7 23
'ii'e 6+ and mate next move.
b) 1 3 . . . iLd7 1 4 'ii'b3 !iJe7 15 !iJxd7
1lt'xd7 16 l0e5 'ii'e 6 1 7 1Wxe6 fe 1 8
0-0-0. White threatens 1 9 litd7 and
must surely have the advantage
because of Black's lazy rooks.

c) 1 3 ... li:Je7 is the safest, though
after 14 '@'a4+ .i.d7 1 5 li:Jxd7
'i!fxd7 1 6 l0xb6 cb 1 7 '@'xd7+
xd7 18 0-0-0+ e8 19 lld3
White probably has ;! in the
ending.
So 9 li:Jbd2 gives White good
chances? Not so! On his way to
becoming world correspondence
champion Palciauskas produced
d5!
the telling improvement 1 1
...
12 l0xb6 ab 1 3 0-0 de 14 'it'xe4

'i!Vg4 (80)
80
w
White is left with a pawn on e5

rather than e4 and this is clearly in
Black's favour since it gives him
complete control of the light
squares. Meanwhile White's a
pawn is a long-term weakness and
the fact that his bishop cannot get
to a3 frustrates his attacking
desires. Estrin-Palciauskas , l Oth
World Corres Final, continued:
15 1We3 l0e7 16 li:Jd4 0-0 1 7 h3 1!t'g6
18 f4 c5! 19 li:Jb5 lidS 20 l0d6 li:Jf5
21 l0xc8 ll axc8 + 22 't!t'f3 h5 23
'it'xb7 ll c6 24 't!fa6 c4 25 'it'a4 lld3
26 llb1 llxh3 27 e6 (27 't!t'xc6 1kxc6
28 gh l0g3 H) 27 . . llxe6 28
.
6 d4 d6
't!fxc4 'i!Vg3 0- l . Black threatens
29 . . . 1!t'h2+ 30 'it>f2 ll f3+ 3 1 xf3
'i!fg3 mate and if 29 't!fd5 lle2
forces mate. A very impressive
game. Perhaps White need not
lose quite so quickly (e.g. he could
exchange queens at move 1 5). But
in the long run he seems powerless
to combat Black's strategic idea,
which essentially consists in going,
not for the two bishops, but for
the better bishop or good knight
v bad bishop.
Conclusion: To take up the question
posed at the beginning of this
section - can White profit from
avoiding the Lasker Defence with
7 't!fb3? - it seems doubtful
whether he can, given the latest
state of the art in variation B (8 . . .
.i.b6). Variation A really has no
business to be good for Black
because of one general consideration:
in our main line of Evans Gambit
1 he could get precisely the same
position as occurs after Black's
9th in A by playing 9 . . . 'i!Vd7. But
with that move order it would
look a silly move to play and
nobody has ever suggested that it
m ight be better there than the
more natural 9 . . . 'i!ff6. Yet our
resultant in A is far from clearly
good for White, and the subsequent
variations c, d and f may indeed
be good for Black. This is quite a
puzzle. I don't know the solution,
but it seems reasonable to suggest
that the fault lies in the Shaposh-

nikov plan 1 0 Ii:d 1 and 1 1 a4.
White should give some t h ough t
to the alternatives 10 .ib5 !? and
10 .idS!?. 10 .ib5 !? is o ften
quickly dismissed with the refer
ence lO . . . 'it'e6 1 1 1Wxe6+ .ixe6
12 .ixc6+ be 13 l0xe5 l0e7 +
Diihrssen-Keres, corres 1935. This,
you wi l l note, is very similar to
note c2 to White's 9th in B . So
probably l O .ib5 !? is not good
enough. But 10 .idS!? h as been
strangely neglected, for such a
normal Evans move. It is, at any

rate, the only thing I can find to
reco mmend.
Peel-back: We have now gone
through all the most i m po r tan t
variations after 5 . . . .ia5, with
White playing either 6 0-0 (sections
1 -3) or 6 d4 ( sections 4-6) . The
next section is a casual look at a
little known third option for
White - 6 't!t'b3 .
Section 7
6 'tlfb3
1
2
3
4
5
6
e4
lbf3
.ic4
b4
c3
'4!t'b3 (81)
e5
lt:lc6
.ic5
.ixb4
.ia5
To be frank, there isn't much to

say here, except that 6 't!fb3 is
rarely played and nobody knows
much about it. Therein lies its
6 d4 d6 79
81
B
charm ! At a very general level,

there are some things going for the
move: it obliges Black to make an
early and difficult decision as to
how to defend f7, inhibits the
sometimes useful riposte . . . d5,
and above all takes the opponent
out of prepared defences (such as
Evans Gambit 1 , 2 or 4). The loss
of flexibility involved in committing
the queen to b3 is, however, a
serious drawback. 0-0 and d4 are
moves that White is going to play
come what may, but the queen
might be better posted on a4 or d l
or so me other square.
A rigorously analytical approach
hardly seems appropriate . Instead
let's follow the one grandmaster
game that has featured 6 't!Vb3 ,
namely Nunn-H iibner, Johannes
burg 1 98 1 :
6
't!ff6
Cafferty and Harding speak

well of 6 ... l0h6 7 d4 0-0 8 .ixh6
gh, but the shattering of Black's
kingside pawns must offer White
some compensation. There is also
6 .. . 1le7 with just a few examples:
'"
I
13 d6 8 d4 ed 9 0-0 i.b6 ,
8A nderssen , Manchester
seems good for Black
1
.
'"- ua e of 1 0 e5 lDa5 ! .
b ) 1 d4 and now:
b l ) 1 ... ll:l xd4 8 lD x d4 ed 9 0-0
.l b6 IO e5 ! with good attacking
chances ( Wills-Jones, corres 1969).
b2) 7 ... .tb6 8 lDxe5 lDxe5 9 de d6
I 0 0-0 d e I I .ta3 "irf6 1 2 lDd2 oo
Wills- H opewe ll, corres 1 969-70.
7
8
.tb6
d6
Much safer than 8 . .. ed 9 e5
'i!t'g6 I O c d lDxd4 I I lDxd4 .txd4
0-0
d4
12 lD c 3 lbh6 (Bird-Chigorin,
Hast ings I 895) 1 3 .ta3 ! .
c!'b as seems a reasonable
8
alterna tiv e, however, and if 9 de
1i'g6 co.
de
In lnformator Filip presents us

with th e heroic idea of 9 lDg5 ! ?
lDh6 I O f4 oo . Easier said than
done ! A ny volunteers for this
missio n fr om which few survivors
can be ex pe cted to return?
9
10
II
82
w
6 'W'bJ
Gambit: Section 7
lDxeS
a4
lD xeS
de
a6 (82)
12
h 1
White prepares f4, a n ambitious

plan. A possible improvement is
12 aS .ta7 ( 1 2 . . . .tc5 I 3 .te3 ! ) 1 3
lD d 2 o r only then I 3 h i
intending f4.
12
13
f4
14
aS
lDe7
.te6
No choice, as I 3 . . . 0-0 I4 fe
't!Vg6 I S .ta3 is very strong indeed.
Black would be very happy with
14 .t xe6 1!fxe6 1 5 't!Vxe6 fe 1 6 fe
lbg6 +.
14
IS
16
17
't!Vxb7
fe
lDd2
.tcs
0-0
't!tg6
White has even contrived to win

a pawn. But really it is only a loan,
since Black can enforce a recapture
on e 5 . In the meantime Black has
taken the lead in development,
and White has to confess that
things are not going too well by
conceding the option of a draw by
repetition ( 1 7 .. . lUb8 I 8 't!Vxc7
lk8 1 9 't!Vb 7 etc).
17
18
19
20
21
22
lD xc4
\!Vb3
1!fa4
.tf4
lladl
.txc4
"i!ke6
ll ab8
lDc6
ll bS
.ta7
The pawn is bound to fall now.

23
h3
h6
Subtle play to gain a tempo. If

at once 23 . . . llc5 the reply would
be 24 lld5 - but what else can
White do anyway?

24
25
li dS
ed
lixd5
1!txd5 (83)
It is clear that Black is better.

H aving got the one pawn back he
can turn his attention to the other
weaklings on e5, aS and c3.
N unn 's next move just loses the e5
pawn. 26 lldl 'i!e4 27 .tg3 ttJxa5
drops one of the others. Best was
26 i.g3, but 26 . . . llb8 is still + .
26
27
28
29
30
ltJ e3?
.txe3
.tf4
i. xe5
1!Vc4
.txe3
ltlxe5
lidS
"t!Vxe5
lidS
White's situation deteriorates

into the tournament player's
nightmare - a game that is pretty
clearly lost, but not decently
resignable. I give the rest of the
game without comments, even
though it is a grim business. In
fact, I give it precisely because it is
a grim business, as a reminder that
this sort of finish is just as
common as an attacking victory:
31 1!Vxa6 llxa5 32 \!fc4 lidS 33
\!fb3 g6 34 11Vc4 <J;g7 35 \!fb3 'i!e4
6 '@b3 81
36 'Wa2 1!t'd3 37 lif3 We4 38 lUI

h5 39 1!fb3 1!Vd3 40 g l lid7 41
'ti'b4 1!te3+ 42 h 1 l:td5 43 1!fb3
'ti'd3 44 Wgl c5 45 "t!rb7 'i!e3+ 46
h l 1!t'e6 47 l!fc7 lig5 48 1Wf4 llf5
49 We i lle5 50 "it'd l lte2 5 1 "t!rd3
'i!e4 52 Wg3 lle3 53 9f2 f5 54 c4
llxh3+ 55 g l lle3 56 lid l lie2
57 1!Ng3 \ie5 58 \ih4 h6 0- 1 .
A fter such a game I would not
think that Nunn would wish to
have his name associated with the
variation. We could call it the
Wills Variation, after the corre
spondence player who has indulged
in it most frequently. Or maybe,
if it is left to slumber for a few
more years, we should call it the
Kipping Variation.
Conclusion: Who knows? It doesn't
seem a particularly promising line
for White, but could be used if you
wish to take a bookish opponent
into uncharted territory.
Peel-back: Sections I -7 have been
devoted to 5 . . . .ta5 defences.
Let's see what happens if Black
puts the bishop on c5.
Section 8
. . .ic5 and the
' N ormal Position
1
2
3
4
5
e4
l!)f3
i. c4
b4
c3
e5
ltJc6
.tcs
.txb4
.tcS (84)
5.
. .
.i.c5
example, 1 4 . . . lle8 1 5 f4 g8 1 6
b2 d5 1 7 c4! Diihrssen
Kramer, Ebensee 1 930, or 14 .
't!rh4 1 5 f4 ll hf8 1 6 .liJd2 g8 1 7
.ltJ f3 1!ff6 1 8 .i.e3 Sokolsky
Kopayev, C hernovitsi 1 946.
.
cd
.tb6 (85)
85
w
I sh all deal rather briefly with

this old-fashioned defensive line.
So far as I can see, it has no par
ticular advantage over 5 . . . .i.a5 ,
whereas th ere i s the significant
disadvantage that after d4 Black is
obliged to play . .. ed, ceding
control and allowing White to rid
himself of a potentially weak
pawn on c3.
6
1
d4
0-0
ed
White could try 1 cd .i. b4+ 8

.i.d2 or 8 \!lfl !? , but there is no
reason why he should.
1
d6
Other moves (apart from 7 .

.i.b6 8 cd d6 transposing) are
dubious:
a) 1
de 8 .i.xf7+ xf7 9 '@d5+
\!lf8 1 0 't!t'xc5+ 'tle7 I I 't!t'xc3 and
White's better development and
safer king more than compensate
for the pawn.
b) 1 d3 8 .ltJg5 .ltJh6 9 .ltJxf7 .ltJ xf7
l O .i.xf7 + \!lxf7 1 1 'lrh5+ g6 1 2
1txc5 d6 1 3 't!t'd5+ .i.e6 1 4 Wxd3
gives White all the benefits of a
gambit at no material cost. For
.
...
..
Thus w e arrive at the Normal

Position - so called because for
much o f the 1 9th century it was
what you c ould expect to get when
playing the Evans Gambit. White
has tried three moves:
A 9 dS
B 9 lbc3
c 9 h3
A
9
dS
This was how Anderssen used

to play it, but I think it is a
positional mistake that really
deserves a question mark. Every
British schoolboy should know
that two abreast' is the ideal
central formation. You should not
voluntarily give it up unless and
until some concrete gain results.
The net result here is that Black
5 . c5 83

can now set to work on exploiting
his queenside majority with . . . c5
and . . . b5. Admittedly, the position
is double-edged and White has his
practical chances (swindles), but
with best play Black ought to
prevail.
ltJaS
9
ltJe7
.i.b2
10
11
.i.d3
Anderssen pointed out that

1 1 i.xg7?! I!g8 12 i.f6 ltJxc4 1 3
\!t'a4+ 1Vd7 1 4 '@xc4 1oses t o 1 4 . . .
lhg2+! 1 5 'Ot>xg2 'ikg4+ 1 6 ..t>h l
1Vx0+ 1 7 g l i.h3 .
11
12
13
0-0
ltJc3
ltJe2
ltJg6
c5 (86)
86
w
Understanding of this position

evolved through a series of games
between A nderssen and Zukertort
in the 1 860s, in which it finally
emerged that Black's best plan
consisted in . . . f6, . . . i.c7, . . . I! b8,
... b5 and then ... c4 ! (but not first
. . . b4, when the advance gets
bogged down in the blockading
mire c4). The Breslau professor
. .
was able to sum up the findings of

the seminars with his promising
student like this: 14 '@d2 i.c7 1 5
li[ac l Itb8 1 6 ltJg3 f6 1 7 ltJf5 b5 1 8
h i c4! (Zukertort's thesis: if
19 i.e2 b4 anyway, because of 20
i.xc4 ltJxc4 2 1 I!xc4 .ia6) 1 9 b1
b4 20 .id4 i.a6 2 1 't!M l c3 +
W. Paulsen - Anderssen, Barmen
1 869.
B
9
lL!c3
Morphy's move, frequently em

ployed by C higorin, and usually
given as best.
9
ltJaS
It is rather curious that this
move works out as well as it does.
On principle, Black ought to
develop an unmoved piece:
a) 9
.ig4 10 i.b5 f8 !? ( 1 0 ...
.i.d7 1 1 e5 ! lL!ge7 12 .ig5 is
attractive for White) 1 1 .ie3 lL!ge7
1 2 a4 aS 1 3 .ic4 .ih5 14 lilc l ! (an
improvement on 1 4 'Ot>h 1 of
Chigorin-Gunsberg, Hastings 1 895)
14 . . . 'Wd7 1 5 .ib5 1Vc8 16 .ie2
and White had a beautiful position
for his pawn in Cafferty-Cadden,
corres 1 967-8.
ltJf6 is best answered by
b) 9
10 h3, with play as in C: 9 h3. The
immediate assault 1 0 e5 !? de 1 1
.ia3 is an untrustworthy oo after
1 1 . . . ltJa5 !
.ig5 (87)
10
The bishop on c4 is of course
indirectly protected ( 'Wa4+), so
White forces Black to make a
decision ...
..
5 . i.c5

87
8
play ! The attack triumphed in

Chigorin-Gunsberg, 9th match
ga me, Havana 1 890: 12
llJ ac6?
13 i.xe7 llJ xe7 14 li:J g5+ g6
( 1 4 . . . Wg8 1 5 \!Vb3 ! ++; 14 . . . wes
15 'it'h5+ Wd7 16 llJf7 'ti'f8 1 7
llJxh8 'ti'xh8 1 8 'i!kf7 ) 1 5 li:Jf4+!
wf6 ( 1 5 . . . wxg5 16 'ti'h5+ wf6 1 7
e5+ mates) 1 6 e 5 + 1 -0, presumably
because of 16 . . . de 1 7 de+ wf5 1 8
'ti'f3 . However, Chigorin
showed that Black can hold the
position with 12 . lieS! 1 3 !i.. x e7
.ll x e7 14 li:Jg5+ Wg8 1 5 'ti'h5 h6
( 1 5 . . . g6 16 llJf6+ Wg7 1 7 'ti'xh7+
wxf6 1 8 e5+ wxg5 19 f4+ wf5 20
'ti'h3+ we4 21 'ti'f3 + Wf5 is
another variant on the theme of
perpetual check) 1 6 'it'g6 hg 1 7
llJf6+ Wf8 1 8 llJh7+ and draws.
82
f6
10
11
!i.. f4 (89)
...
between:
81 10 . .. li:Je7
82 10 ... f6
83 1 0
'ti'd7
..
81
li:J e7
10
11
i.xf7+
This only draws , but then the
same goes for l l llJd5 f6 ( I I . . .
llJxc4?, however, loses to 1 2 i.xe7
'it'd7 1 3 i.f6! 0-0 1 4 'it'c l intending
'it'g5 , \!Vxc4) 12 i.xf6 gf 13 llJ xf6+
f8 1 4 llJg5 li:J xc4 1 5 'it'h5 Wg7 1 6
1!ff7+ h6 and White has t o settle
for a perpetual check with 1 7
'it'h5+.
11
wxf7
12
llJd5 (88)
..
89
B
88
B
The perfect paradigm of gambit
1 1 i.h4 and 1 1 i. e3 are also

playable, but not all that different.
Pollock-Chigorin, H astings 1 895,
was a n exciting attempt to make
something of the action of the
bishop in the h4-d8 diagonal:
1 1 !i.. h 4 llJe7 12 li e l ! ? i.g4 1 3 e5!

de 14 de t!t'xd l 1 5 l:i:axd l lt:lxc4 1 6
ef gf 1 7 i.xf6 'Ot>f7 ! 1 8 i.xe 7 i.xf3
1 9 gf .ta5 20 l:i:d7? (20 l:i:e4! is
better) 20 . . . ltlb6 21 l:i:xc7 l:i:fc8 ! =F.
lt:le7!?
11
This move seems to be a sig
nificant improvement upon 1 1 . . .
ttJxc4 1 2 a4+ t!t'd7 1 3 t!t'xc4 1!ff7
1 4 lt:ld5 when White's advantage
i n space and development m ore
than compensates for the gambit
pawn, e.g:
a) 14 . g5? 1 5 ii:.g3 i.e6 (Steinitz
had been under the misapprehension
that he could play 1 5 . . . h 5 , but 1 6
ttJxb6 ab 1 7 t!t'xf7+ 'Ct>xf7 1 8 l:i: fc l
saves the bishop o n g3 b y win n i ng
a pawn) 1 6 t!t'a4+ i.d7 1 7 "t!t'a3
l:i:c8 1 8 l:i:fe l Chigorin-Steinitz,
London 1 88 3 .
b) 14
i. e6 1 5 a4+ ii:. d 7 1 6 1!t"c2
lic8 17 a4 i.a5 1 8 l:i:fb 1
Chigorin-Dorrer, telegraph game
1 884.
c6
12
h3
Perhaps White should try 1 3
ii:.d3 now, with the idea o f keeping
the knight on a5 out of the game.
13 i.b3 lt:l g6 14 i.g3 e7 15 l:i: e l
lbxb3 1 6 1!fxb3 ii:.e6, as in
Asharin-C higorin, R iga 1 892, is +.
It is revealing that Chigorin, after
some successful performances with
White, was willing to play the
black side of this line.
B3
d7
10
This has the merit of actually
threatening 1 1 . . . lt:lxc4. However,
it does serious harm by obstructing
..
..
5 . . .tc5 85
.
Black's own lines of communica

tion, e.g. 1 1 i.d3 ttJ e7 1 2 ti:ld5!
ti:lxd5 ( 12 . . . tt::l g 6!?) 1 3 ed 0-0 1 4
l:i: e l lle8 1 5 1Wc2 .
c
9
h3
like t h is move. Instead of
doing something frantic, White
decides that he will preserve all the
advantages of his position and
starts by preventing the annoying
. . . i.g4. There a re many cases in
which the most effective way of
handling a gambit is just t o
strengthen your position in a quiet
and methodical way, rather than
forci ng an immediate crisis.
9
lt:lf6
10
l:i:e1
h6
11
i.a3
0-0
12
lt:l c3
l:i:e8
13
l:i:cl (90)
White has a lovely, flexible

development whereas Black's game
is so cramped and uncomfortable
that it is a problem how he is to
connect his rooks. Mariotti-Gligoric,
Venice 1 97 1 , continued 1 3 ... tt::l h 7
1 4 l:i:e3 ltl a 5 1 5 i.d3 i.e6 1 6 t!t'e2
ltlf8 ( 1 6 . . . c5 ! seems much better)
5.
1 7 4ll41 16 18 ltlxb6 ab 19 dS
Ad? 20 .i.bl f4 21 1We2 e6 22
Wc3 16 23 de be 24 .tn eS 25 llJ h4
dS?I (better 25
1We7 - M ariotti)
26 ed IOxdS 27 l he8+ 'irxe8 28
1Wa3 1We4 29 lld1 ll:e8 30 ..id3
1Wa4 31 .ih7+ 'it>xh7 32 llxd5 ll:e7
33 .txf6 ! 'ti'e4 (91) (33 . gf 34
1Wg6+ 'it>h8 35 'it'xf6+ ll:g7 36 ltl f5
)
. . .
. .
.tc5
played. In vanatwn C we have

only an illustrative game, rather
than a theoretical analysis . That is
quite appropriate. White should
just smoothly complete his develop
ment. Let Black worry about how
he is going to get his pieces out!
. .
91
w
Peel-back: Of respectable moves

for Black only 5 . . .te7 remains,
so that is the next section . There
are also two bizarre retreats: 5 . . .
..id6 (once played b y Pillsbury)
and 5 . . ..if8 (once played by
Steinitz). There would be little
point to a detailed analysis. If
Black does that at move 5 , who
knows what his 6, 7, etc. might be?
In either case White replies 6 d4,
with an excellent game.
.
34 lLlg6! 'ft'xg6 (34 . . gf 35 llJxe7

