When Alekhine, with apparent nonchalance, played, for instance, h4 against the French Defence, this appeared to his contemporaries to be ‘the pretentiousness of genius’, but a closer analysis today reveals that behind such moves there lay a great deal of mental calculation and hard logic, and that this move h4 belongs exactly when it was played; a move earlier it would have been weak and on the next it would be too late!
Further illustration of all these points will be better obtained by the reader from the actual games which follow.
Alekhine - Asztalos
Kecskemet, 1927
Queen’s Gambit Declined
1 Nf3 Nf6 2 c4 e6 3 d4 d5 4 Bg5 h6
A little experiment in the opening – and a dangerous one when Alekhine is on the other side of the board. The spatial advantage now obtained by White is worth more than Black’s two bishops.
5 Bxf6 Qxf6 6 Nc3 c6 7 Qb3 Nd7
Having started experimenting, Black ought to have gone on in like manner, e.g. 7 ... Qf5 8 e3 Nd7 9 Rd1 Qg6, by which he would have put more difficulties in the way of White’s plan to gain space in the centre.
8 e4 dxe4 9 Nxe4 Qf4 10 Bd3 Be7 11 0-0 0-0 12 Rfe1 Rd8 13 Rad1 Qc7
Black cannot complete his development as long as his queen remains on f4, for example 13 ... Nf6 14 Ne5! Rxd4 15 Nxf6+ Bxf6 16 g3 Qg5 17 Nf3 Qc5 18 Nxd4 Bxd4 19 Bh7+ followed by Qd3+ and Qd4.
14 Ng3 Nf8 15 Qc3
A useful move, which is intended, among other things, as a preparation for an eventual Nh5, thereby provoking, by virtue of the threat of d5, a weakness in Black’s king position. Black parries this with ... a5 and further moves on the queenside. We shall see at what cost!
15 ... a5 16 a3 a4 17 Ne5
Now 17 Nh5 can be met by 17 ... Qa5. However, as a result of Black’s advance of the a-pawn, his pawn structure on the queenside has lost its flexibility, which in its turn makes it easier for White to favourably resolve the tension in the centre.
17 ... Qa5 18 Qc1 Bd7 19 c5!
This threatens Nc4-b6 and thus induces Black to play b5, which leads to a blocked position on the queenside and restricts his counterplay in the event of White’s launching an attack on the kingside.
19 ... b5 20 Be4
A series of preconditions for an attack on the king has been created: the weakness represented by the black pawn at h6; the aggressive placing of White’s pieces; his control of the centre; and Black’s limited resources on the queenside. White, however, is not in a hurry to attack, for he sees that he can use his pressure on c6 to improve the deployment of his own pieces and, at the same time, to constrict his opponent’s.
20 ... Qc7 21 Qc3 Be8 22 Ne2!
White is thinking of the square b4.
22 ... Ra6 23 Nc1 Nd7
Against the threatening line-up of white knight at b4 and queen at f3 Black has only two expedients, namely, the direct weakening of his king position by 23 ... f6 or an indirect weakening in the form of the exchange of his knight, which defends the square h7. Black decided on the latter course.
24 Nxd7 Rxd7 25 Nd3 Rd8 26 Ne5 Bf8
Black is sure that the enduring pressure on the diagonal against h7 will oblige him to play ... g6. Logically, therefore, he decides upon a delayed fianchetto as the best means of putting up some resistance.
27 h4!
The first move which really commits the attacker; the situation is also ready for it. It is directed against the future weakness on g6.
27 ... Raa8 28 Bb1 h5
He does not like the idea of 29 Qc2 g6 30 h5.
29 Qf3 g6 30 g4 hxg4 31 Qxg4 Bg7 32 Ba2!
Before proceeding any further, White prevents ... f5, which Black cannot prepare by 32 ... Qe7 on account of 33 Nxg6 and 34 Rxe6.
32 ... b4 33 Bc4!
White prefers to avoid 33 axb4 a3 which would allow Black to obtain counterplay.
33 ... bxa3 34 bxa3 Qa5 35 Qe4 Qc7 36 Qf4 Rab8 37 h5
This must come sooner or later.