\!t'xe7 36 'ti'd3+; 34 . ll:f7 35
lLlf8+ 'it>g8 36 llxd7 ; 3 4 .
1!fxd5 35 ltlxe7 1!Yf7 36 '@d3+ h8
3 7 'ft'xd7 gf 3 8 \!t'd8+ 'it>g7 39
lLJf5+ 'it>g6 40 g4 ) 35 .txe7 ..ie6
36 ll:d6 1!Vt7 37 lhb6 l0c4 (37 . .
Wxe7 38 't!re5) 38 l:tb7 ..tfS 39 ll:c7
.ig6 40 .txe5 'ti'f6 41 ..ib4 1 -0.
.
. .
Sectio n 9
5
. . .
An encouraging victory for the

Evans Gambit, showing that it is
still capable of striking down
high-class opposition.
Conclusion: If White gets to the

Normal Position he should not
rush things out of anxiety to prove
that he has enough for the pawn .
Variation B2 needs to be improved
from White's side if 9 lt::l c 3 is to be
92
w
!/Le7
1
2
3
4
5
e4
llJf3
..te4
b4
c3
eS
ltle6
..ic5
..i xb4
..ie7 (92)
What could be more rational
5 . . e7 8 7

than this modest move ? Hartston
has commented: " Most grand
masters consider this the most
sensible response. Black avoids all
the dangers of the old lines . . . ,
keeping instead a safe position,
ultimately gaining the bishop pair
and no structural defects."
True, B lack avoids any of
the minor piece entanglements
associated with 5 . . . a5 (White's
'it'a4 and d5 thrusts) and does not
put it on c5 to be prodded again by
d4. I ns tead, with the gambit pawn
under its cassock, the bishop says
.. mission accomplished" and returns
to base. Yet this is not a
particularly challenging line against
the Evans, for the bishop on e7 is
not playing a very active role.
Indeed it obstructs the defence of
f7 , a factor which more or less
obliges Black to return the pawn
straightaway.
White can choose between:
A 6 d4
B 6 't!i'b3
A
this book we see many instances in

which this i s not a serious blow to
Black's chances.
93
B
7
8
9
IO
lt::l a5
'ifb3
1Jf8
xf7+
1Jxn
'ti'a4
't!t'xa5 (93)
To be frank, this position has

received little serious attention,
the moves cited s o far customa rily
serving as a p retext for passing on
fro m 6 . ed to something more
respectable. I t is not hard to
believe that 1 0 . d6 1 1 cd (as in
Lehmann-Do oner, Munich 1 958)
favours White - big centre, safer
king, better development, n o
material cost. B u t I wonder how
bad for B lack 10 . . de (already
condemned in the Handbuch)
1 1 lt::l xc3 really is.
. .
d4
And now there are three

responses that deserve consideration:
AI 6 . ed
A2 6 ... d6
A3 6 ... lLl a5!
A2
6
7
..
AI
6
ed
The trouble with this is that the

f7 pawn now falls by force. Yet in
de
d6
lt::l x e5
If this is not good, at least it

takes some marvellous play t o
show why. T h e inadequacy o f
alternatives is easier t o detect:
a) 7 . lt::l aS 8 xf7+ 1Jxf7 9
9xd5+ e6 1 0 1Wxa5 .
.
5.
b) 7 . de 8 'llb 3 l0a5 9 .txt7+ f8
1 0 'lra4 m u st su rely be good for

White since here the displacement
of Black's king is aggravated by a
weak e-pawn and an open d-file.
de (94)
8
lDxeS
..
94
... ;. . lj)
& &
. .
X.
- 7.
- 7. .
-
w & -
.-
7.

i.
D
l'flw.fi 0
.
, 1,
-i]
1<!!
' '%\Ult
.
o;:;o
- at:a.
'
;r;-..
z
g
"" "' '-ZJ n 0 13'
"'"/.
.-.
o%
...
.....
I<'
;
Z, .
After 1 1
1 3 lt:J xa3
14 .i xf7+
1 6 f3 .
12
13
14
.ie6 1 2 .txe6 .ixa3

'ii' e 7 White can play
xt7 I 5 'ii'x e7+ xe7
...
0-0
'i!t'g3
ltlxa3
ltl g4
.ixa3
'it'e7! (95)
'
' .i '-
. - D.
.
. "II"'%'-g
' ..
m
1.
Here we have the privilege of

following a game by Fischer
which he himself annotated in My
60 Memorable Games.
9 1!t'h5!
Fischer explains that in another
exhibition game he played the
routine 9 'irb3 but did not gain
any advantage after 9 . . . .ie6 ! 1 0
.ixe6 fe 1 1 .ia3 !? ( I I 'i!t'xe6
'it'd6 ) I 1 . . . 'i!t'd3 ! Black also
seems to be quite all right, if not
better, after 1 1 'it'xb7 liJ f6 1 2
'it'c6+ <$;f7.
g6
9
ltlf6
1 0 'it'xeS
Fischer points out that White
wins after 10 . . . f6 1 1 1i'b5 + c6? 1 2
1i'b3 f8 1 3 .ixg8 ! (intending
.ih6+) . Although I I . . Wf8 is
better, it does not redeem the line
as 1 2 "t!t'b3 g7 1 3 0-0 leaves Black
with a very difficult game.
11
.i a3
l:U8
.ie7
95 . ...
w
l.
1.
&
.
.
a .
,.
. . . .
V.
. .
....7.
Black has some resources:

1 5 ltlc2 'i!t'e5 or 1 5 ltlb5 ltle5. But
the great man found the solution
and blasted B lack off t he board:
15 .i bS+ ! c6 ( 1 5 . . . .id7 1 6 "ii'x c7 ! )
16 ltlc4 "ii'e6 ( Fischer showed that
1 6 . . . cb 1 7 liJ d6+ ..t>d8 I 8 IUd 1
.id7 1 9 ltl xb7+ <$;c8 2 0 ltld6+
..t>d8 2 1 Ild4! ltle5 22 Il ad l gives
White a winning attack) 1 7 Ilad 1
c b 1 8 "ii'c 7 .id7 1 9 ltl d6+ <tle7 (96)
96
20 liJfS+! gf 21 ef Ilac8 (2 I
'itxf5 22 'it'd6+ e8 23 li fe I +) 22

l::txd 7+! 1!t'xd7 23 f6+ tbxf6 24
lie I + tbe4 25 lixe4+ f6 2 6
'i!t'xd7 lit fd8 2 7 Wg4 1 -0 Fischer
Celle , California 1 964. Glorious !
A3
6
llJaS!
The approved move. Black
returns the pawn in order t o
eliminate the menace o n c 4 and
achieve . . . d5.
7 li:l xe5
White has tried two other
moves, but I would advise against
them:
a) 7 .td3!? e d !? (7 . . . d6 8 de de 9
li:lxe5 li:lf6 1 0 0-0 0-0 = is the safety
play) 8 cd d5 9 \!t'a4+ llJ c6 10 ed
'irxd5 1 1 llJc3 .tb4! 12 i.d2 i.xc3
13 .txc3 li:lge7 +.
b) 7 .txf7+ ?! cannot possibly b e
sound in spite of 7 . . . xf7 8
li:lxe5+ <&>f8 9 'it'f3+ llJ f6 1 0 g4 d6
1 1 g5 de 12 gf i.xf6 13 de llJ c6 1 4
ef \!t'xf6 Inkiov-Minev, B u l garia
1 977. 8
e8 1 s one of the
refutations.
7
llJ xc4
8 tb xc4
dS
8 . . . d6 is of course playable, but
concedes an enduring spatial
advantage.
'tVxdS
9
ed
10
tbe3 (97)
The assessment of this position
has to hover between
and t. I n
practice White has done rather
well. It is true that Black has the
two bishops, but White's one
bishop repeatedly threatens to
outshine them by means of i.b2,
=
...
5 . . J.e7 89
.
c4 and d 5 .
97
B
Here are some examples of

further play:
a) 1 0
"it"d8 1 1 0-0 li:lf6 12 c4 0-0
1 3 tbc3 c6 1 4 i.b2 1!i'a5 1 5 d5 i.a3
1 6 i.xa3 "it"xa3 17 "t!t'b3 ! ;!;
Tartakower-Trifunovic, Paris 1 950.
b) 10
\!VaS 1 1 0-0 llJf6 12 c4
b1) 12
c6 1 3 i.b2 .te6 14 li:lc3
li d 8 1 5 d 5 ! cd 1 6 li:l cxd5 li:lxd5 1 7
c d 0-0 ( 1 7 . . . J. xd5 1 8 1!Yh5 ! ) 1 8
'ti'f3 i.xd5?? ( 1 8 ... J.c8 1 9 llfd 1 )
1 9 "t!t'g3 Cafferty-van Geet,
A msterdam 1 972.
b2) 12
0-0 13 d5 ! b5 1 4 li:ld2 be
1 5 llJ d xc4 9a6 1 6 i.b2 lie8
(Minev suggests 16 . . . li:lg4!?, to
w h ich I think 1 7 ll c l is still a good
reply) 1 7 li c l litb8 1 8 .te5 ! lib7
1 9 li e 1 i.d7 (98)
...
...
...
...
98
w
10 ..,., Otlmblt: Stction 9
5 . . . J..e 7
.! ,
, . 4tt od 2 1 .lxf6 gf 22 ll:ld5 .i.e6

. . ll lo3 *hi (23 . . . A xd5 24 llg3+
IU 25 a4 : ) 24 ll:l xe7 llexe7
25 d4 *a' 2 6 llgH ..&f8 21
Wxf'6 *e8 28 ll:le5! 1-0 Timman
Tatai , A msterdam 1 977. If 2 8 . . .
de just 2 9 ll d 1 'ffd 6 30 'ti'h8+
wins.
c) 1 0 ... 9d7 1 1 0-0 ltlf6 1 2 c4 0-0
(afterwards the players agreed

that 12 .. . b5 ! would have been
better, justifying Black's 1 Oth)
lJ ltlc3 c6 1 4 d5! (a good move as
usual) 14 . . . cd 1 5 ltlcxd5 lLl xdS 1 6
ll:lxd5 .i.d8 1 7 ll b 1 'Wc6 1 8 't!rd4 ;!;
Jt:l u nn- L arse n , London 1 980.
B
6
99
w
1fb3
d4
ltlh6
lLl aS (99)
Until the B ulgarian correspond

ence master G .Popo v came up with
\!rbS!
there was cause to regard 6 1!rb3 as

dubious because of 8 a4 ll:lx c4 9
xc4 l0g4 ! and now:
a) 1 0 h3 l0 f6 1 1 de d5 ! 1 2 Wa4+
Ad7 1 3 'W'b3 ll:lxe4 1 4 1Wxb7
c6 (Ravinsky) =t=.
b) 10 de d6! with a good game.
c) ECO surprised me by mysteriously
giving 10 lLlxeS lil xeS 1 1 de d6 1 2

.i.f4 ' ' Skotorenko-Tomaszewski,
corres 1 976. After 1 2......, 0-0 l 3 0-0
.i.e6 this assessment loOks quite
wrong to me.
ltlxc4
8 . . . c6?! 9 't!Vxe5 f6 10 'W'h 5+ g6

1 1 1t x h6 Harding-Parker,
corres 1 974.
9
J..xh6
gh
Now Black's pawns are shattered,

which is the point of 8 1Wb5! . I f the
white queen were on a4 Black
would have the zwischenzug 9 . . .
l0b6. Here 9 . . . lild6 does not have
the same effect because of l O
tfxe5 .
10 'Wxc4 (100)
100
B
Another problematic and in

triguing two knights v two bishops
contest to co m pare with the
resultant of our Lasker Defence
main line. Is it better to have the
bishops with broken pawns (but
still one extra)? Or the knights
with a spatial advantage and
strong centre? My friend and co
analyst Tim Harding has been
successful on the white side of this
line several times in postal games,
5 . . . .te7 91

but still regards the issue as
unresolved.
The evidence we have to go on
may be set out as follows:
a) 10 . . . d6 10 de ( l l ll::l b d2 !?) I I . . .
.ie6 1 2 "t!Vb5+
a l ) 12 ... 'if;f8? 13 1!xb7 'it;g7 14
(}.() Harding-J.H.Hodgson, corres
1 974-5 .
a2) 1 2 .. 1i'd7 1 3 "t!Vxb7 0-0 and
now not 1 4 0-0? .ic4 intending . . .
lifb8. White could try 1 4 1!a6 or
1 4 lll bd2. Against the latter ECO
gives 14 ... de 15 lll xe5 1i'd6 1 6
ltJ f3 f6 intending . . . 't!rd3, which
looks strong. The position is
opening out to the benefit of
bishops, so probably I I de should
be reconsidered.
.
b) 10 . . . dS!? I I ed ed 1 2 0-0 1!fd6
1 3 1!fxd4 Harding-Turek, corres
1974-5.
c) 1 0 ... ed I I cd and now:
c l ) 11
dS!? 12 ed 0-0 is
a suggestion from Konstantino
polsky. The doubled d-pawns are
actually strong, but Black can give
up a pawn for counterplay: 1 3 0-0
c6 14 de be 1 5 -.xc6 .ie6.
c2) 11 ... .if6? 1 2 ll::l c 3 c6 13 0-0 0-0
14 e5 .ig7 1 5 ll::l e4 Rozhlapa
Belova, Moscow 1 972.
c3) 11
d6 12 0-0 0-0 13 ltJc3 c6
has the air of normal play about
it. Harding-Micklethwaite, corres
1 974-5, continued 1 4 l::i: a b l ( 14
d5?! .tf6; 14 l::i: ae l !? 'if;h8 !?
intending . . . f5 - Larry Evans)
14 . . . 'if;h8 15 'if;h l l::i: b8 16 d5! c5
17 e5 , but one presumes that
.
..
..
Black's play could be improved.

Conclusion: 5 . . . .ie7 has enjoyed
a good reputation, but I think it
is overrated. Notice how few
variations in this section terminate
in Black's advantage. Moreover,
as in many E vans Gambit lines,
White scores well in practice.
With 6 d4 White has good
prospects of t and little fear of
anything worse than equality.
6 Wb3 leads to a very interesting
and unbalanced position. But,
beware! Although it might achieve
more than 6 d4, it is significantly
riskier.
Peel-back: White can play 5 0-0 if
he wishes (as in fact Evans did in
his famous victory over McDonnell),
but this will t ranspose into 5 c3
and 6 0-0 variations. So we now go
on to consider what may happen if
Black declines the gambit at
move 4.
Section 1 0
The G a m bit
Decl i ned
1
2
3
4
e4
eS
ll::l f3
ltJc6
.ic4
b4
.icS
.ib6 (101)
The Evans Gambit is usually

accepted. Is this because people
think that Black can advantageously
get away with taking the pawn? Or
is it . because they think the

declined version of the Evans is
inferior for Black ? It is an
interesting fact that the Goring
Gambit is much more commonly
declined than the E vans is, even
though it is doubtful whether the
Evans pawn is really a safer pawn
to take than the Goring pawn. I
uncharitably suspect that many
players imagine that White can
j ust win the e-pawn with 5 b5 and
6 ltl xe5. Another factor is that
there is a plan for W hite against
the decliner (our variation B I )
that has been grossly overrated.
101
w
We shall analyse the con

sequences of:
A 5 b5?!
B 5 a4
Note, however, that another
major option for White is 5 c3
with d3, 0-0 and a4 to follow,
which will transpose into Bird's
treatment of the Giuoco Piano
(see Greco Gambit introduction).
The Gambit Declined

5
ll:la5
5
ltld4 is playable, but
concedes White a stable advantage
after 6 ll:l xd4 .txd4 7 c3 .tb6 or
6 . . . ed 7 0-0 d6 8 d3 . But not 6
ll:lxe5?, which leads to a catastrophe:
6 . . . 'fi'g5 7 ltlxf7 '@xg2 8 lUI
'fi'xe4+ 9 .te2 ll:lf3 mate.
6
ll:lxe5 (102)
5 b5 lacks motive if White isn't
goin g to d o t his. Besides, after the
insipid 6 .te2 both 6 . . . d6 and
6 . . . d5 !? are fine for Black, e.g. 6 . . .
d5 ! ? 7 ll:lc3 !? (better 7 d3) 7 . . . d e 8
ll:lxe4 f5 9 ll:lc3 e4 =t= Spielmann
Burn, Carlsbad 1 9 1 1 .
...
102
B
6
'it'f6
Black has three good moves in
this position, but I would recom
mend 6 .. . 't!Vf6 as the most
straightforward refutation of White's
play. The other two are:
a) 6
\!t'g5 7 .txf7+ e7 8 .txg8
'it'xe5 9 .td5 and now I think that
Black could probably just take the
rook: 9
Wxa l 10 ll:lc3 d6 1 1 0-0
.td4 !?. Levenfish considered this
too risky and gave 9 . c6 10 d4
.txd4 1 1 f4 'it'f6 12 c3 .txc3+ 1 3
...
. . .
A
5
b5?!
Don't do it!
. .

lbxc3 t!t'xc3+ 1 4 .id2 t!t'd4 1 5
.txa5 1!Ve3+ 1 6 1!Ve2 1Wxe2+ 1 7
<t>xe2 cd 1 8 ed b6 1 9 .ic3 .ib7,
and a draw seems likely. But of
course many deviations from this
knotty string of moves are possible.
b) 6
lbh6 7 d4 d6 8 .i xh6 de (8 . . .
gh!? i s also held t o favour Black
after 9 .ixf7+ We7 or 9 lb xf7 1!Vf6
or 9 !i'h5 t!t'f6 1 0 lb xf7 0-0 ! ) 9
.ixg7 llg8 1 0 i.xf7+ Wxf7 I I
.txe5 'fi'g5 oo but Black's piece is
probably worth more than White's
four pawns (S teinitz-Dubois, 2nd
match game 1 862 and Reti-Perlis,
Vienna 19 1 3).
7 i. xf7+
Wf8
d4
8
d6
9 .i xg8
de
.ixd4 (103)
10
.idS
...
103
w
The Gambit Declined 93

B
5
a4
a6
Notice that 5
lbxb4?, a
deferred acceptance oft he gambit,
is very bad : 6 a5 i.c5 7 c3 llJ c6 8
0-0 d6 9 d4 ed 1 0 cd i.b4 1 1 d5 and
't!t'a4+, picking up the bishop on b4.
After 5 . . . a6 White has:
Bl 6 lLlc3!?
B2 6 .ib2!?
Bl
6 llJ c3!?
Soviet theoreticians , and others
following them, have been unduly
i mpressed by this move. Perhaps
the fact that it was devised at the
board by Kan has aroused generous
inclinations towards such invention.
6
lLl f6
lbxd5
7 llJd5
For the second time White was
offering a pawn, which Kan's
opponent unwisely accepted in the
original 1 929 ga me: 7
lLl xe4
8 0-0 0-0? ( better 8 . . . lb f6) 9 d3
llJf6 10 .ig5 d6 1 1 lLld2! i.g4?
12 .ixf6 t!t'c8 13 lLlxb6 cb 14 f3 .
The hapless victim was none other
than the young Botvinnik.
8
ed (104)
...
...
White has a roo k and a mate en

prise. An old analysis (Max Lange
in the Paris 1 867 tournament
book) proceeded 1 1 f4 .ixa 1 12 fe
t!t'g6 1 3 0-0+ We8 1 4 .if7+ 'fi'xf7
1 5 llxf7 xf7, when Black, with
two rooks and bishop for the
queen, must be winning.
1 04
B
'
The standard reference here is

Sokolsky-Goldenov, Kiev 1 945 :
8 . . . e4 9 de 0-0 1 0 J.b2 ! (an
improvement on the 1 0 0-0 ef 1 1
1!t'xf3 de = of Sokolsky-Lilien thai,
USSR Ch 1 944) 10 . . . ef 1 1 1!t'xf3
de 1 2 1!t'c3 lile8+ 1 3 fl 'ifg5 1 4 h4
1!t'h6 1 5 a5 J.a7 1 6 h5 J.e6 1 7 lilh4
with some advantage to White.
There are, however, two reasons
why an assessment of this variation
should not merely reflect this
game.
I n the first place, Neistadt
points out that 9 ... 0-0 is a piece of
misplaced cleverness. Instead 9 . . .
ef 10 '@xf3 ( l.Q cd+!?, trying to win
a pawn, is ignored by both
Neistadt and ECO, though it is the
only way to test Black's play) 10 . . .
1!t'e7+ I I 'it' d ! d e 1 2 J.b2 1 3
lile 1 J.e6 1 4 J.xe6 fe 1 5 'it'g4 0-0-0 !
leaves White not a jot better, e.g.
16 'it"xe6+ 'ifxe6 17 lil xe6 J.xf2 or
16 lil xe6 'it"d7 intending 17 . . .
lilge8.
In the second place, I would
ask : why should Black open the
diagonals for the bishops on c4
and b2 by playing 8 . . . e4? Isn't 8 . . .
ll:ld4 all right? I think i t equalizes
quite easily. After 8 ... ll:ld4 we can
consider:
a) Playing to win a pawn with
9 ll:lxeS - but after 9 .. . 0-0 1 0
0-0 d6 White either has to suffer
I I ll:lf3 J.g4 1 2 J.e2 ll:Jxe2+ 1 3
or +
'it'xe2 lile8 1 4 'it"d3 'it'f6
(Bednarksi-Minev, Warsaw 1 96 1 ),
or else play the awkward 1 1 it)d3
=
The Gambit Declined

(Cafferty and Harding), to which I
think ll . . . 1!t'g5! is a frightening
reply (intending . . . J.g4).
b) Following Alekhine's example
with 9 0-0. Now 9 . . . d6 1 0 it)xd4
J.xd4 1 1 c3 J.a7 1 2 d4 was ;t or
in Alekhine-Fuentes, Melilla 1 945.
However, a simple improvement
is 9 . . . ll:Jxf3+ 1 0 1!Vxf3 d6, as
Alexander recommends in Alekhine's
Best Games of Chess 1 938-45, and
White has nothing. Note that
9 ll:l xd4 J.xd4 1 0 c3? does not get
White into the Alekhine-Fuentes
scheme of things because of I 0 . . .
J.xf2+ 1 1 xf2 1!Vh4+.
B2
6
7
8
9
J.b2!?
d6
bS
ab
lilxa l
ab
J.xa l (1 05)
105
B
it)d4
Others:
a) 9 . . . ll:laS 10 J.a2 J.g4 1 1 d3 ll:Jf6
1 2 0-0 t Kostic-Yates, The Hague
1 92 1 .
b) 9 . . . llJb8 1 0 d4 ed 1 1 J.xd4
J.xd4 1 2 'ffxd4 'fff6! ( 1 2 . . . ltlf6
1 3 0-0 0-0 1 4 it)c3 ;t Tartakower-
Yates, Carlsbad 1 929) 13 eS! de 14

liJxe S .i.e6 I S .ixe6 1i'xe6 16 0-0
liJf6 1 7 lle l 0-0 1 8 liJd3 1i'c8 ;t
Tartakower.
10
lbxd4
Better than 10 .ixd4 ed l l 0-0

liJf6 12 d3 0-0 13 lbbd2 (Tartakower
Schlecter, Baden 1 9 1 4), when
13 . . . d5 ! is +.
10
11
12
13
14
15
c3
0-0
d3
ed
1i'f3
ed
lbf6
0-0
d5!?
liJxdS
lbf6 (106)
106
w
Tartakower- Rubinstein,
The
Hague 1 92 1 , went on 1 6 cd i.xd4

But
1 7 liJc3 lbg4 1 8 lbdS liJ eS
the patient 1 6 h3 lle8 17 liJ d2 ;!; of
Tartakower- Rhodes, Southport
1 950, improves: White's bishops
are more active and he threatens
to assume control of important
squares with 1 8 liJ b3.
=.
Conclusion: The Evans Gambit

Declined is not so bad as
theoretical repute would have it.
One can perhaps understand why
people do not like it much, for it is
unlikely to make adrenalin flow
and pulses race. Only when White
does something silly, as in variation
A, do things get interesting for
Black. White should guard against
ove rreaching. Since the oft lauded
6 ltJc3 (B 1 ) is not at all promising,
if accurately countered, I would
advise White to try Tartakower's
treatment (B2: 6 .i.b2). This line
does not carry a heavy, knockouf
punch, but White seems able to
maintain nagging pressure in spite
of simplification.
SCO TCH
GAMBIT
1
2
3
4
eS
e4
lbc6
lbf3
ed
d4
i.c4 (1 0 7)
107
B
The name derives from a correspondence match between the London

and Edinburgh clubs, played in the years 1 824-28 . Fr.om White's point of
view the only thing wrong with the Scotch Gambit is that it hardly ever
occurs - except as a m ove-order rather than an opening in its own right.
Lines in which Black attempts to hang on to the pawn (Scotch Gambit 1)
giv e White excellent compensation. T h e reason for this is tha t, by
deferring action against the d4 pawn , White hands Black a choice of
developing moves and choices are always opportunities for error.
Mobilizing the bishop on f8 with 4 . .tc5 is natural enough, but puts the
bishop on a somewhat vulnerable square and one that is ill-adapted to
defensive purposes if White then continues in gambit style with 5 c 3 .
However, because Black has been given a choice, the Scotch Gambit is
liable to all sorts of transpositions. The first and most i mportant of these
is 4 . . . lt:Jf6 ( ! ), which i m mediately takes the game into the Two Knights
Defence (briefly surve yed in Scotch Gambit 3).
The second transpositional possibility arises after 4 ... i.c5 5 c3 when
Black can (and I think should) transpose into the Greco Gambit with
5 . lbf6 6 cd i.b4+ etc. As we found Black's resources against the Greco
. .
97
Scotch Gambit: Introduction
Ga mbit t o b e fully adequate, o n e might wonder whether White might

not try to avoid this by deferring confrontation with the d4 pawn for
another move. Let us look at this in the context of a lively old game.
Staunton-von Janisch
Brussels 1 853
1
2
3
4
e4
lbf3
d4
.tc4
e5
lt:Jc6
ed
.t c5
4 . . . .ib4+ is one of the most foolish things to play against the Scotch
Gambit for reasons given in the peel-back note to Scotch Gambit 2. An
example is 4 . . . .tb4+ 5 c3 de 6 0-0 d6 7 a3 .ta5 S b4 .ib6 9 'f!Vb3 1!tf6 10
.ig5 't!fg6 I I lbxc3 .te6 I 2 lt:Jd5 h6 13 .id2 .ixd5 ( 1 3 . . . lbge7 14 a4 is
also extremely unpleasant for Black) 14 ed lt:Jce7 1 5 a4 a6 16 a5 .ta7 1 7
b5 ! lt:Jf6 1 S b 6 t von Bilguier-von der Lasa, Berlin I S 3S.
5
0-0
Crude assault on f7 with 5 lt:Jg5 lt:J h 6 6 't!fh5 1We7 should not work. If
7 0-0 Black can simply reply 7 . . . d6 intending . . . .id7 and . . . 0-0-0, wh ilst
7 f4 is best met by 7 . . . d5! S .txd5 lt:Jb4 9 .ib3 d3 +.
d6
Here is a further transpositional chance for Black - 5 . . . lt:Jf6. If then 6

e5 d5 7 ef de we have the Max Lange Attack (see Scotch Gambit 3).
Alternatively , 6 c3 lb xe4 7 cd d5 is variation 8 of G reco Gambit S.
6
7
8
9
c3
lbxc3
.t xe6
'inl 3
de
.ie6
fe
1Wc8
As in analogous positions in the Goring Gambit (e.g. Goring Gambit

4, variation A) 9 . . . 't!Vd7 ! is preferable . Defence of the b7 pawn is too
trivial a task to saddle the queen with in such an open position.
.t xe3
1 0 .te3
11
12
13
fe
lb g 5
lt:Jf6
lt:Jd8 (1 08)
liac1
I doubt whether this is best despite the result. More forceful is 13 e5!?
de 14 liad l . This inhibits castling - 14 . . . 0-0 15 lixdS li / \!fxdS
15 lt:Jxe6 . The question is whether White can win after 14 . . . h6 1 5
lixdS+ ! WxdS 1 6 lt:Jxe6 We7 . The attempt 1 7 lt:Jb5 lieS 1 S lt:Jxg7+ 't!fxg7
19 't!Ve6+ d8 20 lt:Jxa7 liaS 2 1 lldl+ lbd7 22 l hd7+ 't!fxd7 23 't!ff6+
98
Scotch
Gambit: In troduction
108
w
8 24 'fi'xh 8+ t7 25 'fi'xa8 ird l + allows Black to escape with

perpetual check.
a6
13
1 4 lba4
'fi'd7
h6
e5
15
b5?!
1 6 lbf3
16
lbd5 looks m ore sensible, though White obviously has good
attacking chances.
ba
ef
17
gf
1 8 'it'c2
18
0-0 was quite playable.
e7??
1 9 @g6+
Idiotic. He should have reconciled hi mself to 19
'it'f7 20 'fi'xf7+
lbxf7 2 1 ll xc7 0-0 2 2 lbd4 ;t
de
20 li:\e5!
21
1Wxf6+
1 -0
So try to play the Scotch Gambit if you wish But if you do, you must
be prepared for something entirely different !
. . .
...
...
109
w
Section 1
M a i n Li n e :
5
de
. . .
1
2
3
4
5
e4
e5
ll)f3
lb c6
d4
.ic4
c3
ed
.ic5
de (109)
Main Line 5 . de 99
Scotch Gambit: Section 1
I have chosen this as the main

line because it yields real, red
blooded gambit-play. Black's moves,
however, do not appeal to modern
taste and so this position is rarely
seen in practice.
6
lbxc3
White can play 6 .txf7+ but
there is a danger that it will lead to
simplification: 6 .txf7+ 'i!7xf7 7
'ti'd5+ 'i!7f8 8 "t!txc5+ fi'e7 9 1!ixe7+
(9 't!fxc3 is more lively and White
may have the value of his pawn
after 9 . .. 'i!t'xe4+ 1 0 .te3) 9 . . .
tJge7 1 0 lb xc3 and now 1 0 . . . d5
1 1 ed .!Db4 = Schlechter-Spielmann,
Baden 1 9 14. Black can also play
8
d6 here, an interesting recent
example of which was 9 1!t'c4 !?
.tg4 (9 . . . cb !?) 1 0 lb xc3 .txf3 1 1
gf "t!ff6 1 2 f4 "t!tf7 l 3 'fi'b5 tJd4 1 4
'ti'd3 lbe6 1 5 f5 .!D e S 1 6 'fi'c2 "t!tc4
1 7 .te3 lbf6?! ( 1 7 . . . .!Dd3+!)
1 8 0-0-0 Ile8 1 9 f3 Sveshnikov
Kupreichik, Hastings 1 984-5.
It is worth noting that the
position after 6 lb xc3 could also
be reached via a Goring Gambit
( I e4 e5 2 lbf3 tJc6 3 d4 ed 4 c3 de 5
lbxc3 .tc5 ? ! 6 .tc4). White may
draw encouragement from the
fact that in this sequence Black's
5th move is inferior to 5 . .i.b4.
...
d6
1
.tgs
The most promising way of
prosecuting the attack. Others:
a) 7 0-0 tJ f6 8 .i.g5 (useless is 8 e5?
de 9 'fi'e2 0-0 1 0 i.g5 .ig4 I I lit ad 1
tJd4 12 Wxe5 .i.xf3 13 Wxc5
.txd l 14 litxd 1 tJe4! + Morschel

Reissman, Lugano 1 968) 8 . . h6
(8 . .te6 9 lbd5 poses problems)
9 .th4 is a reasonable alternative.
Black has difficulties over com
pleting his development.
b) 7 'fi'b3 Wd7 8 tJd5 tJge7 9 'it'c3
0-0 1 0 0-0 ( 1 0 .ih6 Wg4 1 1 lbg5
.ib4 ! 1 2 lbxb4 gh =F) and now not
10
tJg6 1 1 b4 intending .ib2
Schlechter-Hromadka, Baden 19 14,
but Euwe's 1 0 . tJxd5 ! 1 1 ed lbe5
12 .!Dxe5 de 13 't!rxe5 .td6 +.
.!Dge7
7
There is nothing immediate
against 7 f6, though it is as ugly
as sin. But 7 .. 'fi'd7 may be better.
Lasker gave the following variation:
7 . . . 't!rd7 8 "t!td2 h6 9 .th4 .!Dge7 1 0
0-0-0 .!Dg6 1 1 .ig3 a 6 1 2 lb h4 b 5
1 3 .t b 3 .i b 7 1 4 'i!7b l 0-0-0
1 5 llc l , claiming an advantage for
White. This is far from clear. All
that one can say is that White has
very free play for his pieces, whilst
the pawn he has conceded plays
no significant role - for the time
being.
8
lbd5
Going in for a piece sacrifice.
8 0-0 is a less drastic alternative
and seems to preserve a dangerous
initiative. Two variations from
Schleeter's Handbuch:
a) 8 . h6 9 .ih4 .tg4 10 tJd5 'ird7
1 1 b4! lbg6 12 .if6!? .i.xf3 l 3
Wxf3 .td4 1 4 .txd4 tJxd4 1 5 1!fc3
with good attacking chances.
b) 8 0-0 9 lbd5 h8 10 b4 .ib6
I I b5 f6 1 2 be fg l 3 cb .txb7 1 4
.
. .
. . .
...
...
100 Scotch Gambit: Section 1
Main Line 5 .
lt:lxb6 ab 1 5 lt:lxg5 .
f6 (1 10)
The crucial test. In any case 8
i.e6 9 0-0 0-0 1 0 b4 .t xd 5 1 1 e d
lt:lxb4 (Mieses-Salwe , Carlsbad
1 907 ) 1 2 lil b l ! intending a3 is
awkward for Black.
...
..
de
fact 5 . . . l0f6 ! - see the Greco

Moller Gambit complex after 6 cd
.ib4+ etc . 5 . . d3 is another,
though less satisfactory, possibility,
which we take as Scotch Gambit 2.
.
Section 2
D ec l i n i ng with
d3
/10
w
. . .
1
2
3
4
5
e4
lt:l f3
d4
.tc4
c3
e5
l0c6
ed
.tcs
d3 (1 1 1)
Ill
w
9
.txf6
gf
10
lt:l xf6 +
f8
11
1!t'cl !
This idea of Keres' gives White a
very strong (I think winning) attack,
viz:
a) 11
h6 12 1!t'f4 t.
b) 11
7 12 lt:lh5+ and mates.
c) 11
lt:lg8! (the only move) 1 2
l0h5 ! .ib4+' 1 3 fl 1!Ve7 14 'tWf4+
e8 1 5 lt:lg5 lL!h6 1 6 l0f6+ d8 1 7
lt:ld5 t!Vf8/ e 5 1 8 'i!t'h4 with decisive
threats.
...
...
...
Conclusion: The main line of the

Scotch Gambit gives White a very
vigorous attack. Unwise for Black
to enter here!
Peel-back: The next question is
whether Black has anything better
than 5 . . de. His best move is in
.
A safety first move that may

be compared with the similar
procedure against the Goring
Gambit (Goring Gambit 6: variation
C). But is it really so safe? Black
aims to avoid the i mmediate
dangers associated with hanging
on to the pawn and allowing the
gam biteer a lead in development.
The general idea that return of the
material surplus is a n i mportant
defensive resource has a host of
applications. But this o ne I do not
approve of because the defender's
'
'
Declining with 5 . . d3 101

back payment ought to extract
concessions from the gambiteer.
I n this instance White has eliminated
Black's e-pawn and thus remains
with a spatial advantage in the
centre - not an overwhelming plus,
but the basis for a persistent !.
0-0
6
White can also go into action on
the queenside immediately with 6
b4 .i.b6 7 a4 a6 8 a5 .i.a7 and now:
a) 9 'ib3?! 'ife7 10 0-0 and now:
a 1 ) 1 0 . d6 ? 1 1 b5 ab 12 .i.xb5
.id7 1 3 a6! Kostic-Nielsen ,
Munich 1 936.
a2) 1 0 . lLlf6 1 1 .!t:lbd2 0-0 1 2
i.xd3 d 6 13 b5?! ab 1 4 'ifxb5
.i.d7 + Kostic-van Scheltinga,
Stockholm 1 9 3 7 .
b) 9 0-0 .!t:lf6 ( 9 . d 6 1 0 \!fb3
transposes to the main line)
1 0 \!fxd3 i.
6
d6
.i.b6
7
b4
a4
8
a6
8 . . a5 9 b5 gives White tempi:
a) 9 . .!t:le5 10 .!t:l xe5 de 1 1 :S:a2!
intending .ll d 2xd3 .
b) 9
.!t:lb8 - here I used to think
that 1 0 'ii' b 3 was strong, but 10 . . .
\!ff6 i s not bad for Black. So
simply 10 1!Vxd3 i ( 1 0 . . . lLlf6?
I I e5 ; 1 0 . . . lL!d7 ! ) or 10 i.g5
.!t:le7 1 1 \!fxd3 t.
9
a5
.i.a7
1 0 3!?
Of course 10 1!t'xd3 can also be
played and ought to be t- if the
assessment of analogous positions
is right. 1 0 1!t'b3 is more vigorous .
..
. .
but less flexible.

10
1!t'f6
10
1!t'e7 I I b5 ! transposes into
Kostic-Nielsen above.
ab
b5!
11
S hying away from the tactics
that follow by playing 1 1 .. . .!t:l e5
has the disadvantage that after
12 .!t:l xe5 de 1 3 ba ba 14 \!fa4+ .i.d7
15 \!fd 1 White has lasting pressure
against Black's weakened queenside.
Sveshnikov - A. Petros ian, USSR
1974, was an impressive exploitation:
1 5 . . . .!t:le 7 1 6 "t!fxd3 i.c8 1 7 lL!a3!
0-0 18 .!t:lc2 lt:lg6 1 9 .!t:l b4 lid8 20
"t!fg3 lie8 (20 . . . .!t:lf4 21 'Lld5 ! )
2 1 i.g5 1!t'd6 2 2 lifd l t!fc5 2 3 i.d5
lib8 24 i.e3 "t!fxc3 25 .!t:la2 "ii'x a5
26 i.xa7 .!t:lf4 27 i.c4 lib2 28 1!t'c3
'ii'x c3 29 .!t:l xc3 .i.g4 30 lidc l .ll a 8
31 li xa6 li b4 1 -0.
12
a6! (1 1 2)
...
..
...
Black has at most one way out

of deep trouble:
a) 12
be 13 ab .
b) 12 . . ba 13 i.d5 .!t:lge7 ( 1 3 . . .
i.b7 1 4 li xa6 ! , e.g. 1 4 .i.xf2+
15 li xf2 li xa6 16 .!t:lg5) 14 i.g5
...
. ..

'irg6 1 5 i.xe7 t;J xe7 1 6 i.xa8
0-0 1 7 'itth 1 c6 1 8 t;Jbd2 i.e6 1 9
't!Va3 ! lixa8 2 0 'i!Vxd6 li e 8 2 1

lixa6 I . Zaitsev-Aronin, Mos
cow 1 964.
c) 1 2
t;Jge7! is best . l . Zaitsev
gave 1 3 i.d5 i.b6 1 4 i.g5 'i!Vg6 1 5
i.xe 7 rj;xe 7 1 6 lt:lbd2 i n annotations
to the game with Aronin. Black
still has all eight pawns ! But not
for long: the two-pawn surplus
ca n quickly be erased (e.g. by
'i!Vxb5xd3). Does White have an
advantage? I am not entirely
convinced. Black has several
reasonable moves - 1 6 . . . lib8,
16 . . . .i.e6, 1 6 . . . .i.d7 . Unclear.
Declining with 5 .
.
Peel-back: Retracting to Black's

4th move we can dismiss 4 . . .
i.b4+?! a s a move that asks for
trouble : 5 c3 de 6 0-0 (6 be i.a5
7 0-0 is also good) 6 . . . cb 7 i.xb2
with a slashing attack on the way,
e.g. 7 .. . lt:lf6 8 a3 i.c5 9 lt:lg5 0-0
1 0 lt:l xf7 ! li xf7 1 1 i.xf7+ rj;xf7 1 2
e5 . Much more sensible is
4 . lt:lf6, peeling out of the Scotch
Gambit altogether.
. .
ili
Section 3
4
lbf6
(Two Knig hts Defe nce)
. . .
1
2
3
4
...
Conclusion: Analysis does not

conclusively prove 5 . . . d3 to be
bad. My objections to it are based
on principle rather than detailed
calculation. Black's prospects are
at best unexciting. But White
needs to give careful consideration
to the choice between the calm
10 1fxd3 and the forcing 10 1fb3
(followed by b5 and a6).
e4
t;Jf3
d4
i.c4
e5
t;Jc6
ed
lt:lf6 (I 13)
1 13
w
The trouble with the Scotch

Gambit is that it doesn't happen.
For Black usually prefers 4 t;Jf6
to 4 .. . .i.c5 , and that takes us
straight into the Two K nigh ts
Defence. In the light of the two
previous sections we can see the
sense in opting out like this. It's
certainly what I would play as
Black .
T h e normal Two Knights move
order is 1 e4 e5 2 t;Jf3 t;Jc6 3 i.c4
t;Jf6 and now either 4 t;J g5 or 4 d4.
Has White lost or gained by
adopting the Scotch Gambit move
order? I cannot see any gain u nless
you h ave some special reason for
not wishing to face 3 . . . i.c5
(leading to the Greco Gambit, the
Evans Gambit, or some other
...
form of the Italian Game). What

White has passed up is the chance
to play 4 ll:Jg5. But I don't
consider that much of a loss,
inclining to Tarrasch's view that
it's a duffer's move - even though
some pretty distinguished 'duffers'
have played it !
But what can one advise for
White, now that he has got there ?
I have difficulty in answering this
question because I don't know
any really good line against the
Two Knights. I don't believe that
there is a line that promises White
any advantage. Nor can I give a
proper survey of the possibilities for that you would have to consult
some work devoted to the Two
Knights, such as Estrin's 1 983
book The Two Knights Defence
(Batsford).
What we can do is discuss some
of the salient features of the two
main lines, commencing with 5 e5
and 5 0-0. Ta ke 5 e5 first. This
is a move I don't like very
much. Black gets a solid central
position and the bishop pair,
and he can eventually prepare
. . . f6, eliminating White's central
pawn an initiating play along
the f- and e-files. The main line
runs:
5
6
7
8
9
10
e5
.ib5
ll:J xd4
.ie3
.ixc6
0-0
d5
ll:Je4
.icS
.id7
be
'f!Je1 (1 1 4)
4 . . . lilf6 (Two Knights Defence) 103