37 ... gxh5 38 Kh1 Rb7 39 Rg1 Qe7 40 Rxg7+!
White cannot count on success without this sacrifice to eliminate the fianchettoed bishop, Black’s main line of defence.
40 ... Kxg7 41 Rg1+ Kh7 42 Nxf7!? 1-0
Having played so excellently and – as regards the method of preparing the attack on the king – so instructively. Alekhine concludes the game with a sacrifice that is not fully correct! I say ‘concludes’, because Asztalos took fright and resigned when he could well have played on. The continuation which Alekhine had in mind and which he expounded in the tournament book contained a loophole giving a draw, and even against White’s best play Black would have had drawing chances.
Let us first look at Alekhine’s line:
42 ... Qxf7 43 Bd3+ Qg6 44 Bxg6+
If 44 Qg3, Black replies 44 ... Rg7, while 44 Qf3 can be met by 44 ... Rd5. Thirdly, 44 Qg5 Qxd3 45 Qg8+ Kh6 46 Qh8+ Qh7 47 Qf8+ Qg7! 48 Rxg7 Rxg7 49 Qf6+ Bg6 50 Qxd8 Be4+ probably results in a draw.
44 ... Bxg6
45 Rxg6?
So far this is Alekhine’s analysis. However, 45 Qf6! Rg8 46 Qxe6 is stronger and does indeed win, though White still has a hard task ahead.
JN: A distinct exaggeration. White has queen and pawn against rook and bishop. Black’s remaining pawns are all weak and White has a passed f-pawn – I wouldn’t expect many technical problems here!
45 ... Kxg6 46 Qe4+ Kg7 47 Qe5+
Here Alekhine concludes his analysis with the remark that Black ‘after a few further checks, would inevitably lose one of his rooks.’ In fact, he loses after all moves except 47 ... Kg6 followed by keeping the king on g6 or g7. After 48 Qxe6+ Kg7 49 Qxc6 Rb1+! 50 Kg2 Rxd4 51 Qc7+ Kg6 the checks cease, whereupon Black’s rooks begin to deploy themselves for play against the king or the c-pawn. One cannot see how White can win; the position is very likely drawn. Alekhine unintentionally bluffed the unfortunate Asztalos and the jury, who on the basis of his analysis awarded him the Brilliancy Prize.
We must now point out what is objectively the strongest move and the continuation by which White wins. Thus (instead of 42 Nxf7!?) he should have proceeded as follows:
42 Qg3 Qf8
Forced; if 42 ... f5, then 43 Qg8+ followed by Qh8+ and Qf6+ is decisive; while if 42 ... Qf6, then 43 Bd3+ wins.
43 Qh4! f6
Both 43 ... Rd5 44 Ng4 and 43 ... Rdb8 44 Qf6 Rb3 45 Bxe6 also win for White.
44 Bd3+ Kh8 45 Bg6! Rh7
Or 45 ... fxe5 46 Qxh5+ Kg7 47 Bxe8+ and 48 Qg5#.
46 Bxh7 Kxh7
After 46 ... fxe5 47 Qxd8 Qf3+ 48 Rg2 Black has one more check before White mates him.
47 Ng4 Bg6
Or 47 ... Kh8 48 Nxf6 Bf7 49 Nxh5 and White wins.
48 Nxf6+ Kg7 49 Nxh5+ and Black loses his bishop with check.
Capablanca - Schroeder
New York, 1916
Queen’s Gambit Declined, Orthodox Defence
1 d4 d5 2 Nf3 e6 3 c4 Nf6 4 Nc3 Nbd7 5 Bg5 Be7 6 e3 0-0 7 Rc1 a6 8 Qc2
This is better than either 8 c5 or 8 a3, which Capablanca played in a number of games of his 1928 match against Alekhine, but best of all is 8 cxd5.
8 ... Re8 9 Bd3 dxc4 10 Bxc4 b5 11 Bd3 Bb7?
At this point Black cannot spare the time to continue preparing 11 ... c5; he should have played it at once.
12 a4!
In order to entice the black b-pawn to b4, which is useful for further play against the square c5.