1 14
w
11
lle1
White wants to kick the knight on

e4 with f3 , but I I f3? lild6 favours
Black.
11
12
13
f3
'it'd2
0-0
lZJ gS
f6!
Black is able to get this in

because after 1 4 ef 1!xf6 15 .ixg5
the knight on d4 hangs. So Black
has completed his opening scheme
and has a good game. In favour of
the 5 e5 line one can at least say
that it does not terminate in early
forced draws (as so many sharp
openings do). A complex middle
game battle should ensue. But in
my opinion there is a tendency
towards + rather than t.
What of 5 0-0? Here everything
depends on the level at which you
are operating. At an advanced
level , against an opponent who
knows the theory, the serious
winning prospects are practically
zero . At a humbler level it is an
excellent weapon since many of
the side variations end in quick
wins for White, who is able to play
natural ,active moves with little

risk of getting into trouble. The
main line runs:
5
6
7
8
9
10
11
0-0
lit e 1
i.xd5
lt:lc3
lt:l xe4
i.d2
ttJxe4
d5
1!Vxd5
'ii'a 5
i.e6
1i'd5
i.g5 ( 1 15)
1 15
B
There were plenty of other

moves on the way! But for Black
at least deviations from this path
are disadvantageous , if not disas
trous. It is still a bit tricky. For
example, 11 ... i.c5? neatly loses a
piece to 1 2 c4! ( 1 2 . . . 1!Vxc4 1 3 IIc l ;
1 2 . . . de 1 3 lt:lf6+; 1 2 . . . 1i'f5 1 3
lt:lh4). 1 1 . . . i.e7?! is also suspect
because of 12 i.xe7 xe7 13 c4 !
1!Vxc4 ( 1 3 . . . de 1 4 1!t'c2 ! and the
black king is dangerously exposed)
14 lite 1 1!t'd5 15 l:tc5 'it'd 7 16 'ii'c 1
liteS 1 7 b4 and White has good
attacking chances. But what every
well-briefed Black should know is
that the safe and solid choice is
1 1 .. . i.d6 !, when White hardly
has anything better than 12 i.f6
4 . . . tDf6 (Two Knights Defence)
0-0 1 3 lt:l xd4 ttJxd4 1 4 1i'xd4 't!Vxd4

1 5 i.xd4 with a completely equal
ending.
But what if Black wants to win?
Unfortunately, I can hardly hope
to please both sides! However, if
Black's ambitions overreach the
boring equality that emerged at
the end of the last line, he really
ought to go into the Max Lange
Attack. Objectively this may not
be any stronger than the 5 . . . lt:lxe4
line against 5 0-0, but at least
Black adopts a more aggressive
and uncompromising posture and White's best moves are not so
easy to find. It is n't all that helpful
to set out the main line of the
Max Lange without explaining
why it is the main line (which
space precludes me from doing),
but for the sake of completeness,
here it is :
5
6
7
8
9
10
11
12
0-0
e5
ef
llel+
lt:lg5
lt:lc3
lt:lce4
g4
i.c5
d5
de
i.e6
1i'd5
1Wf5
0-0-0
'ii'e 5
fe
13 lt:lxe6
14
fg
1 4 i.g5 ! ? is an important alter
native, best answered by 1 4 . . . g6
1 5 f7 i.e7 16 f4 1i'g7 1 7 i.xe7 lt:lxe7
18 lt:lg5 d3 ! 19 lt:lxe6 1!t'xfl 20 lt:lxd8
litxd8 with fair compensation for
the exchange.
14
lithg8

15
1 16
B
.i. h6 (1 1 6)
White has made a lot of

progress on the far right wing, but
Black has an impressive centraliz
ation to set against this. The best
course seems to be 15 . . . d3 ! 16 c3
d2 ! 17 lil e2 ( 1 7 t0xd2? .i.xf2+)
17 . . lild3 when there are two
.
4 ... t0f6 (Two Knights Defence) 105
plausible continuations, which

both end in a draw:
a) 1 8 'W'fl 'W'd5 19 lild 1 t0e5 20
t0f6 1Wf3 2 1 t0 xg8 1Wxg4+ 22 'it;h 1
(22 1Wg2? 1Wxe2 23 t0e7+ \t>d7! H)
22 . . . 'ii'f3+ 23 g 1 'ii'g4+ with
perpetual check.
b) 18 t0xc5 'ii'x c5 1 9 lilxd2 t0e5 20
lilxd3 cd 21 \t>g2 "@d5+ 22 g3
t0f7 23 'ird2 'ird6+ 24 \t>g2 e5 25
g5 'it'g6 26 <M1 'ire4 27 g1 'ifg4+
and Black forces the draw (his
rook being no more available for
active service than White's bishop).
The Two Knights is an essentially
sound defence which has been
very heavily analysed. New ideas
will be hard to come by.
4
1
2
3
4
GORING
GA MBIT
e4
li:Jf3
d4
c3 (I 1 7)
117
B
e5
lt:Jc6
ed
The Goring Gambit - the name comes from a game played at Leipzig
in 1 877 between Dr K . Goring and W .Paulsen - is an accelerated form of
the Scotch Gambit in which White immediately challenges the d4 pawn,
unequivocally declaring his intention of offering a true gambit. The
gambit was in fashion in the 1 960s and early 1 970s, when it was doing
great execution in the hands of positive and aggressive players like
Velimirovic, Ljubojevic, Penrose, Levy and Ciocaltea .
In practice the offer is still frequently treated with respect and the gift
refused. Hence our section on the possibilities of declining (Goring
Gambit 6) is both lengthy and important. Yet the real test lies in the
acceptance of the gambit by 4 ... de . Then we have a question c rucial to
the assessment of the gambit's prospects. Can White afford to offer a
second pawn with 5 .tc4, or should he play 5 ltlxc3?
The disadvantage o f playing 5 ltlxc3 is that 5 . . .tb4 is a very good
reply. You will be able to see this from a comparison of sections 1 and 2
(in which B lack plays 5 . .tb4) with section 4 (with 5 . d6). The general
virtues of the 5 . i.b4 defence are: i) it enables Black to eliminate the
knight on c3, otherwise a dangerous attacker; ii) it is better not to
.
..
. .
107
Goring Gambit: Introduction
consign the bishop to an ineffective role behind the d6 pawn ; iii) after the
exchange on c3 White incurs a structural weakness because his queens ide
pawns are split.
If 5 J.c4 could not safely be answered by 5 . cb, then Black would
have nothing better than 5 . . d6 6 lDxc3, transposing into Goring
Gambit 4, and the gambit would be in business. Unfortunately for
White, there seems to be no serious obj ection to the resolute and
acquisitive defence 5 . cb 6 J.xb2 d6! (Goring Gambit 5, variation B).
This is a serious blow to the gambitee r's prospects.
The Goring Gambit is such a stylish and appealing opening, full of the
spirit of daring enterprise that made for so many enj oyable games before
the Karpovian softly-softly approach became the order of the day. It
would have been nice to be able to report that it w as a playable gambit,
offering fair chances even against accurate defence. The one discovery
that helps in this direction is the move I6 c4 ! in our main line (Goring
Gambit I , with 4 . . . de 5 lDxc3 J.b4 6 J.c4 d6 7 0-0 J.xc3 8 be i.g4).
This does at least prevent 8 . . i.g4 from constituting an outright
..tg4
positional refutation. However, even if the innovation against 8
works out wel l , White's rejoicing will be muted by the fact that Black has
good alternatives at the ready in 8 . . . J.e6 and 8 . . ttJf6 (section 2).
.
. .
. . .
Here are two games by way of illustration of the chances available .

The first chugs along one of the maj or theoretical roads until White's
attack runs out of fuel.
Hawley-Mitchell
corres 1 970-71
1
2
3
4
5
6
7
8
e4
lt)f3
d4
c3
ltJxc3
..tc4
0-0
be
e5
lDc6
ed
de
.ib4
d6
J. xc3
lDf6
8 . . i.g4 is considered in section I , and 8 . . . J.e6 is variation A of

section 2.
.
9
10
11
e5
ttJ xe5
1i'b3
lDxe5
de
1!1e7
108 Goring Gambit: Introduction

12
13
i.a3
i.b5+
c5
c.tf8!?
A better winning atte mpt than interposition on d7.

14
f4
i.e6 (1 18)
14 . . . e4 is taken as the text move in section 2, variation B3. Neither
seems bad , which is the better is unclear.
1 /8
w
15
c4
Although this gives the queen access to squares along the 3rd rank it
can hardly be good to restrict the scope of the bishop on b5 so drastically.
15 .ic4, 1 5 1!Va4 and 1 5 'Wc2 are other moves to try, though White's
co mpensation for two pawns is nebulous.
15
16
17
18
19
20
21
22
llxf4
'it'e3
nan
i.b2
'it'e5
lle4
.tel
ef
h5
ll c8
ll h6
lbe8
f6
li:Jc7
g8
White has pressed hard, but can go no further. The reaction now sets
m.
23
1!rg3
h4!
Black can happily contemplate the liquidation 24 ll xh4 llxh4 25

'ifxh4 lbxb5 =F.
24
25
ll g6
'it'e1
'it'xh4? (l l 9)
Loses instantly, but constructive moves are no longer available, e.g.

25 .if4 is parried by 25 . . . f5 .
1 09
1/9
B
25
26
cb
li:lxb5
lil:xg2+ !
0-1
Upon 27 'it>xg2 comes 27 . . . i.d5 28 lle l lle8 29 'it>f3 f5 , and 27 'it>h l
i.d5 ! 28 ll xe7 :iit g4+ mates.
The second game shows that even when the theoretical assessment i s in
his favour Black must still tread carefully.
Ciocaltea-M. Kovacs
Baja 1 971
e5
e4
1
li:le6
2 lilf3
d4
ed
3
de
4
e3
li:lxc3
5
i.b4
d6
6
i.c4
't!fb3
7
Even the winner did not consider this a very impressive alternative to
the usual 7 0-0.
7
i.xe3+
8
be
In his lnformator notes Ciocaltea gave '8 't!fxc 3 ! ' , but did not explain
what the intention might be i n t he event of 8 . . . 't!ff6. (Perhaps 9 't!rb3 h6
10 i.d2 !?.)
8
1td7
9 't!fc2
li:lf6
0-0
0-0
10
lle8
h3
11
1 10 Goring Gambit: Introduction

Hereabouts an assessment of + is appropriate, as in the peel-back
note to Goring Gambit 2. Black only needs to find an effective role for
the bishop on c8 .
h6
12
.id3
13
Il:bl
a6
14
Il:e1
lt:l eS
de
15
ltJ xeS
16
f4
'it'd6 ( 1 20)
There was a good alternative in 1 6 . . . ef 1 7 .ixf4 'it'c6.
1 20
w
rs
11
. See how persistently White plays for the restriction of the u ndeveloped
bishop on c8 - h 3 , Il: b 1 , f5 . Nonetheless, Black has the better chances
after 17 . . . Il:d8! 18 .ifl b6 (intending . . . .ib7 with pressure on e4).
bS?
17
c4!
18
Now White gets his bishop back onto.the a2-g8 diagonal and is able to
show that Black still has a weak point o n fl .
Il: d8
18
be
19
Il:b3
20
.ixc4
'it'd4+
21
'Ct>h2
.id7
An unsatisfactory attempt to develop the problem piece.
Il: ab8
22
Il: g3
Black is in trouble - after 22 . . . ltJh5 23 l:i d3 'it'b6 24 Il:ed 1 it becomes
obvious how badly the bishop on d7 is located .
23
.ixh6
lt:lhS
li! b2
24
JigS
'ilxd 1
25 d 1
Il:b6
26
li!xd1
111
27
llxhS
gh (1 2 1)
Although material is level Black is quite lost because his bishop o n d 7
i s in a permanent pin a n d all h i s pawns are vulnerable. The rest of the
game speaks for itself.
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Section 1
M a i n Li ne with
8
.i; g 4 !
. . .
1
2
3
4
e4
lLlf3
d4
e3
eS
lLle6
ed
de
lidS
i.b3
ll xh6
lld3
llhd6
f6+ !
i.e2
ll3xd4
c;t> g3
<J.7f4
i.fS
ll xd8
g4
g5
ll h8
1-0
ll b4
ll xe4
e6
wg7
ll d4
g6
gS
ed
wh6
eS
i.xf5
i.e6
c;t> g 6
i.xh3
5 lLl xe3
i.b4
6 i.e4
d6
7
0-0
i.xe3
8
be
i.g4! (122)
Since 8
lLlc6 and 8
i.e6 are
also good moves, perhaps it is just
a matter of taste which of them
o ne prefers . I like 8
i.g4
.. .
. . .
. . .
112 Goring Gambit: Section 1

122
w
because it allows White very little

scope for attack and enables Black
to trade in his extra pawn for
something which is more valuable
- a permanent positional advantage.
Playing this line should not
ove rload the memory since Black
can operate on princi ple rather
than tackling a fresh set of
problems in calculation at each
move.
9 b3
Regrettable, since it allows the
f-pawns to be doubled . But what
else is W hite to play? 9 i.a3 lt:Jf6
transposes into Penrose-Smyslov,
Munich 1 958 (see Go ring Gambit
2, variation B, p. 1 1 5 ) - and that is
known to be inadequa te .
9
..txf3
'i.t>f8
I0
..txf7+
11
gf
lUeS
12
i.xg8
It is a serious matter to part
with this bishop, but 1 2 i.e6 is
neatly refuted by 12 . . . lt::l x f3+ 1 3
'i!;lg2 1!Vf6 1 4 1Vxb7 lt:J h4+ 1 5 'i.t>g3
1!t'f3+ 1 6 <t>xh4 h6! =t=F O 'Kelly.
12
llxg8
Main Line with 8 . .

.
Jl.g4i
f4
13
l 3 fi'xb7 g5 + c a n hardly be
contemplated .
13
lt::l f3 +
1 4 'i!;>g2
lt::l h 4+
I S 'i.t>hl
d7! (123)
16
fS? !
Certainly not 1 6 fi'xb7 ?? fi'h 3
++ . However, White has just one
chance of parryi ng the threat of
. . . h3 and avoiding crippling
positional disadvantage
16 c4!
That move increases the potential
of White's queen and bishop, in
p articular clearing lines along the
third rank and the a 1 -h8 diagonal.
16 . . . g4 can be met by 17 g3
and if then 1 7 . . . e2 1 8 i.e3 with
a good ga me. After 16 f5 by
contrast White's pieces are not
active enough to compensate for
his weak pawns.
16
1lkc6!
In the game we a re following
Black actually played 1 6 ... lle8,
which is inaccurate because it
allows 17 'ikxb7. O'Kelly claimed
that 1 7 . . . 'ti'a4 1 8 f3 'it'c2 was +, but
-
Main Line with 8 . . . .tg4! I I 3
Goring Gambit: Section I

White has 1 9 'it'b2 'i!t'd3 20 'fff2 g5
21 .txg5! Ilxe4 22 .txh4 Milukas
Sutkus, corres 1 976.
Ile8
f3
17
18 'i!t'c2
In Winter-Wason , corres 1 98 384, White tried 1 8 .tg5! ? and
after 18 . . . Il xe4 19 .txh4 Il xh4 20
ae l 1!c4? 2 1 1!xb7 obtained
good chances . However, there was
no need for B lack to give the pawn
back: after 20 . . . Ilc4! the
threatened penetration with 1!e6
is prevented and it is clear that
Black's extra pawn and sounder
formation should tell in the long
ru n .
Since writing this I found that
20 . . . Ilc4! has been tested in
Richter-Volz, 19th German Corres
Ch 1 983-85: 20 . . . Ilc4 2 1 Ile3
g6? ! 22 f6 rtifl 23 Il e4?! b5 24 a4
a6 25 'i!t'a2 Ile8 26 ab ab 27 1!d2
Ucxe4 28 1!h6 xf6 29 fe+ rtie7 30
'i!t'xh7+ d8 and Black was
clearly winning (0- 1 , 35). This is a
potentially misleading example,
since the obvious 23 Ilfe 1 would
have given White dangerous attack
ing chances (as Richter indicates
i n his annotations in Fernschach).
In my opinion Black runs fewer
risks by answering 21 Ile3 with 2 1
. . . fl (intending to untangle with
. . . Ilf8 and . . . g8 ) and if then 22
Ilfe 1 Ilf8 23 Il e7+ g8 24 Il g 1
Ilfl +.
18
g5!
19
fg
lt:Jxg6
<tJe7 (124)
20
.th6+
Black stands better because

White's pawns are split into fou r
'islands ' . In Ciocaltea-Karaklajic,
S mederevka Palanka 1 97 1 , there
followed 2 1 f2 <tJd7 22 Il ad 1
Ile5 23 c4 Il h5 24 1!d2 lt:Je5 25
.tf4 1Wxc4 + (0- 1 , 52).
Conclusion: This is an elegant

positional line of defence . Most of
White's moves 1 0- 1 8 were forced ,
but he does have one chance of
striving for piece activity to
compensate for his pawn weaknesses
16 c4! Further investigation of
this possibility is needed: one idea
is 1 6 c4 'ffc6 1 7 f3 Ile8 18 1Wb5 ! ,
for the pawn would be more
desirably located at b5 than at c4.
-
Peel-back: Let us see how much

hangs on this. How do Black's
other 8th move choices fare?
Section 2
Black's 8th M ove
Alternatives
1 14 Goring Gambit: Section 2

1
2
3
4
5
6
7
8
e4
lt:\f3
d4
c3
ltJ x c3
.tc4
0-0
be (125)
e5
lt:\c 6
ed
de
.t b4
d6
.t x c3
125
B
Black's 8th Move A lternative$

exam ple , 1 3 . . . 't!Vd7 I4 .ta3 't!Vc6
1 5 \t>h i lt:\e7 1 6 Ilae l 0-0 17 f5
a 5 ? ! ( 1 7 . . . 'i!Yb6 must be better) I S
ltJf4! a4 1 9 't!Vd l e5 20 lt:\h5 Ribli
Kovacs, Hungary I970 .
ll
't!Vxb7
Now 1 1 ltJg5 ltJd8 12 f4 h6
1 3 ltJ h 3 1c6 puts Black a vital
tempo ahead of Ribli-Kovacs.
Ilb8
ll
12 't!Va6
lt:\ge7
0-0 ( 1 26)
1 3 .t g5
126
w
Black has two other p er fe c tly

reasonable moves besides 8 . . .
.tg4:
A 8 . .. .te6
B 8 ... lt:\ f6
Indeed, there is even a fou rth 8th
move option - 8 . . . 't!Ve7. Bu t that
Jacks i n depende nt significance if
best play is (as it s ee m s to be) 9 e5
ltJxe5 10 ltJxe5 de l l 't!Vb3 lt:\f6 1 2
.ta3 c5 I 3 .tb5+ with transposition
into variation B (8 . . . lt:\f6).
A
Bjerring- Lein , Varna I 974, con

tinued 1 4 li fd1 ltJg6 1 5 l't'a4 Ilxf3!
16 gf lif8 17 .ie3 e5 IS 'it>fl lt:\h4
+ . Clearly White must not allow
Black this s o rt of chance in the f
file. So the correct course is 14
.txe7! Il b6 I5 't!Va4 ( 1 5 't!Vxb6 ab
I6 .txf8 \t>xf8 I take to be + in
view of Black's better pawns) I 5
. . . 'ii'xe7 1 6 ltJd4
=.
8
9
10
.txe6
't!Vb3
.te6
fe
'ikd7!
Black should shun the pa ssive

10 . .. 't!Vc8?! I I ltJg5 ltJd8 I 2 f4 h6
1 3 lt:\h3 when he is left with
problems of mobilisation. For
B
8
9
lt:\f6
e5!
White needs to open avenues of
attack before Black can consolidate.
9 .ta3?! works fine in the event of
Goring Gambit: Seciion 2
Black's 8th Move Alternatives
9
0-0? l O e 5 ! , but p roves
inadequate against 9 . i.g4! as
demonstrated i n Penrose-Smyslov,
Munich 1958: 9 i.a3 i.g4 ! 10 i.b5
( 1 0 't!fb3 .!Da5 ! l l i.xt7+ c;i>f8 1 2
't!i'a4 i.xf3 1 3 gf ..t>xt7 1 4 't!i'xa5
lle8 +, a variation give n by
Alekhine, is a near relative of the 8
i.g4 scheme) 1 0 . . . 0-0 1 1 i. xc6
b e 12 e5 .!Dd5 13 't!i'd3 (perhaps 1 3
c4 .!D b6 is less bad for White than
the game) 13 . . . lle8 14 ed .!Df4! 1 5
'tli'c4 .!De2+ 1 6 c;i>h 1 i.xf3 1 7 gf cd
1 8 'ti'xc6 llc8 19 'ti'xd6 "i!Vh4 +.
i.a3 was given by Levy. But

Black has better in 12 ... 'it'f6! .
de
1 0 .!Dxe5
11
't!b3
1 1 't!t'xd8+ xd8 12 i.xt7 <t/e7
13 i.b3 i.e6,Yukhtman-Furman,
USSR Ch 1 959, offers White
equality at be st.
11
'fke7
12
i.a3
c5
1 3 i.b5+ (12 7)
White does not succeed with 1 3
't!rb5 + .!Dd7 1 4 lil:ae l 0-0 1 5 f4 a6 1 6
"i!Vb3 e4 1 7 lle3 lle8 1 8 l:Ue l .!D f6
1 9 'it'b6 i.e6! 20 i.xc5 'ti'd7 +
Penrose-Unzicker, Leipzig 1 960.
.. .
..
. . .
.!Dxe5
9
White has more tangible pressure
after 9
de, although this should
not be dismissed too lightly:
a) 10 l2Jg5 i.e6! (Best . If 1 0 . . .
'ft'xd 1 1 1 i.xf7+ c;i>f8 1 2 ll xd 1 .
And if 1 0 ... 0-0 1 1 i.a3 't!i'xd 1 1 2
llaxd 1 i.f5 1 3 i.xf8 ll xf8 1 4 ll fe l
Yukhtman-Tal , USSR C h
1 959) 1 1 i.xe6 fe 1 2 't!fb3 'i!fd5 1 3
.!Dxe6 'tWxb3 1 4 a b t7 Aroni n .
b) 1 0 't!t'e2 0-0 1 1 i.a3 i.g4 1 2
i.xffi Velimirovic-Haag, Vnijacka
Banja 1 966. But 10 . . . i.e6! is
again the best defence.
c) 1 0 'ti'b3 0-0 1 1 lld l may be the
strongest:
c1) 11
i.d7 1 2 .ia3 .!Da5 1 3 1!t'b4
.!Dxc4 (relatively best is 13 . . . c5 ! ?
1 4 1!t'xc5 llc8 1 5 1!t'xf8+ 'ti'xf8 1 6
.ixf8 .!Dxc4 1 7 i.e7 when White
should win , but it's a long job) 1 4
Wxc4 lle8 1 5 .!Dg5 't!t'c8 1 6
1Wxf7+ c;i>h8 1 7 .ie7 .!D g8 1 8 .if8 !
1 -0 Keffler-Kocem, corres. 1 97 1 .
c2) 1 1
.!Dd7!? 1 2 i.d5 .!De7 1 3
1 15
...
...
There are
consider:
three
replies
to
81 13 ... .!Dd7
82 13 ... i.d7
83 13 . . c;i>f8! ?
It is a significant comment o n
the state o f the gambit that the
.
real question here is whether Black

has winning chances.
81
13
.!Dd7

Not this way! The result should
be Yl Yl . After 14 i.xd7+ i.xd7 1 5
1!t'xb7 0-0 1 6 llad l ll fd8 White
equalises mate rial and position
with 17 i.xc5 't!fxc5 18 ll xd7
llxd7 1 9 'it'xa8+ 'tiffS 20 1!t'e4 . The
more ambitious attempt 17 lld2
1!t'e8 1 8 .txc5, Velimirovic-Littleton,
The Hague 1 966, is also equal.
Perhaps the most accurate is then
18 . . . i.b5. But there was nothing
wrong with 18 . . . i.c6 19 ll xd8
and now 1 9 . . . i.xb7 20 ll xe8+
llxe8 = (instead of 1 9 . . . ll xd8 20
'it'c7 llc8 2 1 'ti'xa 7 lla8 22 '@'c7
llxa2 23 lld l ;!; as played).
B2
i.d7
13
'ti'xd7
1 4 i.xd7+
But not 1 4 . . . li:lxd7 15 'it'xb7 0-0
16 li ad l with strong pressure.
'ti'c6
1 5 i.xc5
15 . . . lt::l e4 is riskier because
White can ignore the fork with 16
i.a3! li:ld2 1 7 'ti'b4 and now:
a) 17 ... li:lxO 18 ll d l ! : 'ti'e6 19
1!t'b5+ ''c6 20 9xe5+ ''e6 2 1
9xg 7 li:ld2 2 2 9xh8+ 'it>d7 23
''xa8 l -0 Ribli-Imre, Budapest
1 968.
b) 1 7 . . . 0-0-0! is still O K : 18
lUel ''c7 1 9 c4 lld3 20 c5 f6 oo
was Prieditis-Kiss, corres 1 977. I
would prefer 1 8 . . . lihe8 followed
by . . . 1!Vc6.
1 6 i.a3
lt::l d 5
Ribli's 16 ... 0-0-0 (!) 1 7 't!t'xf7
lld7 1 8 'it'b3 and now 1 8 . . . lihd8
is probably an improvement.

Black has a strong centralisation.
17
llae 1
0-0-0
18
llxe5
lihe8 (128)
would assess this as equal,

though in practice Black has
done rather badly . H ere are a
couple of games in which the
draughtier location of Black's
king played a part:
1 . S mederevac-Lengyel, Polanica
Zdroj 1 966: 1 9 lixe8 ll xe8 20 g3
li:lxc3 2 1 h4 li:le2+ 22 h2 f5
( Black is crushed with amazing
ease after this natural attacking
continu ation) 23 ll d l ! (stops . . .
li:ld4 and controls d6) 23 . . . 't!t'e6
(23 . . . f4 !?) 24 ''a4 'it>b8 25 i.d6+
'it>a8 26 i.c5 a6 27 't!t'a5 ''c8 28
1!t'b6 'ti'b8 29 lld3 +t- llc8 30 lia3
ll xc5 3 1 'i!t'xc5 f4 32 lld3 1 -0.
2. Menvielle-Medina, Las Palmas
1 9 72: 1 9 : lif5 lid7 (very solid is l 9
. . . 't!t'e6 20 llf3 f6 2 1 .te l li:lb6 22
't!t'a3 a6 = S mederevac-Clarke,
Wij k aan Zee 1 970) 20 h3 g6 2 1
li D f5 22 .te l lle4 (22 . . . li:l b6 ! ?)
23 ll d 3 ''c4 24 ll fd l lie ! +? ( A
little combination that misfires
badly. 24 . . . li:lb6 was better) 25
lixe I 'it'xd3 26 lie8+ c7 27 1!Va3 !
Goring Gambit: Section 2
Black's 8th Move A lternatives 1 1 7
't!fxc3 2S .tf4+ c6 29 l:le6+ b5

30 'it'xa7! 't!fa5 31 't!fd4 W'b4 32
'tfd3+ a4 3 3 lieS lLlb6 34 liaS+ !
lLl xaS 35 'it'xd7+ 'tfb5 36 'ti'xh 7
't!fb l + 37 h 2 1!t'e4 3S W'd7+ a5
39 .te3 'tre5+ 40 g3 lLlc7 41 .tf4
1 -0
I used t o think this position

was clearly better for Black arid
that White was struggling in vain
to justify his material investment.
But now I a m not so sure. Two
game examples tell very different
stories:
1 . Smederevac-Minic, Yugoslav
Ch 1 959: 16 liad1 a6 (this looks
like the start of a bad plan) 1 7
.te2 b5 1 S c4! b4 1 9 .te l a5
(Forgetting about the pieces! But
how else to meet the threatened
.tf4-d6 ?) 20 .tf4 lia6 21 1!i'g3 h4
22 '@g5 ltJeS? (22 . . . .tb7! ?) 23
li!d5! f6 24 't!fg6 and White has
formidable pressure. For example,
he can answer 24 . . . .tb7 with 25
lifd I .txd5 26 cd ll:d6 27 .th5 ! .
2 . Prizant-Eley, Southport 1 972:
16 liae1 b6 1 7 .tc6 l:lbS IS 'i!t'a4
.tb7 ! 1 9 't!fxa7 't!fc7 20 .txc5+ be
2 1 li b 1 li:lg4 22 g3 lih6! 23 l hb7
li xb7 24 .txb7 h4 +.
B3
13
f8! ?
This is the move Black should
choose if he is playing to win .
f4
14
e4
Also quite playable is 1 4
.te6. After 1 5 .tc4 .txc4 1 6 'ti'xc4
b6 1 7 fe 'it'xe5 1S liae 1 1!t'd5 1 9
'tfh4 White retains attacking
chances, but it is doubtful whether
he has genuine compensation for
the two-pawn deficit.
15
f5
h5 (1 29)
Black thus restricts White's
kingside and enables the rook on
hS to come into play via h6. There
is a viabl e alternative in Nikitin's
1 5 ... g8 (intending . . . h6, . . .
h7), e.g. 1 6 li ae 1 b 6 1 7 .tc6
i.b7 1S .txb7 't!fxb7 19 .txc5 h6
20 .td4 h7 21 .txf6 gf 22 lif4
liae8 +. (But maybe just 22 lixe4!?.)
129
w
So what should happen? 1 6

li ad 1 is the better place for the
rook . The appropriate way for
Black to get his pieces working is
surely 16 ... b6, plausibly continuing
1 7 .tc6 .tb7 ( 1 7 . . . l:lbS ? 1 8 .te l !
with the terrible threat of 1 9 .tf4)
I S .txb 7 'it'xb 7 1 9 .txc5+ gS 20
.td4 and whilst Black's game is
probably tenable, the results of
Nikitin 's 15 ... 'it>gS seem preferable.
Conclusion: S . . . .te6 is quite safe ,
although Black cannot count o n
any advantage that way ( 1 4

.txe7 !). 8 . . . lt:lf6 remains a
problematic variation , in which
Black will have to play 1 3 . .
<M8 ! ? if he is after the full point.
But this does involve some risk.
.
Peel-back: We need not devote

much space to alternatives in the
sequence 7 0-0 .txc3 8 be. Moves
other than 7 . . . .txc3 give White
the dangerous possibility of lt:ld5
and leave the bishop on b4
hanging tactically loose . 7 'ti'b3
has been played, but no good
came of it after 7 . . . .txc3+ 8 be
\!rd7. E.g. 9 1lVc2 lt:lf6 1 0 0-0 0-0
1 1 h3 l:le8 + Ciocaltea-M.Kovacs ,
Baja 1 97 1 ; or 9 0-0 lL!a5 1 0 1!t'b4
lt:lxc4 1 1 11xc4 lt:le7 1 2 l:ld l "i!Yc6
=F Minev-Korchnoi, Moscow
1 960.
So we turn back to look at
Black's alternatives to 6 . . . d6.
Secti on 3
Black's 6th M ove
Alternatives
1
2
3
4
5
6
e4
lL!f3
d4
c3
eS
lL!c6
ed
de
.tb4
lL!xc3
.tc4 (130)
In addition to 6 . . . d6 Black has
also tried:
A 6
B 6 ...
C 6
D 6 ...
..
1re7
'@f6
lL! ge7
lL! f6
130
B
A
6
"i!Ye7
This was made to look extremely

bad in Joksic-Cvetkovic, Belgrade
1 966, after 7 0-0 .txc3 8 be lt:le5?
(8 . . . d6! ) 9 lL!xe5 1!Yxe5 10 1!t'b3
1lt'h5 I I e5 ! because Black can
hardly develop. The finish was I I
. lL!e7 1 2 .ta3 b6? ( 1 2 . . . lt:lc6 1 3
f4 ) 1 3 .txe7 xe7 1 4 1!t'a3+
d8 15 l:l ad 1 .tb7 1 6 "i!Ya4 c6 1 7
i.a6 ! Ilb8 (Black i s finished
because 17 . . . .txa6 is a nswered by
18 1!t'xc6) 1 8 .txb7 l:l xb7 l 9 1!t'xc6
lk7 20 "i!Ya8+ l:lc8 2 1 1!Yxa7 1!t'f5
22 l:lfe 1 11t'e6 23 Ild6 "i!Ye7 24 l:led 1
e8 25 l:lxd7 1!t'e6 26 Wb7 l -0.
The real culprit here was Black's
8th move.
.
1!t'f6
The only instance of this that I

have seen transposed into A above
after 7 0-0 .txc3 8 be lL!e5 (8 . . . d6
could hardly be worse) 9 lt:l xe5
"irxe5 1 0 "iVb3 Ciocaltea-Baretic,
Vrsac 197 1 . This time Black

tried 10 . . . 9/e7 but again landed in
an awful mess after 1 1 .ta3 c5 12
litad l ltlh6 13 litd5 b6 1 4 f3 0-0 1 5
lUd l 'it>h8 1 6 .te l ! .
c
ltl ge7
6
This looks rational, but it turns
out that f7 is too sensitive.
7
8
9
10
11
12
llJ g5
.t b3
f4
fe
be
0-0
llJeS
h6
hg
.txc3+
ltlc6
llJxeS (13 1)
13 1
w
Black's 6th Move Alternatives 119

D
6
7
An opportunity that should not

be passed over. If instead White
just plays 7 0-0, then Black is
advised to reply 7 . . . .txc3 8 be d6,
transposing into variation B of
Goring Gambit 2 .
7
dS
Notice that 7 .txc3+ 8 be d5 9

.tb3 t'Lle4 1 0 .txd5 ltlxc3 ? 1 1
.txc6+ be 1 2 1!rc2 t'Llb5 1 3 .tb2
Penrose-Butcher, London 1962, is
undesirable for Black. But 1 0 . . .
.tf5 ! ? , angling for a transposition
into Levy-Mednis below, seems
playable. So 9 ef is best here too.
...
Just as Black is about to get

things together White strikes a
shattering blow: 13 .txf7+! t'Llxfl
1 4 litxfl d5 ! (Black is groggy but
still fighting. 14 ... 'it>xfl 15 ..d5+
is a quick knockout: 15 . . . 'it>e8 1 6
.txg5 o r rs . . 'it>g6 1 6 'it'f5+ 'it>h5
17 g4+ 'it>h4 18 1t'f3) 15 l:lf2 'ird6
16 e5 'it'h6 1 7 g3 .tf5 1 8 'it'a4+ c6
19 .ta3 Penrose-Soderborg,
Budapest 1 959. As is often the
case, the opposite-coloured bishops
aid the attacker here.
t'Llf6
eS!
ef
Less convincing is 8 .tb3, with

the continuation 8 . . . ltle4 9 0-0
.txc3 10 .txd5 .tf5 1 1 be t'Llxc3 1 2
.txc6+ be 1 3 We 1 11M3 Levy
Mednis, Hastings 1 969-70 - when
14 .te3 ! ? may give White enough
for his pawn (though I doubt it).
In a later game Levy preferred 8
ef, and rightly so.
132
B
8
9
de
'irxd8+ (132)

Which way should Black recap
ture ?
1 . After 9
ltlxd8 10 fg lii: g 8 1 1
i.h6 .ixc3+ 1 2 be lt!e6 1 3 0-0-0
Levy-Karaklajic, Cienfuegos 1970,
continued 1 3
ltlc5?! 14 ltlg5
lLld3+ 1 5 Il xd3! cd 1 6 lt!xh7 rtie7
17 lite ! + .ie6 1 8 f4 f5 19 g4! . In
view of this Black has to content
himself with 13
ltlxg7. This is
sometimes assessed as equal (e.g.
in ECO) , but that underestimates
White's superior development.
Levy himself analysed 13 . . . lt!xg7
14 libe l + ltle6 1 5 .if4 as
favouring White through 1 5 . . .
.id7 1 6 .ixc7 llc8 1 7 .ig3 i.c6 1 8
lt!d4! i.xg2 1 9 f4 rtif8 20 ltlxe6+ fe
2 1 ll xe6 . I do not see that the
loss of the c7 pawn at move 15 can
well be avoided: if, for example,
15 . . . rt;f8 16 .ig3 Black cannot
develop his QB in any satisfactory
way .
2. Taking with the king has the
advantage that Black's rook does
not end up immobilised on g8.
Even so, Ljubojevic-Lombardy,
Manila 1 973 , went in White's
favour: 9
rtixd8 10 fg Ile8+ 1 1
i.e3 i.xc3+ 1 2 be rtie7 1 3 0-0-0
f6 1 4 ll he l .ie6 1 5 lt!d4 rtixg7
1 6 lt! xc6 be (Riddle: when can
you be a pawn down and a pawn
up at the same time? Answer:
when your opponent has trebled
pawns!) 1 7 i.d4+ rtig6 1 8 lle3 !.
...
...
...

6th move alternatives are not
attractive from Black's point of
view. White has a long-term plus
against 6 . . . ltlf6 after 7 e5 ! and 8
ef, and can generate a strong
attack after 6 . . . ltlge7. The very
best chance for Black in this
section is that 6 . . . 'ti'e7 may turn
out to be merely a harmless
transposition on the line 6 . . . d6 7
0-0 .ixc3 8 be lLlf6 9 e5 (Goring
Gambit 2, B) via the sequence 6 . . .
"f!/e7 7 0-0 i.xc3 8 b e d 6 9 e 5 etc.
Peel-back: 6 .ic4 is almost an
automatic choice for White, but if
you wanted to avoid all the theory
you could try 6 .ig5 ! ?. The
solitary precedent known to me is
6 .ig5 lt!ge7 7 't!fc2 d6 8 0-0-0
.ixc3 9 "f!/xc3 0-0 1 0 h4 .ie6 +
Gufeld-Stein, USSR 1 959. Our
next section looks at what happens
if Black plays 5 . . . d6 instead of 5
. . . .ib4.
Secti on 4
5
. . .
d6
...
Conclusion: In so far as they have

independent significance, these
1
2
3
4
5
e4
e5
ltlf3
lt!c6
d4
c3
de
eel
lt!xc3
d6 (1 33)
Experience has shown that this
gives White more encouragement
than 5 ... i.b4. Note, by the way,
that 5 . . . i.c5 transposes after 6
.ic4 into Scotch Gambit 1 (very
5 . . d6 121
A1
133
w
'llc8
lLld8
f4
.te7
There isn't anything better:
a) 10 . lLlh6 1 1 f5 ef 1 2 0-0 J.e 7 1 3
ef c 6 1 4 .i.d2 lLlhf7 1 5 l:tae 1 'llc 7
16 lLlxf7 lLl xf7 1 7 f6! gf 1 8 'lle 6
Pietrusiak-Bzjuszka, Poland 1965.
b) 1 0
c6 1 1 0-0 'ird7 1 2 f5 e5 1 3
lDe6 lDf6 1 4 .i.g5 h6 1 5 i. h4
Mikhalchishin-Puc, Yugoslav Ch.
1 962.
c) 10 ... h6 1 1 lbf3
c l ) 1 1 . lLlf6 1 2 lLl h4 'ild7 1 3 lD g6
l:tg8 1 4 e5 de 1 5 lDxe5 1!Yd4 1 6
i.d2 lDe4 1 7 0-0-0! Tseitlin
Pimonov, USSR 1 972.
c2) ll . lbe7 1 2 0-0 lDec6 (or 1 2 . . .
b 6 1 3 i.e3 'itd7 1 4 nac l lLldc6 1 5
lLlb5 a6 1 6 lLlbd4 F e1d m an
Obovsky, USSR 1 977) 13 lDh4
i.e7 1 4 lLlg6 l:tg8 1 5 f5 Sonner
Sacarello, Siegen 1 97 0
ef
f5 !
11
Black has two other defensive
tries:
a) 1 1 . . i.xg5 1 2 i.xg5
a 1 ) 12
f7 1 3 fe! lDxg5 14
1!Yb5+ c6 1 5 'ilxg5 g6 1 6 0-0 1i'xe6
8
9
10
g5
...
promising for White).