12 ... b4 13 Bxf6
This exchange is advantageous in two respects; first, for its ‘clearance’ of the square c5 and secondly, because it eliminates the possibility of a defence later on by ... Nf8.
13 ... Nxf6 14 Ne4 Nxe4 15 Bxe4 Bxe4 16 Qxe4 c5
Sink or swim, Black must play this, for otherwise 17 Qc6 follows, fixing the backward c-pawn.
17 dxc5 Qa5 18 b3 Bxc5 19 Ng5!
Can it be that White is justified in switching to an attack on the king after preparations which were clearly directed at the c-file? How does this fit in with our thesis on the need to fulfil a series of precond
itions before opening an attack on the king? In fact, there is no contradiction; the preconditions are present, but they have arisen in a different way from those in the Alekhine-Asztalos game.
It is true that the preconditions for the attack on the castled king in this game have come about suddenly and somewhat incidentally; at the same time, this has occurred quite legitimately. The main point is that White’s play, starting with 12 a4, serves two purposes: it maintains the pressure on the c-file and simultaneously improves the preconditions for an attack on the king. As a result of 12 a4 b4 and 18 b3 the position on the queenside became blocked, and this has reduced Black’s prospects of counterplay, while 17 dxc5 resolved the tension in the centre.
Furthermore, 13 Bxf6, 15 Bxe4, and 16 Qxe4 represent, in effect, a far-reaching elimination of Black’s defending pieces; in the same way, the struggle revolving around the pawns on the c-file has forced Black (in 17 ... Qa5 and 18 ... Bxc5) to displace two pieces which are important for the defence of his king position. This is without doubt a full set of preconditions, built up with true Capablanca virtuosity and élan. Now White’s knight, a sort of ‘equus ex machina’, stirs up a whirlwind attack on the king, against which there is hardly an adequate defence.
19 ... h6?
Although Black has no really sound reply at his disposal, 19 ... f5 was the lesser evil. Then White does not gain much by 20 Qe5 Bb6, while if 20 Qb7, Black plays 20 ... Qb6, and after 21 Qf7+ Kh8 22 Qh5 h6 White’s attack stops; for example, 23 h4? Bxe3! 24 Qg6 is not sound, since after 24 ... hxg5 25 hxg5+ Kg8 White cannot give perpetual check, while 26 Rh7 is weak on account of 26 ... Qd4. The best line is 20 Qc4 Rac8! 21 Nxe6 Bb6 22 Qxc8 Rxc8 23 Rxc8+ Kf7 24 Ng5+ Kg6 25 Nf3 Qd5 26 0-0 Bd8 followed by ... Bf6 with some chances of a draw. This continuation would have been relatively the best way out for Black.
JN: After 27 Rc6+ Bf6 28 Nd4 I see no real chances for a draw; White has two rooks and a pawn for the queen, and Black’s king is exposed. One line is 28 ... Kh5 29 Rc4 Bxd4 30 Rd1, followed by Rdxd4.
Better than the text move, but worse perhaps than 19 ... f5, is 19 ... g6. An analysis of this move, which, though it may emanate from Capablanca. needs to be corrected, runs as follows: 20 Qf3 Rf8 (everything else is weaker; 20 ... Ra7 in particular, does not work on account of 21 Qc6) 21 Qh3 h5 22 Nxe6? fxe6 23 Qxe6+ Kh7 24 Qd7+ Kh8 25 Qc6; now the analysis continues with 25 ... Rac8?, missing the fact that Black can win by 25 ... Bxe3! 26 fxe3 Qf5 27 Qf3 (forced) 27 ... Qd3 28 Qe2 Qxb3, obtaining an attack and, before long, a material advantage as well.
JN: This allows a defence by 27 Qc2, but Black can improve by 26 ... Qg5!, which is indeed a win for Black.
This analysis can be improved by playing, instead of 21 Qh3, 21 Qf6 (threatening either 22 Rc4 and 23 Nxh7 or at once 22 Nxh7) 21 ... Rae8 (if 21 ... h6 then 22 Nxe6 wins, while 21 ... Rac8 is not good because of 22 Rc4) 22 Qe5 Bb6 23 Qxa5 Bxa5 24 Ke2 Rc8 25 Ne4, when White has much the better ending, for example if 25 ... Rfd8, then 26 Nc5 Rc6 27 Nb7 Rxc1 28 Rxc1 Rd5 29 e4.