6
.i.c4
There are three moves worth
considering:
A 6
i.e6
B 6 . . . lLlf6
C 6 ... .te7
..
A
6
7
8
.te6
fe
.i.xe6
1i'b3 (134)
134
B
..
..
...
1 7 lLld5 ! ? cd ( 1 7 . . . 'it?d7 ! ? is less

easily disposed of) 1 8 ed 'lld 7 1 9
llae l + lDe7 20 'ilf6 Liskov
Soloviev, Moscow 1958.
a2) 1 2
f6 13 i.xf6 gf 1 4 0-0 e5
1 5 lLld5 nf8 1 6 1!Vh3 lif7 ( 1 6 . c6
1 7 1i'h5+ d7 1 8 1txh7+
Wise-Hoogendorn, Hastings 196566) 1 7 1i'h5 c6 18 lDxf6+ 'it?e7 19
lLlxh7 Mubedi-Hess, Hanau
...
. .
Should Black bother to defend
h i s b-pawn?
A1 8
1!Yc8
A2 8
'ild7 !
...
...
5 . . . d6
. 122 Goring Gambit: Section 4
1 970
b) 1 1
eS ! ? may be the best way
to resist, e.g. 12 ltld5 J.xg5 1 3
J.xg5 ltl f7 1 4 'irg3 c6 oo B onner
Rubinetti, Siegen 1 970. However,
after 1 2 0-0 Black's position
remains difficult.
12
0-0 (135)
de 20 lldl Levy. In this

variation there has been some
confusion over how White should
meet 1 2 . . . h6 1 3 lbh3 ! fe. Not with
14 '@a4+ c6 1 5 1Wxe4? 't!le6 =t= de
Visser-Belder, corres 1 970. Instead
very strong is 14 ltlf4 ! , e.g. 14 ...
lDf6 15 l0g6 l0f7 1 6 l0xe4! .
135
B
A2
..
'4i'd7!
In such positions this is usually

preferable.
9
10
11
llb8
J.e7
J.f6 (136)
't!fxb7
't!la6
0-0
A less harmonious development

results from 1 1 .. lbf6 as in
K1ovan-To1ush, Riga 1 962: 1 2
lld 1 0-0 1 3 1We2 e 5 1 4 ltld5 ltlxd5
15 ed l0d8 1 6 J.e3 .
.
Any true gambiteer should

scent blood in this position.
Black's king is in the centre, his
pieces are not working together,
and there are wide avenues into
the heart of his position. The
defender is not likely to survive
White's onslaught . For example:
a) 12
fe? 13 ltlf7 .
b) 12 . g 6 13 ef gf 14 lbd5 !
intending 'irc3 .
c ) 12 . ltlh6 1 3 ltl d 5 J.xg5 1 4
J.xg5 ltlhf7 1 5 J.xd8 Wxd8 1 6
llxf5 1Wd7 1 7 1rxb7 K1ovan
Darsnik, Riga 1 962.
d) 1 2
h6 13 ltlh3 ! (intending
f4) 13 . . . lDf6 14 lDf4 (stronger
than 14 ef c6 1 5 J.e3 1Wd7 16 ltlf4 ,!
Liptay-Bokor, Budapest 1 960) 1 4
. . . c6 1 5 lbg6 llg8 1 6 e f lb f7 1 7
J.e3 'lrct 7 1 8 litae l l0e5 1 9 l0 xe5
136
w
..
..
Compare this position with the

resultant of variation A in Goring
Gambit 2 (Diagram 1 26). Velimir
ovic's suggestion of 12 J. gS?!
could almost lead to a transposition
via 1 2 . . . J.xc3 1 3 be lOge 7, the one
difference being that Black would
be a tempo down. Howeverj Levy

convicts 1 2 .i.g5? ! of wilful neglect
of the b-pawn ( 1 2 . . . Ilxb2 with . . .
Ilb6 available) and I bring i n the
same verdict.
So how should White play?
I . He got a big plus in S mith
Bisguier, USA I 970, with 1 2 li[d i
lt:Jge7 1 3 't!fe2 0-0 ? 14 e5 ! . 1 3 . . .
lt:Jg6 i s an obvious improvement,
with chances for both sides.
2. The most interesting continuation
is 12 eS! ? lt:lxe5 1 3 lt:l xe5 .i.xe5 I 4
't!fxa7 Ilc8 of Levy-Feller, Praia
da Rocha 1 969. A difficult ,
double-edged position: Black has
an imposing centre to set against
White's outside passed a-pawn.
The game continued I5 W'e3 lt:lf6
1 6 f4 lt:Jg4 I 7 W'e2 j.d4+ I S ct>h i
lt:Jf6 I9 lt:le4 0-0 20 lt:Jg5 e5 2 I f5
(2 I lt:Jf3 ! ?) 2 I . . . c5 22 ll:le6 Ilf7 2 3
j_g5 d5 oo. I would prefer Black's
chances, though it was actually
White who won.
B
6
lt:\(6
7 'tib3
1t'd7
The defiant 7 . . . lt:laS should
always be examined when White
lines up on f7 like this. Here the
answer would be 8 .txf7+ ct>e7 9
1t'a3 r3;xf7 IO W'xa5, which is
awkward for Black because as
soon as White has played 0-0 he
threatens e4-e5 .
8 lt:lgS
lt:leS
Even though it is clear that the
5 . . . d6 123
knight on e5 is soon going to be

kicked by f2-f4 this is better then
the abject 8 ... lt:ld8 9 f4 and now:
a) 9
h6 10 lt:lf3 .i.e7 I I e5 lt:lh7
I 2 .i.e3 intending 0-0-0 Botterill and Harding.
b) 9 ... c6 10 e5 ( 1 0 0-0?! b5 ) 10 . . .
de ( 1 0 . . . d 5 I I i.d3 ! ) I I fe lt:lg4
12 i.f4 't!ff5 13 Ilfl ll:le6 ( 1 3 . . . h6 ?
I 4 ll:lxf7) I 4 lt:Jxe6 fe (Smit) and
n ow just I5 0-0-0 gives White an
overwhelming position.
9
10
.tbS
f4 (137)
c6
137
B
The position poses a problematic

decision between:
B1 1 0 ... cb
B2 1 0 . . . lt:l eg4
Against 1 0 . . . lt:lg6 the obvious
I I .tc4 is not so good because of

I I . . . d5! 1 2 ed .i.c5, Lutikov
Lisitsin, USS R 1 962, or I2 lt:lxd5
ltlxd5 1 3 ed .tc5 with a dangerous
counterattack. Kajkamdjozov sug
gested 1 1 .td3 h6 I 2 ll:lf3 .i.e7
13 J.d2 intending 0-0-0, but more
incisive is Velimirovic's idea 1 1 eS!
de 1 2 .tc4 or 1 1 . . . cb 1 2 ef gf 1 3
lt:lge4 .
5 . . . d6

B1
10
cb
lbg4
11
fe
The older move 1 1 ... de should
cost Black the game after 12 i.e3!
(138)
138
B
the check at b4 is a saving resource

enables us to discover the rebuttal
13 0-0! , e.g. 13 . . . a4 14 'W'xb5
intending Itd 1 or 1 3 . . . .tb4 14
lld 1 'W'c7 ( 14 . . . 1!Ve7 15 liJd5
liJxd5 16 llxf7) 15 liJxb5 'W'e7 16
a3 .tc5 17 .txc5 9xc5+ 18 rl;;h 1
0-0 1 9 liJxf7 ! :: .
12
h3
A slow move, but best in view of

a) 12 0-0 de 13 lbxf7 .tcS+ 14 <Sh 1
llf8 =F
b) 12 e6 fe 1 3 liJxb5 (intending
liJxe6 - Levy) 1 3 . . . a6 1 4 lbd4 e5
=F
a) 12
i.d6 1 3 lld l (but not 1 3
0-0-0 0-0 1 4 i.c5? when Black can
happily let the queen go: 14 . . .
.txc5 1 5 llxd7 .txd7 =F ) 1 3 . . . 0-0
14 lb xb5 lbe8 1 5 0-0 1!Ve7 1 6 lbxd6
lbxd6 17 1fa3 lld8 ( 1 7 . . . h6 1 8
l!Jf3 lld8 1 9 .tc5 ::) 1 8 lbxt7 !
:: Alekhine-Verlinsky, Odessa
19 1 8 .
b ) 1 2 .. . a 6 1 3 ll'd 1 "ifc7 ( 1 3 . . . 'fle7
14 .tc5 forces 14 . . . 1!Vc7 in view of
14 . . . .te6 1 5 .txe7 .txb3 1 6 .txf8
.txd 1 1 7 .txg7 ::) 14 .tb6 1!Vc4
1 5 lld8+ rtJe7 1 6 'W'd 1 Wc6 1 7
liteS+! and mates is a pretty piece
of analysis by von Minckwitz.
c) 12 . . . a5 1 3 lld1 a4 14 'W'xb5
'W'xb5 1 5 lbxb5 .tb4+ 1 6 rl;;f l 0-0
1 7 liJc7 llb8 1 8 j,a7 .tg4 ! =F is a
variation given by Sosonko and
van der Sterren (New in Chess 1 )
which might appear t o overturn
previous theory. But noticing that
12
13
14
15
0-0
lbxe5
f6
lbxf3+
lbf3
Itxf3 (139)
White has full compensation

for the two pawns; Black may be
able to develop or he may be able
to keep his material advantage,
but he can hardly do both. In
Levy-U nzicker, Hastings 1 969-70,
the sequel was 15 . . . b6 16 lt!d5
.tb7 1 7 llxf6 .txd5 1 8 1Wxd5 gf 1 9
"ti'xa8+ <M7 2 0 .tf4 llg8 2 1 1!Vd 5+
11re6 and now, instead of 22 1fh5+
5 . . d6 125

<jJg7 23 1tf3 J.e7 24 lilc l Iic8
Y2-Y2 , White should play 22 Ii c l !
when he wins back the second
pawn with an advantageous end
game.
B2
10
11
lileg4
J.e2!
The tempting 11 i.c4 is again

dubious because of 11 d5 ! 1 2 ed
i.c5 =F. 1 1 h3 cb 12 hg h6, as in
Bronstein-Fuderer, USSR-Yugo
slavia, Kiev 1959, should go in
White's favour after 1 3 lil xb5 !
(Kajkamdjozov). However, an
argument against I I h3 is the
improvement 12 ... b4! , as in
Ljubojevic-Smej kal, Wijk aan Zee
1972: 12 . . . b4 1 3 li:ld5 ( 1 3 'ilxb4
d5) 1 3 . . b6 14 lilxf6+ gf 15 Ii xh7
Iixh7 1 6 ltl xh7 i.e7 (140)
...
Wxf6 Wxe4+ =F.

11
h6
A recent attempt was 11 d5! ?

1 2 h 3 lil h 6 1 3 e5 ltle4 14 ltlcxe4 de
15 .te3 .te7, Minguell-Femandez,
Barcelona 1 983, when I think 16
.i.c4 and 1 6 g4 are both .
d 5 (141)
ltl f3
12
He has to do something to
rescue the knight on g4 from the
threatened h3. 12 . . . h5 does not
appeal.
141
w
1 40
w
1 7 g5 fg 1 8 f5 'ilc6 1 9 't!t'xb4 f6 20
't!t'd4 and now Black dallied with
20 . . . 'llfl ? (he must have been in
time trouble), which quickly proved
fatal: 2 1 J.e3 J.b7 22 li c l Wxe4 23
Iic7 lilh8? 24 'ilxf6+ l -0. Correct
was 20 . . . J.b7 2 1 ltlxf6+ J.xf6 22
Now 1 3 ed is met by 1 3 . . . J.c5

with counterplay. But it is thought
that White can secure an advan
tage in either of two ways:
a) 13 e5 ltlh5 1 4 ltld4! (intending
15 e6 or 15 h3) 14 . . . 1!fd8 15 g3
1Wb6 (or 1 5 . . . i.c5 same reply) 1 6
'it'd l Euwe.
b) 1 3 h3 d4 with the alternatives:
b l ) 14 h g ! ? de 1 5 lile5 't!t'e7 16 J.c4
J.e6 1 7 .txe6 fe oo. The position
was analysed in depth by C. B .
Wood, who revealed that 1 8 ltlg6
1lt'b4! is surprisingly strong for
Black. But I think 1 8 be probably
does favour White.
b2) 14 lild1 ! ? lile3 1 5 ltlxe3 de 16
l0e5 'fle7 1 7 .ic4 .ie6 1 8 .txe6

tfxe6 19 'tlbb7 according to an
analysis by Kenneth H amilton .
However, 1 8 . . . fe is again rather
unclear and, at any rate, no better
for White than the similar position
he gets in b2 .
c
6
7
8
9
.te7
ltJaS
tfb3
i.xf7+
'i!tf8
't!t'a4 (142)
142
B
Well, there must be some

variation in which .txf7+ is
strong and perhaps this is it ! It is
amazing how harmless that threat
often turns out to be.
9
'i!txf7
I would say that 9 . . c6 errs by

preserving the wrong knight: 10
.ixg8 litxg8 11 0-0 and now:
a) 1 1
h6 1 2 .if4 b5 13 9c2 .ig4
1 4 l0d4 .if6 15 lDf5 .txf5 1 6 ef d5
17 life 1 f7 1 8 lite6 P enrose
Szabo, Hastings 1956-57.
b) 11
bS 12 twc2 .ig4 1 3 l0d4
.if6 1 4 .ie3 '@e8 1 5 f3 .id7 1 6
llad l l0c4 1 7 .te l Varnusz.
..
5 ... d6
Szabo , Hungarian Ch 1 96 1 .
c) 1 1
'i!tf7 (Levy) is u ntested, but
does at least get the priorities
right.
10
't!t'xaS
c6
Tal-Russel, Munich 1958, was

too easy for White: 1 0 . . . .ie6 1 1
0-0 'i!tf8? (a step in the wrong
direction) 1 2 l0d5 ! c6 1 3 l0c7 i.f7
1 4 l0d4 .
11
12
't!t'xd8
.if4 t
.ixd8
Braun-Kostro , Kienbaum 1 958.

White has a significant lead in mobi
lisation and the Black d-pawn is
weak.
Conclusion: If White could steer
Black into the variations examined
in this section, then the Goring
Gambit would be in business . In
fact the only line after 5 . . . d6 that
gives Black a reasonably safe
route to a playable midd1egame is
A2.
Peelback: The vital question now
is whether White can play 5 i.c4
instead of 5 lDxc3. For by
adopting that move order he
hopes to avoid all the lines with 5
lDxc3 .tb4.
Section 5
5 i. c4
1
2
3
e4
l!Jf3
d4
eS
l!Jc6
ed
5 .ic4 127

4
c3
de
1.c4 (143)
5
143
B
lbf3 1.b4 etc. But that sequence

has the defect of allowing 5 . . . d5 ! .
lDf6
7 lb c3
8 't!Vc2
8 e5?! fails to 8 . . . d5 9 ef 'ti'xf6!
10 0-0 1.xc3 1 1 bc3 "t!Vxc3 1 2

't!Ve2+ .ie6 1 3 1.xd5 0-0 =F Stein
Spassky, Tallinn 1 959.
8
9
d6
0-0
0-0-0
.ixc3 10 11t'xc3 .ie6 1 1
9
libe l .ixc4 1 2 11xc4 0-0 1 3 e5
..
Now, if Black would only oblige

with 5 .. d6, White plays 6 ltJxc3
and he has done the trick. White
has lots of attacking chances (see
the previous section) and he has
avoided our main line with 5 . . .
.ib4. But there i s a second pawn
on offer, and why shouldn't Black
take it?
.
5
6
cb
1.xb2
lbe8 is an impressive attacking

position for White. However,
game experience and reliable
analysis are lacking. The Handbuch
gave 14 h4 't!fc8 1 5 e6 fe 1 6 ll xe6
cii>h 8 17 lbg5 lbf6 18 't!fd3 't!fd7 1 9
llde l ( ?). Fine suggested 1 4
lie3 ! ? .
10
e5
lbg4 (1 44)
144
w
Black's choice here lies between:

A 6 ... .i b4 +
B 6 ... d6
A leads to fantastic complications,

but, unfortunately for White, B
just seems to be solid and good.
A
6
.ib4+
Thus we enter the cut and thrust

action of the Danish Gambit,
introduced by Severin From in
Paris 1 867. The traditional move
order - a sort of accelerated
Goring - is 1 e4 e5 2 d4 ed 3 c3 de 4
1.c4 cb 5 .ixb2 lbf6 6 ltJc3 lbc6 7
One cannot really hope . .to

produce a definitive analysis in
such a complicated position. But I
think that any player with attacking
flair ought to be happy with
White's chances.
11
lbd5
Possibly good, but very messy is
5 .t.c-4

1 1 h4, e.g. 1 1 . . . l0cxe5 1 2 il:l g5 g6
1 3 l0ce4 .tf5 14 'iVb3 and if now

14 . . . .txe4 15 il:l xe4 il:l xc4 1 6
1Wxc4 .ta5 1 7 f3 ::: (Botterill and
Harding).
11
12
.tc5
ed
The sacrifice 12 l0f6+ ! ? has

been played, but I don't think it
can really be sound: 1 2 . . . gf 1 3 ef
.txf2 (to answer 14 llJ g5 with . . .
.te3+) 14 lithe 1 .te3+ 1 5 b 1 ( 1 5
litxe3 ! ?) 1 5 . . . .th6 oo . Black can
also defend with 12 . . . l0xf6 13 ef
gf 14 g4 il:le5 1 5 g 5 ! ? 'iVd7 !
(Botterill and Harding).
12
13
cd
h4
White could also try 1 3 lithe !

intending 1Wc3.
h6
13
Since this does n o t stop il:lg5

one might wonder whether it is the
best defence. 13
llJce5 comes
into consideration too. I am no
longer convinced by 13 . . . il:lce5 14
il:lg5 g6 15 llJe4! ? (Botterill and
Harding) 15 . . . .tf5 oo .
...
14
15
16
17
18
145
B
hg
llJ g5!
hg
1Wxg5+
.trs
f4
.txc2
fg
xc2 (145)
White is still two pawns down

and the queens are off the board,
but the attack persists thanks to
the activity of White's pieces, the
open h-file and the important g5
pawn. Here are some of the
possibilities:
a) 1 8 b5 1 9 .td3 f5 20 g6 ! il:le3+
21 llJ xe3 .txe3 22 .txb5
Ashcroft-Harding, corres 1 97 1 72.
b) 1 8
litfe8 19 lith5! ( 1 9 lith4
lite4!) 19 . . . il:le3+ (but now 19 . . .
lite4 2 0 .td3 ) 20 il:lxe3 .txe3 2 1
litdh 1 f8 22 lit e 1 ! g8 2 3 g6 .
c) 1 8 ... llJ ge5 19 lith3 l0g6! is
unrefuted since 20 llJf6+ gf 2 1
.txf6 intending lit h 1 and lith8 can
be met by 2 1 . . . .td4 - though
White still has compensation for
the pawn after 22 .txd4 l0xd4+ 23
litxd4.
..
6
146
w
d6! (1 46)
Black's 8th Move Alternatives 129
It is the doom of the Goring

that against this simple little move
White lacks any convincing attack
ing plan.
14 liJc3 c6 I S liad l llg8 1 6 lL'lf3

'tWf7 =t= Valdes-Jensen, corres
1 972-74.
0-0
Two other moves have been

tried:
a) 7 1!Vb3 lL'laS ! (As usual ! 7 . . . Vd7
and 7 . . . ll:l h6 are obscure) 8
.txf7+ <tie7 9 "ti'dS c6 (but not 9 . . .
lL'lf6? 1 0 i.xf6+ gf 1 1 .thS or
10 . . . <tixf6 1 1 "ti'gS+ winning the
queen) 10 'W'gS+ lL'lf6 and the
attackers are in a tangle, e.g. 1 1
i.hS liJc4 1 2 .tc3 h6 1 3 'tWh4
(what else?) 1 3 . . . gS 14 liJxgS hg
I S WxgS liJeS ! 16 i.xeS 'tWaS+
+F.
b) 7 lL'lc3
b 1) 7 . . . i.e7 8 'W'b3 lL'l h6 (8 . . .
lL'laS ! ? i s probably the best here as
well) 9 liJdS f6 10 0-0 liJaS =t=
Csom-Barczay, Hungary 1 967.
b2) 7 . . . i.e6 8 lL'ldS Vd7 (8 . . .
liJaS ! and 8 . . . lL'l e S ! both look
good to me) 9 llcl liJge7 1 0 0-0
lL'lg6 1 1 'tWa4 ! and Black was in
trouble in Ljubojevil:-Kovacs, Sara
jevo 1 969.
7
8
.txe6
1!fb3
fe
Wd7!
One of the advantages of taking
the b2 pawn becomes apparent: 1 0
1!rxb7?? lib8 H .
10 liJg5
lL'ld8
11
f4
lbf6 (147)
White does not have much to
show for his pawns after 1 1 ... h6
either - e.g. 12 1Jfh3 eS! 1 3 fS lL'lf6
147
w
White has a big lead in

development, but what can he
make of it? So far as I can see, he
just bruises himself on Black's
wall of pawns. Experience and
analysis supply the following
data:
a) 1 2 f5 eS 1 3 ll:lc3 h6 14 lL'le6 c6 1 S
liad l lL'lxe6 1 6 fe @c7 =t= E . Szabo
Kocsis, corres 1970. The rest of
the game was really depressing for
the gambit player: 1 7 Cibh l i.e7 1 8
1!rc4 @aS 1 9 a4 1WcS 20 @b3 Wb6
21 @a2 liJg4 22 .tel llf8 23
llxf8+ <trxf8 24 llfl + <tig8 2S h3
lL'lf6 26 llxf6 .txf6! (26 . . . gf? 27
We2 gives White hopes) 27 e7+
<tih7 28 Vf7 @f2 29 e81!r lixe8 30
'fixeS .th4! 0- 1 . A game which
neatly displays the difference
between the ghost of an attack and
the real thing.
b) 12 e5! ? de ( 1 2 . . . liJdS 13 fS ! ) 1 3
fe liJdS 14 lL'ld2 gave a misleadingly
rosy picture of White's chances in
Krantz-Sellberg, corres 1 974: 1 4
. . . 11fc6 1 S h 1 1t'b6 1 6 Wa4+ \!Vc6

1 7 1fe4 e7 1 8 \!Vg4 hS 1 9 1Wh 3
11fc2 20 lilad 1 ! ll:lfS (20 .. ; 'ifxb2 2 1
c4 and 2 2 xe6 ) 2 1 de4
'ifxb2 22 xe6 1!YxeS 23 lil xfS !
'ft'xe6 24 JileS 1 -0. The true
perspective is restored by 14 ...
J.e7!, e.g. 1S ll:lde4 h6 =F or 1 S
ge4 f7 1 6 1!t'xb7 0-0 =F or l S
Wg3 J.xgS 1 6 1!YxgS 1re7 =F.
which is quite understandable if

Black has not come fully prepared
for all the complicated tactical
play that may result from acceptance.
There are three main ways of
declining:
A 4 . . . ll:l f6
B 4 . . . dS
c 4 ... d3
A
4
Conclusion: S J.c4 does not give
White sufficient compensation for

his two pawns against 5 . . . cb 6
J.xb2 d6 ! .
Section 6
The G a m b it
Dec l i ned
1
2
3
4
e4
ll:lf3
d4
c3 (148)
S
eS
c6
ed
148
B
Although it is really stronger to

. accept the gambit, this section has
considerable practical importance.
Over the board the gambit is
declined more often than not,
f6
The most interesting possibility

and one that can quickly reverse
the roles by turning a gambit into
a counter-gambit. It can also arise
from the Ponziani Opening via the
move order 1 e4 e5 2 f3 c6 3 c3
ll:lf6 4 d4 ed etc.
eS
e4
The only other move worth

considering is 5 ... 1We7! ?:
a) 6 1!Ve2? d3 ! is the point. If then 7
ef de 8 fe efW+ 9 llxfl J.xe7 H.
b) 6 cd d6 7 .i.bS (7 1!Ye2? de 8 de
ll:lg4 9 J.f4 'tWb4+ 10 1!fd2 i.cS =F)
7 . . de (7 . . . ll:ld7? 8 0-0 de 9 dS
cb8 10 xe5 Chigorin
Winawer, Monte Carlo 190 1) 8
0-0 and I think Black is in
difficulties, e.g. 8 . . . J.d7 9 de
xe5 10 xeS J.xbS 1 1 ll e 1 or 8
. . . e4 9 e5 J.d7 1 0 .txc6 J.xc6 1 1
c3 .
6 We2
fS! (149)
Lasker's idea. Others:
a) 6 . cS 7 cd e6 8 d5 ed4 9
xd4 xd4 1 0 1!fe4 Chigorin
Vainstein and Navarovsky, Moscow
1904. I had the pleasure of this
.
..
The Gambit Declined! 131
Lasker wrote: 'White's task is

by no means simple. For instance
8 liJfd2 'irxf6 9 f3 d3 1 0 1!re3 'Wg5
and White has had his trouble for
nothing.' But it is the other knight
that should go to d2.
myself more recently: 1 0 . . . .i.b4+

1 1 o!Dc3 c5 12 .i.e3 'W'e7 13 0-0-0 d6
14 ed 'W'xe4 15 liJxe4 .i.f5 1 6 f3
0-0-0? ( 1 6 . . . :c8 was a must) 1 7
a3 .i.a5 1 8 b4 .t.xe4 1 9 fe .i.b6 20
b2 lbb3 2 1 .i.c4 1-0 Botterill
Trevelyan, Gwbert 1977.
b) 6 . . . d5 is t ricky but inadequate:
7 ed f5 8 o!D xd4 liJ xd4 9 cd .i.xd6 I 0
f3
b l ) 10 . . . 1!fh4+ 1 1 g3 .t.xg3+ 1 2 hg
'W'xh 1 and now both 1 3 fe 'W'xe4 1 4
'iVxe4+ fe 1 5 .i.f4, Sanguinetti
Re inhardt, Mar del Plata 1 95 8 ,
and 1 3 o!D c 3 0-0 1 4 fe fe 1 5 .tf4,
Velimirovic-Trifunovic, Yugoslavia
1 963, are .
b2) 10 . . . .i.b4+ I I liJd2 Wxd4 12 fe
fe 1 3 a3 - Botterill and
Harding.
This has superseded:

a) 8 fg ? .i.xg7 9 liJxd4 0-0 10 .i.e3
lbxd4 1 1 cd liJxf2! 12 .t.xf2 :e8 =F
Casa-Boey, Lugano 1 968 .
b) 8 liJxd4 liJ xd4 9 cd and now:
b l ) 9 . . . f7! 10 fg .i.b4+ 1 1 d 1
( 1 1 liJd2 1Ie8, Zwaig-Boey, Varna
1 962, gives Black an even stronger
attack) 1 1 . . . lle8 12 .i.e3 g8
and Black has very active play.
Levy-Boey, Siegen 1 970, continued
13 1!rh5 .i.e6 14 J.d3 'ti'd7 1 5 h3
.i.f5 1 6 .i.c2 (150)
149
w
/50
B
ef
7
7 lbxd4 is dubious because of 7
. . . .t.c5 ! 8 liJxf5 0-0 9 'ti'xe4 d5 1 0

ed .i.xf2+! 1 1, xf2 .i.xf5 with a
tremendous attack (Wade). lt was
Wade who reintroduced 4 . . . liJf6
into tournament play in the 1 95 5
British Championship.
7
d5
liJbd2! ?
1 6 . . . c5 (it has been suggested

that 1 6 . . . liJ g3 ! ? would have been
more potent) 17 1!rf3 cd 18 .t.xd4
1Iac8 1 9 .i.b3 :c4 20 .i.e3 liJc5 2 1
.i.xc5 (2 1 .t.xc4? loses to 2 1 . . . de+
22 liJd2 .i.xd2 23 .t.xd2 'Wa4+ 24
b3 o!Dxb 3) 2 1 . . . .txc5 22 liJc3
:d4+ 23 c 1 :d3 24 liJxd5 !
xg7 25 1tf4! Jlxb3 26 gS+
132 GiiTing Gambit: Section 6
The Gambit Declined
'it>h8 27 \i'f6+ 'it>g8 28 \i'g5+ 'it>h8
lO f7+! 'it>xf7 I I lll g 5+ (152)
- .
152
B
b2) 9
J.b4+ l O J.d2 .txd2+ I I
lt!xd2 0-0 I 2 lll xe4 lieS ! I 3 0-0-0
llxe4 I4 \i'h5 g6 I 5 f7+ ( I 5 \i'g5
J.f5 I 6 f3 lle6 I 7 g4? ll xf6!
intending . . . llc6+ Wade) I 5 . . .
'it>g7 I 6 f8\i'+ \i'xf8 I 7 \i'xd5
\i'f4+ I S 'it>b i (15 1)
...
151
B
11
'it>f6?! (A mysterious king
move . Surely l i . . . e8 ! was
better. If then 12 lll x e4 'fke7 is not
bad for Black) 1 2 g4 h6 13 h4 .ig6
14 lt!xe4+ .txe4 1 5 'fkxe4 \i'e7
(loses, but so does everything else
by now) I6 .tg5+ ! hg I 7 hg+
'it>xg5 1 8 f4+ 'it>f6 19 g5+ 20
J.c4+ 'it>e8 2 1 1!t'xe7+ lll xe7 22
litxh8 de 23 lith2 cb 24 llxb2
b6 25 lld l c6 26 lii: e 2 b5 27 J.b3 c5
28 .te6 g6 29 f5 1 -0.
Another possibility is 8 "ti'xf6
9 lll xe4 de 1 0 1i'xe4+ 'fke6, which
is reasonably safe (and hence
hardly in the spirit of this lively
variation ! ) but perhaps not quite
equalising: 1 1 J.d3 ! ( 1 1 \i'xe6+
J.xe6 1 2 cd 0-0-0 is easy for
Black) 1 1 . . de 1 2 0-0 \i'xe4 1 3
J.xe4 .td7 (Against 1 3 . . cb
Velimirovic gives 14 .txb2 J.d7 1 5
.txc6 intending lii: fe l ' '. However,
I don't see why this should be
worse than the game) 1 4 J.xc6
.txc6 1 5 lii: e l + 'it>f7 1 6 lt!e5+ 6
1 7 lt! xc6 be 1 8 be t Velimirovic
Ree, A msterdam 1976.
. . .
Wade here proposes I S . . . c6 I 9

"t!fb 3 llxd4 as a n improvement
on IS . . J.f5?! I9 'it>a i liaeS 20
\i'xb7 lite I 2 I J.b5 Gheorghiu,
Goulandris and Vrondissis v.
Wade, Georgopoulos and Rose,
consultation game, London 1 972.
b3) 9
\i'xf6?! lO f3 .tb4+ is
reminiscent of something we have
seen before 6 . . d5 instead of
Lasker's 6 . . f5 . Black does not get
enough for his piece after 1 1 J.d2 !
.txd2+ ( 1 1 . . \i'xd4 1 2 \i'b5+) I 2
lt!xd2 \i'xd4 1 3 fe "ihb2 1 4 \i'b5+.
=
...
d3
Boey recommended this. 8

.tf5 was roughly treated in
Diickstein-Wittmann, Kapfenberg
1976, though there is an obvious
improvement available: 9 lt!xe4 de
...
...

9 'ife3
Naturally White could just play
9 'ifxd3 lDxf6 - probably roughly
equal, but untested . There is also
the po ssibility 9 fg .i.xg7 10 'ifxd3,
when White is a pawn ahead, but
this looks hazardous after 10 . . .
.i.f5 .
.i.c5
9
Now that White can capture on
d3 with the bishop on fl 9 .i.f5 is
less likely to be a useful move, e.g.
9 . . . .if5 10 .ixd3 .i.c5 1 1 fg lii: g8
1 2 lDd4 lhg7 1 3 .i.xe4 de 1 4 ltJ xc6
.ixe3 15 lbxd8 .i.xd2+ 16 .i.xd2
Supancic-Flear, London 1 978.
1Ig8
fg
10
ltJd4 (153)
11
...
153
B
The Danish IM Iskov has had

two disasters with the white pieces
from this position:
1. Iskov-Hebden, Benedictine Inter
national, Manchester 1 979: 1 1
1!e7 1 2 ltJ2b3 lb xd4 1 3 cd .i.b4+
14 d l ?? ltJxf2+! 0- 1 . For if 1 5
'ifxf2 .i.g4+.
2. Iskov-Kaiszauri, Oslo 1 980: 1 1
. . .i.xd4 1 2 c d .i.f5 1 3 .i.xd3 ( 1 3
...
f3? lDb4 +) 1 3 . . . 'ife7 14 .i.b5

0-0-0 1 5 .ixc6 be 16 lD xe4 de
(Kaiszauri observes that 16 . . .
.i.xe4! ? was i n fact possible a s 1 7
f3 ?! \!fxg7 1 8 fe allows 1 8 . . . l!t"xg2
19 llfl 1Idf8 ! 20 'ife2 lii: x fl + 2 1
'ifxfl 1!xe4+ 2 2 d l 'ifxd4+ 23
c2 llgl +. So White should play
1 7 0-0 'ifxg7 1 8 g3 oo) 1 7 'ifc3 e3 ?!
(j ust 1 7 ... 1Ixg7 - Kaiszauri) 1 8
.i.xe3 lixg7 1 9 'ifxc6 (Kaiszauri
laconically gives ' 1 9 0-0 ! ' . I take
the point to be that the obvious
counter 19 . . . .ih3 is met not by 20
.i.g5 ? .i.xg2 2 1 .i.xe7 .i.f3+ and
mates, but by 20 g3 ! .i.xf1 2 1 lii: xf 1
when White's advantages in quan
tity and quality of pawns outweigh
the exchange) 19 . . . .i.e4 20 1!t'c5
1!f7 21 llc l 1Ixg2 22 1!t'e5 .i.b7
(not 22 . . . 1Ie8? 23 1Ixc7+ ! ) 23
1Ic5 lig4 24 f3? (24 llfl was
better) 24 . . . 1Ig2 25 .i.f4 1Id7 26
d l .i.xf3+ 27 c l lig4 28 .i.g3
.i.xh 1 0- 1 .
These results make the position
of the last diagram seem unhealthy
for White. But is that really so? In
the second game White's loss is
not primarily attributable to the
opening. Meanwhile, Iskov m ust
have thought that there was a
worthwhile improvement on his
game with Hebden, or he would
not have repeated the line. Well,
14 .i.d2 certainly improves on 1 4
d 1 ?? , though whether it is good
for White is another matter. An
interesting possibility is 12 .i.xd3! ?
lDxd4 1 3 c d .i.xd4 14 1!rxd4 lDg3+

1 5 d 1 ltlxh 1 1 6 lt:Jf3 ; Black has a
problem about getting the knight
on h i out, but it is unfortunate
that the king is one of White's
centralised pieces.
All I cim say is that the 4 . . . ltlf6
line offers chances for both sides
and needs further investigation. I
do not know of any secure way for
White to get a plus.
B
d5
The most common way of

declining the gambit, not so frisky
as 4 . . . ltlf6, yet bespeaking a less
craven tendency than 4 . . . d3.
5
6
ed
The Gambit Declined

middlegame accentuates the dy
namic potential of White's isolated
d-pa wn, and the half-open c-file is
bound to pose a threat to Black's
king after . . . 0-0-0.
I will slice variations into:
Bl 6 ... lt:Jf6
B2 6 ... .tg4
.i. b4 +
B3 6
.
- with the warning that one must

be on the alert for transpositions.
The Capablanca plan (B3) involves
. . . .i.g4 and . . . .i.b4 and it does
not matter much which Black
plays first, though he m ight wish
to lure White into B2 1 with 6 . . .
.tg4.
'ti'xd5
cd (154)
154
B
Bl
6
7
8
lt:Jf6
ltlc3
.i.b4
lt:Je4
.tel
Black needs to play actively
here as White is better after 8
.
0-0 9 0-0:
a) 9 ... 'tWaS 10 .i.d2 l::t d 8 ( 1 0 . . .
Although Black has tried a

number of different moves, the
basic choice is between two policies.
Either he can go in for a middlegame
fight with . . . 0-. or else he can
seek simplification through ex
changes designed to reduce White's
attacking capacity. Capablanca
favoured the second course and he
was probably right. A complex
ltl d 5 1 1 ltlxd5 'ti'xd5 1 2 .i.xb4

lt:Jxb4 1 3 lt:Je5 Reti-Breyer,
Baden 1 9 1 4) 1 1 a3 .i.e7 12 d5 !
(the isolated pawn shows its
strength: if 12 . . . ltlxd5 1 3 lt:Ja4
) 1 2 . . . ltlb8 1 3 .i.c4
Penrose-Fairhurst, Glasgow 1955.
b) 9 . . . 'ifd8 1 0 .tg5 h 6 1 1 .i.h4
.te7 1 2 l::t c l .i.g4 13 lt:Je5 ! .i.xe2 1 4
ltlxe2 lt:Jb4 1 5 'ti'b3 ;!;: Velimirovic
Holmov, Yugoslavia v USSR,
Sukhumi 1 966.
9
10
.i.d2
be
.i.xc3
lt:Jxd2
A more enterprising plan for

Black is to avoid this exchange
and play for s trength on c4 and e4
with . . . ti)aS, . . . bS and . . . fS .
This procedure never seems to be
adopted in modern tournaments,
but there is a fascinating historical
precedent, Nyholm-Alekhine, Stock
holm 19 1 2: 1 0 . . . 0-0 I I 0-0 ti)aS 1 2
liel b S 1 3 .td3 fS 1 4 a4 llb3 (155)
Ljubojevic-Szabo, Wijk aan Zee

1 973.
12
13
b6
0-0 (156)
156
B
155
w
I S lia3 ba 1 6 l ha4 ti) bxd2 1 7

lJxd2 lJxc3! 1 8 \!t'c2! ( 1 8 Wa 1
'i!Vd7 1 9 liaS ..-xd4 20 .tc4+ <&11 8
2 1 lidS ti) e2+ !) 1 8 . . . 'it'd7 1 9 :laS
..-xd4 20 llc l lidS 21 'it'b3 + .te6 !
(21 . . . ti)dS ? 22 lJf3 'i!Vf4 23 li xdS !
't!t'xc l + 24 .tfl ) 22 1!Vxe6+ h8
23 lieS (23 ti)b l ..-xd3 24 lJxc3
1!Vxc3 !) 23 . . . 'it'xd3 24 lice 1 h6 2S
't!Vg6?? (25 llf3 ! ) 25 . . . \!t'xd2 0- 1 .
Did White miss something powerful
here?
11
12
...d2
0-0
lib1
White has frequently castled

first, but that allows . . . .tf5 , e.g.
12 0-0 .tf5 13 ll fe l i.g6 1 4 't!t'b2
b6 1 5 i.b5 ( 1 5 lJd2 ! ? intending
.tf3) 15 . . . a6 16 .tfl 't!t'd6
=
Perhaps one cannot speak of an

advantage in this position, but
White's game is slightly more
pleasant to play. Exampl1 . 13 . .. li) a5 14 't!t'f4 ...a-6 1S 'it'h4
.tf5 16 lib5 t Velimirovic-Smejka1,
Yugoslavia 1 970.
2. 1 3 ... 't!t'd6 14 .td3 h6 (definitely
inferior is 14 . . . .tg4? 1 5 lCJgS h6
16 ti)e4 't!Vd8 17 f4 Velimirovic
Tringov, The Hague 1966) 1 5
libe l .tg4?! (Still not good.
Smej kal gives 15 . . . .id7 16 lJe5
ti)xeS 17 de co . More sensible for
White would be 1 6 lie3 liae8 1 7
llfe l t Klovan-A verbakh, USSR
Ch 1969) 16 'it'e3 ! :ll a d8 17 .tb l
(or, slightly better, 1 7 .tc2 ! ) 1 7 . . .
.txf3 1 8 '@xf3 ti)e7, Velimirovic
Smejkal, Arandelovac 1976. The
game went 19 h4? ti)dS! =, but 1 9
..-e4 ! is very strong, e.g. 1 9 . . . lJg6
( 1 9 . . . life8? 20 'i!Vh7+ f8 2 1
't!rh8+ lJg8 2 2 .th7 ) 20 f4 f5
2 1 't!re6+ (Smejkal) is .
The Gambit Declined

3 . Probably the best is 13
...
.i.fS ! ?
anyway, since after 1 4 llb5 't!d7

White's rook on b5 may turn out
to be more of a liability than an
asset.
10
ltJxaR
.i.hS (158)
158
W
B2
.i.g4 (15 7)
!57
w
The transpositional possibilities

hereabouts can be rather bewildering, since after 6 . . . .i.b4+ Black
usually plays 7 . . . .i.g4 and
sometimes plays 7 . . . .i.b4 (+) after
6 . . . .i.g4.
White now has:
B21 7 lt:lc3 ! ?
B22 7 .i.e2
B21
.i.xf3 ! ?
7 lt:lc3!?
It used t o b e thought that this
was good for Black, but Sax has
challenged the verdict. If he were
right, then Black's best course
would be 7 ... .i.b4 8 .i.e2 .i.xf3 9
.i.xf3 'i!fc4 with transposition into
B3.
8
lt:l xdS
.i. x d 1
9 lt:l x c 7 +
d7
Is Black just gomg to swallow

the knight on a8 and gain the
material advantage of two minor
pieces against a rook?
dS
11
The Handbuch al ready gave 1 1
.i.d2 .i.b4 1 2 d5 .i.xd2+ 1 3 xd2
lt:lb4 1 4 .i.c4 lt:lf6 =F. But 14
.i.bS+! t ransposes into Sax-Vogt
below. The right reply to 1 1 .i.d2
is 1 1 . . . lt:lf6!, j ust threatening . . .
.i.d6 and then . . . n xa8.
11
ltJd4
In SaxVogt, Budapest 1 976,
Black tried 11 . . . .i.b4+ but this
failed to 12 .i.d2 .i.xd2+ 13 'it>xd2
lt:lb4 14 .i.b5+ 'it>d6 1 5 nac 1 lt:lf6
1 6 lt:lc7 lt:lxa2 (if 1 6 . . . nc8 not 1 7
lt:le8+? nxe8 1 8 .i.xe8 lt:l xe8 + . but
simply 17 a3 ! ) 17 llc4 ! . Note
that 1 1 ... ltJb4? is no good because
of 1 2 .i.b5+ c8 1 3 .i.f4 ! .
12
.i.d3
Sax's analysis continues:

a) 12 . . . .i.g6 13 .i.xg6 hg 14 .i.e3
lt:lc2+ 1 5 d2 lt:lxa1 1 6 .i.xa7! .
b) 1 2 . . . .i.b4+ 13 .i.d2 .i.xd2+ 14
xd2 lt:le7 1 5 l:l:ac l nxa8 1 6 nc4
The Gambit Declined I 37

ltJdf5 1 7 g4 ltJd6 1 8 gh i.
There is a big hole in the first line.
Instead of 1 5 . . . ltJxa 1 Black plays
15 . . . ltJxe3 followed by . . . .i.b4+
(or j ust . . . .i.d6), . . . ltJf6 and . . .
llxa8 =F.
I would not recommend 7 ltJc3
for White. Even if Sax's analysis
were sound, the minute endgame
advantage in line b after 1 8 . . .
ltJxc4+ 1 9 .i.xc4 'i!i>d6 would
hardly justify the risks involved.
B22
7
0-0-0! ?
.i.e2
Of course 7 . .i.xf3 8 .i.xf3

't!Vxd4?? loses to 9 .i.xc6+. But 7 . ..
.i.xf3 8 .i.xf3 1Wc4! is quite a good
alternative, distinguishable from
B3 only by the fact that Black's
bishop still has the option of going
to d6 or e7 rather than b4. 7 . . .
.i.b4+ 8 ltJ c 3 is transpositional: 8
. . . .i.xf3 9 .i.xf3 1Wc4 puts us
straight into B3, whilst 8 ... 0-0-0
leads to the . . . .i. b4 options in the
notes to moves 8 and 9 of the
present line.
.
ltJc3
't!fa5
This has the obvious disadvantage

of making the queen a potential
target for White's attacking efforts
(a3 and b4 or .i.d2 followed by
knight moves). But the alternatives
are not appealing:
.i.b4 9 0-0
a) 8
@'d7 1 0 .i.e3 ltJf6 1 1 'ti'a4
a1) 9
lbd5 12 ltJ xd5 'ti'xd5 and White
gets a powerful attack after either
1 3 lifc 1 .i.d6 14 b4!, Mieses.
Lowy , Vienna 1 907, or 1 3 a3 .i. d6