JN: Following 21 ... Rac8 22 Rc4 Black can defend by 22 ... Rc7!, for example 23 Nxh7 Be7 or 23 Rh4 Be7. Thus it is hard to demonstrate an advantage for White after 19 ... g6.
20 Qh7+ Kf8 21 Qh8+ Ke7 22 Qxg7 hxg5 23 Qxg5+ Kd6 24 Ke2
Two pawns, the exposed position of the black king, Black’s commitments at c5, and White’s rooks on the c- and d-files – all this is obviously more than sufficient compensation for the sacrificed knight. Capablanca also particularly stresses Black’s inability to stop the advance of the h-pawn while White is tying him down on the queenside.
24 ... Rac8 25 Rc4 Kc6 26 Rhc1 Kb6 27 h4 f5
Capablanca had reckoned on the following transition into a won ending: 27 ... Rc7 28 h5 Rec8 29 h6 Bd6 30 Qxa5+ Kxa5 31 Rxc7 Rxc7 32 Rxc7 Bxc7 33 f4 Bd8 34 g4 Bf6 35 g5 Bh8 36 e4 Kb6 37 f5 followed by g6-g7.
28 Qg7 Re7 29 Qe5 Rc6
This loses at once, but 29 ... Rec7 has no future either in view of 30 h5.
30 Rxc5 1-0
For if 30 ... Rxc5 then 31 Qd6+ is decisive.
Capablanca’s virtuosity in integrating all the features of the position into a coherent whole is worthy of close attention.
Alekhine - Kmoch
San Remo, 1930
Nimzo-Indian Defence
1 d4 Nf6 2 c4 e6 3 Nc3 Bb4 4 Bd2
One of the most harmless continuations.
4 ... 0-0 5 e3 d5 6 Nf3 c5 7 a3 Bxc3 8 Bxc3 Ne4 9 Rc1 Nxc3 10 Rxc3 cxd4 11 exd4 Nc6 12 Be2
If 12 c5 then Black can immediately play 12 ... e5.
12 ... dxc4 13 Bxc4 Qf6
White has not achieved very much in the opening. In view of the reduction in material his spatial advantage offers no more than approximate compensation for the isolated pawn.
14 0-0 Rd8 15 Rd3 Bd7 16 Re1
By 16 Qd2 White could have prepared to obtain clear equality by d5. However, he does not like the idea of a draw.
16 ... Be8 17 Qd2 Ne7 18 Ng5 Nd5
As this threatens 19 Nxe6 fxe6 20 Rxe6, Black does not have time for 18 ... Bc6 or 18 ... Nf5.
19 Rf3 Qe7 20 Rg3 h6
White has proceeded cautiously and prepared the ground for Qd3, with the aim of provoking a weakening of the enemy king position in the shape of ... g6. For this reason Kmoch prefers to go in for ... h6, which weakens him less. White can only obtain results by attacking the king, for Black is the better placed for the endgame.
21 Nf3
If 21 Ne4 Black can reply 21 ... Qh4.
21 ... Qf6 22 Re4 Ne7
Parrying the threat of Reg4.
23 Ne5 Nf5 24 Rd3
White has managed to weaken the enemy king position, but he still lacks the preconditions for a real attack. A little regrouping is therefore needed, and above all the square d4 must be safeguarded. If 24 Rf3 then 24 ... Bc6 25 Nxc6 bxc6 is quite good for Black. Of the minor pieces the knights are more valuable than the bishops, from the point of view of both the attacker and his opponent.
24 ... Rac8 25 h3 Nd6?
Up to this point the author of The Art of Defence has defended correctly and preserved equality. Now, however, he goes astray in playing to exchange off his important knight, which is the main defender of his king position. Correct was 25 ... Bc6.
26 Rf4 Nxc4 27 Nxc4 Qg5 28 Rg3 Qd5 29 Ne3 Qc6 30 Kh2 Qc1 31 Qb4 Qc7 32 d5!
Apparently, Black has manoeuvred himself into a new defensive position, one which also affords the possibility of his inconveniencing White’s queen further by ... a5. The only weak side of his manoeuvre is that it does not give sufficient attention to centralization. As a result, the alleged pawn weakling on d4 is transformed – ironically – into an aggressor, punishing Black for his neglect of the central squares.