1 4 h 3 .i.h5 1 5 b4, Levy-Kraidman,
Lugano 1 968.
a2) 9 .i.xc3 1 0 be ltJf6 1 1 .i. e3
Niedermayer-Fridh, Vesely 1 967.
b) 8 ... 'i!Yh5 ! ? 9 h3 ltJf6 does at
least give White a chance to go
wrong:
b l ) 10 0-0? .i.d6! 1 1 hg ltJxg4 1 2
lle 1 .i.h2+ 1 3 'i!i>fl .i.e5 1 4 .i.d3
1t'h l + 15 'i!i>e2 ti'xg2 and Black's
attack won through in Stein
Levin, Kiev 1 960.
b2) 10 .i.e3 ! .i.d6 1 1 "@a4 lit he8 1 2
0-0-0 - Botterill and Harding.
ltJf6
0-0
9
Other options:
a) At this stage probably most
.i.b4 a
people wou l d think 9
poor move. But the position could
well arise by some other order of
moves, with 617/8 . . . .i.b4. The
game Estrin-Sevecek, 6th World
Corres Ch, 1 968-7 1 , shows how
the queen on a5 I bishop on b4
line-up can accelerate White's
attack: 9 . .. .i.b4 (by transposition)
1 0 .i.e3 ltJge7 ( 1 0 . . . .i.xc3 1 1 be
@'xc3 is said to be too dangerous,
but in truth Black might as well try
it. Keres gives 12 li c l 1t'a3 1 3
llxc6 be 1 4 ltJe5 , but 1 3 . . .
.i.xf3 ! ? i s not s o clear. Estrin
suggests 1 3 ltJe5 i.xe2 14 t!lxe2
ltJxe5 1 5 de 'and B lack's defence is
difficult') 1 1 ltJa4! ltJg6 ( 1 1 . . .
ltJxd4 1 2 .i.xd4 ltJc6 1 3 a3 ! .i. e 7 14
b4 't!t'h5 15 b5! - Estrin. Relatively
best is 1 1 . . . ltJf5 12 a3 i.e7 1 3 b4
'ird5 as in a game N eimanis..
Saldre, 1 970. Estrin suggests

that 14 4Jc3 '4Wd7 15 'it'a4 would
give White a dangerous attack but I wonder what the intention is
after 1 5 . . . J.xf3 1 6 J.xf3 4Jcxd4)
12 h3 .te6 1 3 a3 J.e7 14 b4 t!ld5 1 5
4Jc3 'ttd 7 1 6 'tlra4 J.xh3 1 7 d5
4Jb8 ( 1 7 . . J.xg2 1 8 de! . But
note 17 . . 4Jce5 18 4Jxe5 4Jxe5 1 9
'trxa7 'ttf5 20 '@aS+ 'i!?d7 2 1 J.b5+
c6 22 tlrxb7+ .<e8 23 J.xc6+ 'i!?f8
24 d6! - Estrin . White's last is a
necessary precaution against . . .
4Jf3+) 1 8 4Jb5 ! It.de8 1 9 gh 1 -0.
b) 9 .t cS 10 tlrb3 ! .txf3 1 1 J.xf3
J.xd4 1 2 J.xc6 be 1 3 t!rxf7
.
..
Smit.
10
J.e3 (159)
159
B
I would judge that White's

attack on the queenside is more
likely to be effective than anything
Black can do on the other wing.
But it is a sharp position and, at
least i n practice, both sides have
chances.
10
h6
O n e move among many. There

is no established view as to how
Black should handle the position:
Dec/1
The Gambit
a) 1 0 . . J.d6 intending . @hS is

the most aggressive. We have seen
this before (Stein-Levin , Kiev
1960), but Black is a vital tempo
behind in comparison with that
game: 10
J.d6 I I h3 ! 'it'h5 ( 1 1 . . .
h5 keeps the bishop a t g4 b u t does
little else) 12 hg 4Jxg4 1 3 lie l
J.h2+ 1 4 'i!?fl and if now 1 4 . . .
J.e5 1 5 J.d3 'tth 1 + 1 6 'i!?e2 'tlrxg2
17 J.f5+ intending It.g l
Botterill and Harding.
b) 10
J.e6 1 1 a3 4Jd5 12 J.d2
b6 enables White to display a
useful little trick: 1 3 d5 ! 4Jxd5 1 4
4Ja4 t: Levy-Pritchett, Glasgow
1 969 .
c ) 1 0 . . . J.cS ! ? leads t o interesting
play: 1 1 a3 .tb6 1 2 4Jb5 J.xf3 1 3
gf 4Jxd4 (otherwise 1 4 b4) 1 4
xd4 lid6?! 1 5 b 4 tlrh5 1 6 :Et c l
Penrose-Prameshuber, M unich
1958. 14 ... ltJhS! is more dangerous
(e.g. 1 5 'i!?h l 't!Pe5 ! de Villiers-Karl,
Switzerland 1 972), but White can
respond with 15 f4! .txd4 1 6
.txd4 4Jxf4 (intending . . . It.xd4,
. . . 4Jxe2+, . . . 'ttg 5+) 1 7 .tg4+! f5
1 8 J.f3 lLle6 1 9 'i!b3 .
.
. . .
...
11
12
13
a3
ltJxdS
lLldS
'it'xdS
b4
White intends to proceed with
't!Pa4, It.fc l and b4-b5.
83
6
7
8
4J c3
.tb4+
.t g4
J.e2
.txf3
8 . . . 0-0-0 would put us back
The Gambit Declihed 139
Goring Gambit: SeCtion 6

into B22.
9
i.xf3
't!t'c4! (160)
160
w
Feeble is 11 1!rb3 1!rxb 3 1 2 ab a5

1 3 0-0 0-0-0 1 4 ll:la2 ll:le7 1 5 ltlxb4
ab 1 6 lia4 ltl d 5 1 7 .i.d2 'it>d7 +
Kuzovkin-Tseitlin, USSR 1 976.
Black has the better minor piece.
It
12
13
xe2
i..e3
1!Yxe2+
0-0-0
ltle7 (161)
161
w
By preventing White from

castling Black induces the exchange
of queens. This idea was introduced
by Capablanca against Marshall
at Lake Hopatcong in 1926. The
game went 10 .i. e 3 (the dashing
American refuses to skip the
middlegame and go straight into
Capa's favourite phase) 10 . . .
i.xc3+ 1 1 b e 1!Yxc3+ 1 2 fl 1!rc4+
1 3 g 1 ll:lge7 14 ne t 't!Yxa2 1 5
lla l 'ic4 1 6 ll e t Y2 - Y2 . This still
looks to be best play after 10 i.e3.
So unless White is looking for a
quick draw (perish the thought ! )
h e should turn to the alternatives:
B31 1 0 .i.xc6+
832 1 0 'i!b3
There is no reason to think that

Black stands worse. His doubled
c-pawns are quite usful, giving a
firm grip on important squares.
There have been quite a few
boring draws from this position.
A fitting conclusion is 14 llacl
ll:lf5 1 5 l:I hd 1 lilhe8 16 cctf3 ll:lh4+
1 7 <ot>g3 ltl f5+ 1 8 <ot>f3 with
perpetual, Ghizdavu-Sydor, Skopje
1 972.
B32
B31
10
'in>3
i.xc6+
be
'i!xc6 is unthematic and
A better t ry for advantage, if

only because it is slightly provocative,
inferior (notwithstanding an ECO

assessment of ' ='), e.g. 1 1 0-0 ltle7
1 2 1!t'b3 ! i.xc 3 1 3 be 0-0 14 c4
Penrose-Barden, London 195 8.
10
1!rxb3
10
1!rxd4? would be foolish:
1 1 i.e 3 1!rd6 12 li d l 1!re7 1 3 0-0
10
10
...
11
We2+
...
ll:lf6 1 4 i.g5 Enklaar.

11
ab (162)
Declined
The Gambit
Ghizdavu-Thornally, Los Angeles

1975
162
B
13
14
15
16
llxa7
ll a8
.ixa8
.ie4
.ic5
lha8
lt:le7
0-0 (163)
1 63
w
ltl x d 4
11
Black can also refuse to be
tempted: l l . . ltlge7 1 2 0-0 a6 1 3
lla4 ( 1 3 ltld5 0-0-0 1 4 ltl xb4 ltlxb4
1 5 .if4?! lt:lec6 + Raaste- Wester
inen , Finland 1 979) 1 3 . . . .id6 1 4
.ig5 f6 1 5 .ih5+ lt:lg6 1 6 ll e I +
lt:l e 7 1 7 .id2 0-0-0 Lju bojevic
Ree, Amsterdam 1972.
ll b8
1 2 .ixb7
Black plays it safe . There is
more fun to be had with 12 . . .
ltlc2+ 1 3 <t>e2 lt:l x a 1 and now:
a) 14 .i x a8 ltlxb3 1 5 .ic6+ ct>d8 1 6
lld I + <t>c8 1 7 ltld5 .ic5
Ljubojevic-Stein, Yugoslavia-USSR
1 972.
b) 1 4 .ic6+!? is a supposed
improvement on this, when play
has continued 14 . <t>f8 ( 14 . .
ct>d8 ! ? has always seemed better to
me, although 1 5
lld l + ..t?c8 1 6
.ixa8 xb3 1 7 .ie3 ltlf6 1 8 ltld5
.
..
may
be :f) 1 5 .txa8 lt:lxb3
b 1 ) 16 .ie3 ltlf6 17 ltld5 ! .id6

lt:lxf6 g f 1 9 .id5 ltlc5 20 ll a 1
Velimirovic-Toth, Nice 1 974.
b2) 1 6 .if4 ltld4+ 17 ct>d3 e6
.ie3 a5 1 9 ltld5 .id6 20 lla 1
18
18
It seems that White has no

advantage at all here, in spite of
the bishop-pair:
a) 17 .if4 lt:le6 1 8 .ig3 llb8 1 9
.ic2 ltlc6
Ljubojevic-Parma,
Yugoslav Ch 1 972.
b) 17 0-0 llb8 18 lld l ltl e6 1 9 ltld5
ltlxd5 20 .ixd5 lld8
Levy
Harandi, Skopje 1 972.
After sixty years Capablanca's
defensive plan stands as solid as
ever.
=
c
4
d3
This is analogous to the 5 d3

line against the Scotch Gambit,
though a bit more respectable as
Black does not have a bishop on
c5 hanging around for White's
pawns to harry . There is no
gambiteering excitement in this
line as Black's reticence means
. . .

that there is to be no early contact.
The two sides just develop normally
with White retaining slightly the
better prospects because he has
more space. I suppose a few
precedents may be helpful.
5
i.xd3
d6
h3
6
Taking g4 away from Black's
pieces . A good alternative is 6
ll:l d4 ltJf6 7 f4 i.e7 8 o-o 0-0 9 1!Vc2
g6 1 0 ll:ld2 ;!; Moe-Brinck Claussen,
Danish Ch 1 969.
ll:lf6
6
A rare instance of enterprising
Black play in this variation was
Velimirovic-Antoshin, Budapest
1973: 6
i.e7 7 ltJbd2 ll:lf6 8 ltJd4
ll:le5 9 i.c2 c 5 ! ? 10 ll:le2 i.d7 1 1 f4
ltJg6 1 2 0-0 i.c6 1 3 c4 - the 'hole'
on d5 ought to give White some
advan tage, but Black has compen
sating piece activity. Another
approach is to fianchetto with 6 ...
g6, e.g. 7 i.g5 ltJf6 8 ll:lbd2 i.g7 9
ll:ld4 0-0 10 ltJxc6 be 1 1 f4 l:tb8 1 2
0-0 "t!e8 ! , Raaste-Westerinen, Hel
sinki 1 979, when I would suggest
13 "t!c2 intending 1 3 . . . ll:ld7 1 4 b4
;!;,
7
0-0
Equally good is 7 i.f4 as in
Ljubojevic-Olafsson, Las Palmas
1 974: 7 . . . i.e7 8 ll:lbd2 ll:ld7 9 ltJc4
ll:ld7 10 i.c2 0-0 1 1 0-0 ll:lde5 1 2
ll:le3 i.e6 1 3 ll:ld2 b5 1 4 i.g3 ;!;.
7
i.e7
8 ltJ d 4
0-0
lieS
9 ltJd2
10
f4 (164)
...
The Gambit Declined 1 4 1
A position characteristic o f the

whole line. White enjoys a spatial
advantage. Black has no weaknesses,
but his cramp is not the sort that .
can be described as a coiled spring
tre mbling with energy. A couple
of game sources :
a ) 10
i. f8 ( 1 0 . . . ltJ xd4 1 1 c d in
creases White's central preponder
ance) 1 1 ..c2 g6 1 2 ltl2f3 i.g7 1 3
i.d2 i.d7 1 4 li ae l ;!; Velimirovic
Keres, Sukhumi 1 966.
b) 10
ll:ld7 1 1 ll:l2f3 i.f6 1 2 i.e3
g6 1 3 "t!c2 i.g7 14 liae 1 ;!;
Velimirovic-Ivkov, Skopje 1 976.
...
...
Conclusion: 4 . . . ltJf6 is the best

way of declining the gambit if you
like lively play. 4 . . . d5 is sufficient
for equality if you adopt Capa
blanca's line (B3) - a sound, but
boringly unambitious procedure.
4 . . . d3 is colourless and ;!;, but has
at least the merit of producing
the sort of game in which the
players' own ablities and j udge
ment are mor iitlyerf:anJ than book
knowledge. <;; / , _,, _
Index of Complete Games

Anderssen-Dufresne
Barczay-Portisch
Bateman-Boisvert
Botterill-Trevelyan
Botterill-Williams
Chigorin-Gunsberg
Ciocaltea-M . Kovacs
Davis-Peters
Di.ickstein-Wittmann
Ecke-Schonewald
Emery-Menchik
Erlandsson-Demidenko
Estrin-Kondali
Estrin-Palciauskas
Estrin-Sevecek
Evans-McDonnell
Fischer-Celle
Fischer-Fine
Ghizdavu-Sydor
Hartoch-Eslon
Hawley-Mitchell
Iskov-Hebden
Iskov-Kaiszauri
J oksic-Cvetkovic
Keffler-Kocem
Krantz-Sellberg
49
21
18
131
54
84
1 09
68
1 32
53
22
67
66
78
1 37
56
88
72
1 39
72
1 07
1 33
1 33
1 18
1 15
1 29
Lazard-Gibaud
Levy-Boey
Levy-Unzicker
Lj ubojevic-Smej kal
Mariotti-Gligoric
Marshall-B urn
Marshali-Capablanca
Menvielle-Medina
Miles-Korchnoi
Nunn-H i.ibner
Nyhol m-Alekhine
Ribli-Imre
Roikov-Orlov
Santasiere-Marshall
Schwarz-Teschner
Skotorenko-Ahman
Skotorenko-Timejer
S mederevac-Lengyel
Staunton-von Janisch
S teinitz-Lasker
Sveshnikov-A. Petrosian
E. Szabo-Kocsis
Thomas-Markwell
Tim man-Tatai
Wedberg-Kaiszauri
30
131
1 24
1 25
85
46
1 39
1 16
46
79
1 35
1 16
73
57
41
54
60
1 16
97
36,37
101
1 29
31
89
51
Index of Variations - Italian Gambits

1
e4 e5
2 lbf3
lb f6
c3
4
ed
d4
5
(Part I
Greco Gambit)
cd
6
-
lbc6
(Part 2
4
lbc3
..
d5
(Moller A ttack)
9 be 35-3 7
9
9 .
i.f6
lt:le5 & others 32-35

lt:le7
10
l:te1
10 . . . 0-0 30-32
l:t xe4
d6
i.g5
1 2 g4! ? 28-30
12
13
.:.
lb xg5
13 . . h6!? 1 7-21
13 . . . 0-0 21-28
.
c3
i.a5
5 . . . i.e5 8 1-86
5
.te7 86- 9 1
6
0-0
6 't!i'b3 79-8 1
6 d4 ed (6 . . d6 74- 79 ) 7 ed:
7 . . . lt:lge7 65- 71
7 . . de 71-74
.
d6
6 . . lt:lge7, 6 . . . 'i!t'f6, 6
6 1-65
.
. .
11
12
Evans Gambit)
i.xb4
. ..
7 lt:lbd2, 7 fl 45-47
lt:l x e 4
7
i.xc3
8
0-0
8 . . lt:lxe3 38-42
9
5
i.b4+
b4
4 . . i.b6 9 1-95
6 0-0, 6 e5 42-45
6
i.c4 i.c5
i.xg5
. .
. lb f6
d4
7 . . . i.b6 52-55
7 . . . i. g4, 7 . . . i.d7 56-61
Index of Variations - Scotch Gambits

1
4 .ie4
(Part 3 Sco tch Gambit)
.ie5
.
4
. .ib4+ 102
. lt:\ [6 / 02- 1 05
e3
5
de 98- / 00
. . d3 /00- / 02
. . . lilf6 see Part 1
-
4
4
5
5
5
e4 e5 2 lLlf3 ltJc6 3 d4 ed
..
..
. . .
e3
4
(Part 4 - Goring Gambit)
de
4
lilf6 130- 134
4
d5 134- 1 40
4
d3 140- 1 4 1
4
5 li:J xe3
5 .ic4 126 :
5 . . . .ib4+ 127- 1 28
5 . . . d6! 128- 130
...
5
.i b4
5 . . . d6 120- 126
d6
.ie4
6
'i!Vf6, 6 . . . li:Jge7 ,
6 . 'i!Ve7, 6
6 . . . lt:\ [6 J l 8- 1 20
...
. . .
...
. . .
. . .
. .
0-0
7 'it'b3 1 18
.
7
.ixe3
be
8
8 . i.g4! 1 J l- J 13
8 . .ie6 1 14
lil f6 J l 4- l l 7
8
..
..
. .
...
FOR CHESS . . . READ BATS FORD
FOR CHESS . . . READ BATS FORD
T h i s book dea l s with fou r romantic gambits w h i c h were fi rst p l ayed

i n the last century and a re now enjoyi ng a resu rgence of popu larity,
parti c u l arly at correspondence p l ay.
The Greco (1 e4 e5 2 Nf3 Nc6 3 Bc4 Bc5 4 c3 Nf6 5 d4), the Scotch
(1 e4 e5 2 Nf3 Nc6 3 d4 ed 4 Bc4 ), the Eva ns (1 e4 e5 2 Nf3 N c6 3 Bc4
Bc5 4 b4) and the Cori ng (1 e4 e5 2 Nf3 N c6 3 d4 ed 4 c3) a re the
gambits featu red i n t h i s thorough l y researched work which bri ngs
the theory of these old ope n i ngs u p to the present d ay.
Gambit p l ay has an appeal for p l ayers of a l l strengths and i s an ideal
way for the average p l ayer to i m p rove h i s tactical awareness. I n
add ition the complexity of the open i ng provides the we l l-prepared
p l ayer with many opportu n ities for an early victory, often i n
s pectac u l a r style.
George Botteri l l is one of Brita i n's l ead i ng theoreticians and
co-author of Pirc Defence and Modern Defence. H e is an
I nternational Master and was B ritish Champion i n 1974 and 1977.
1 67 d i agrams
Batsford Gambit Series
T h i s exc i t i n g new series of ope n i n g works has been designed to meet the
needs of the competitive p l ayer. Each vol u me deals with a part i c u l a r
ope n i n g and the e a r l y attem pts t o obtai n sharp a n d i nteresti n g play b y a
pawn sacrifice. A l l the a uthors are top I nternatio n a l Masters and
G ra n d masters a n d the series i s u nder the general ed itors h i p of
GM Raymond Keene.
Also in this series.

Spanish Gambits
Leon i d S h a m kovich a n d Eric Sch i l ler
Other recent opening books include.

Alekhine for the Tou rnament Player
Lev A i b u rt a nd Eri c Schi l l er
Anti-Sicilian: 3 Bb5 ( + )
Yuri Razuvayev and A l exander
Matsukevitch
Grand Prix Attack: f4 against the

Sicilian
J u l ia n H odgson and Lawrence Day
Spanish without . . . a6
M i kh a i l Yudov i c h
Vienna a n d Bishop's Opening
A l exander Konsta nti nopolsky and
V l ad i m i r Lepesh k i n
Caro-Kann: Classical 4 . . . Bf5
For a complete l ist o f Bats

books please write to
B. T. Batsford Ltd,
Gary Kasparov and A lexander

Shakarov
4 Fitzhardinge Street,
london W1 H OAH.
ISB
0 7134 5085 1
7
I

Open Gambits - Botterill

Uploaded by

Copyright:

Available Formats

Open Gambits - Botterill

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Open Gambits - Botterill

Uploaded by

Copyright:

Available Formats

George Botteri II

B.T.Batsford Ltd, London

First published 1 986

ISBN 0 7134 5085 l (limp)

A BA TSFORD CHESS BOOK

A dviser: R . D. Keene G M , OBE

Symbols and References

Part 1: Greco Gambit

Section 1 : Moller Attack - Main Line

Moller Attack - 1 3 . . . 0-0

Part 2: Evans Gambit

Section 1 : 6 0-0 - Lasker's Defence

Part 3: Scotch Gambit

Section 1 : Main Line 5

Part 4: Goring Gambit

Main Line with 8 . . . .ig4 !

Index of Complete Games

W or B at the side of a diagram indicates which player is to move. A

Other major sources:

Do I come to praise these gambits or to bury them? Well, at the most

But it really is impossible to generalize about when material should be

We have to learn from results in practice, but games cannot always be

After 2 1 . . . fe 22 f4 White's major pieces would quickly kill off Black's

White was threatening not only 23 llJxb6 but also 23 e5 . So Steinitz

What a bishop on c8! For all practical purposes White is a piece up

With the 'point' that he wins the rook on d2.

16 Greco Gambit: Introduction

Here 'quietness' has been acco mpanied by avoidance of exchanges

Greco Gambit: Introduction

and a steady build-up of tension . A complicated struggle will develop

The position that is the focal point

of our investigation into the

analyses have firmly established

It is the assessment of the move

Nothi n g could be more natural

18 Greco Gambit: Section 1

ordinated and the black pawns

This is Zek's analysis, sub

The obvious drawback to this

Where should the black knight

Moller Attack Main Line 1 9

Greco Gambit: Section 1

do not combine well.

2 4 b4! also leaves Black helpless to

It is hard to be sure about such

This is one of the messy

20 Greco Gambit: Section 1

doubt White's sacrifice is inade

This is t h e move that makes 1 3

Moller A ttack Main Line

trouble (20 . . . 'ti'e7 2 1 lLlf5).

White lacks convincing alter

Greco Gambit: Section

White has little choice, since

Estrin calls this 'an unclear and

Moller Attack Main Line 21

this invalidates the whole gambit

Simply regaining the piece with

ll GJYco Gtlmblt: Stction 2

Moller Attack 13 . . . 0-0