32 ... a5
After 32 ... exd5 33 Qd4 g6 (or 33 ... g5) 34 Nf5 Black would quickly come to grief; equally bad is 32 ... e5 33 Rfg4 g6 34 Nf5.
33 Qe4 Rd6 34 Qe5 g6 35 Qh5!
If 35 Rc4, Black could bring the attack to a halt by 35 ... Rc6!.
35 ... Rxd5
Or 35 ... Kh7 36 Ng4! gxh5 37 Nf6+ and mates.
36 Nxd5 exd5 37 Qxh6 Qe5 38 Rh4 Qg7 1-0
Colle - Capablanca
Carlsbad, 1929
Colle System
1 d4 Nf6 2 Nf3 b6 3 e3 Bb7 4 Nbd2 e6 5 Bd3 c5 6 0-0 Nc6 7 c3
The formation with pawns on c3, d4, and e3, bishop on d3, and knights on d2 and f3 characterizes the Colle System. White’s plan is to force through e3-e4 and then exchange his d2-knight for Black’s f6-knight on this same square; he thereby creates some of the preconditions for an attack on the king without entailing any commitments. The Colle System, however, lacks positional depth and, consequently, Black does not have any great anxiety as regards the centre.
7 ... Be7 8 e4 cxd4
Capablanca’s treatment of the opening, adopting a Queen’s Indian set-up, is both sound and also suitable for avoiding lines which equalize out to a draw. White’s strong point in the centre is now eliminated, for he must recapture the pawn with his knight, otherwise 9 cxd4 Nb4 1
0 Bb1 Ba6 leaves Black with the better game.
9 Nxd4 0-0 10 Qe2
First 10 Nxc6 would have been sounder.
10 ... Ne5 11 Bc2 Qc8 12 f4
Preparation for an attack on the king, but also a weakening of the g1-a7 diagonal.
12 ... Ba6 13 Qd1 Nc6
Obviously not 13 ... Bxf1 on account of 14 fxe5. Moving the bishop to a6 had other points: on the one hand, to drive the white queen back and, on the other, to induce White to play Rf3, whereby he undertakes a commitment to attack before the preconditions for a successful attack have matured.
14 Rf3
Against 14 e5 Black has 14 ... Nd5, after which 15 Bxh7+ does not work (because of 15 ... Kxh7 16 Qh5+ Kg8 17 Rf3 f5 18 Rh3 Nxf4) while, given a free move, Black can consolidate his defences by playing 15 ... f5.
In the event of other, slower methods of attack Black has a sound defensive plan based on the effective use of the diagonals for his bishops and a centralized knight on d5. This will be seen in the further course of the game.
14 ... g6!
In view of the restricted range of activity of White’s dark-squared bishop, this weakness is negligible. It is important for Black to exclude short-term dangers on the kingside, since he needs time to increase his pressure on the central squares.
15 N2b3 Nxd4 16 Nxd4 Bb7 17 Qe2 Bc5 18 Rh3 Qc6!
Black forces e5 in order to gain the d5-square for his knight and open up the diagonal.
19 e5 Nd5
20 Qf2?!
Up to this point Colle’s play has been a little pretentious but not really incorrect, for his commitments have not been decisive ones. Still, he should now have admitted that it was pointless to continue playing for an attack and striven, by paying careful attention to the central squares, to keep the game within the well-known ‘margin of a draw’; for example, 20 Be3 Bxd4 21 cxd4 Rac8 22 Bb3!, when Black, after 22 ... Nxe3 23 Rxe3, clearly has nothing either in the centre or on the c-file; nor does 22 ... b5 or the like offer anything very much. (In Golombek’s book, Capablanca’s Hundred Best Games of Chess, instead of 22 Bb3!, the erroneous 22 Rc1? is given; then Black does indeed get the upper hand by 22 ... Nxe3 23 Rxe3 Qc4.)
Art of Attack in Chess Page 30