Capablanca,J - Janowsky,D [C48] New York INT (3), 1913
This game was discussed at Exeter Chess Club training sessions on 13th/20th October 1998. The notes are based on those of Irving Chernev (Capablanca's 60 Best Chess Endings (OUP)) and liberally supplemented by comments and questions of the group, led by Tony Dempsey.
1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.Bb5
Less experienced players are likely to be more comfortable in the open positions that arise after 4. d4, the Scotch Four Knights Game.
4...a6
The challenge to the Bishop, routine on move 3 of a Ruy Lopez, is here rather dubious. White has more than one way to maintain an advantage. Preference should be given to one of the alternatives:
[4...Bb4 is the reliable main line 5.0-0 0-0 6.d3 d6 7.Bg5 threatening Nd5 7...Bxc3 Defending against the threat; as ever, Black must be wary of maintaining symmetry for long. 8.bxc3 Black now unpins (the Metger unpin) 8...Qe7 9.Re1 Nd8 10.d4 Ne6 11.Bc1 is preferable and recommended by theory (11.Bh4?! is dubious due to the sensitivity of the f4 square. 11...Nf4 12.Nd2 Kh8 13.Bf1 Kamsky-Timman, 1991) 11...c5 12.Bf1
Most of you may be scratching your heads wondering what White is up to. According to theory the position is fairly balanced with both sides having opportunities for active play. But how confident would you be about handling White's positioon in practice? Do you close the centre with d4-d5, exchange with dxe5 or maintain the tension? How do you activate your Rooks and Bishops? Are those queen's-side Pawns going to be liabilities?;
4...Bc5 unusual 5.0-0 0-0 6.Nxe5 Nxe5 7.d4 Bd6 8.f4 Nc6 9.e5 with unclear complications 9...a6 10.Be2 Bb4 11.d5 Bc5+ 12.Kh1 Nxd5 13.Nxd5 d6 14.Bd3 dxe5 15.fxe5 Nxe5 16.Bxh7+ Kxh7 17.Qh5+ Kg8 18.Qxe5 Bd6 19.Qh5 f6 20.Bf4 Be6 21.Rad1 Bf7 22.Qf3 Bxf4 23.Nxf4 Qc8 24.Nd5 Bxd5 25.Qxd5+ Rf7 26.Rd3 c6 27.Qh5 Re7 28.Rh3 Qf8 29.Qh8+ Kf7 30.Rg3 Ke8 31.Rd1 g5 32.Qh5+ Qf7 33.Qh8+ Qf8 34.Qh5+ Qf7 35.Qh8+ Qf8 36.Rh3 Rg7 37.Re3+ 1-0 Short Nigel D-Adams Michael/Ch England (play-off), England 1992;
4...Nd4 Rubinstein's dynamic move]
5.Bxc6 dxc6 6.0-0
Capablanca charateristically opts for a quieter, more positional, approach to the position, one in which is however sufficient to retain some advantage.
[After 6.Nxe5! Black maintains material equality with 6...Nxe4 7.Nxe4 Qd4 8.0-0 Qxe5 but after 9.Re1 Be6 10.d4
Black's uncastled King and lack of development portend a winter of discontent
A) 10...Qf5 has been tried several times 11.Bg5 is natural, continuing development while preventing castling 11...Bd6 (11...h6!? 12.Qd3! and White's initiative persists) 12.Qd2 h6 (12...0-0? 13.Nxd6 cxd6 14.Be7) 13.Bh4 g5 14.Nxd6+! cxd6 15.Bg3 White enjoyed a clear plus in Yudasin-Sagalchik, Kemerov 1995;
B) 10...Qd5 was tried, but after 11.Ng5 (11.Bg5 plays for an attack on the King 11...Kd7 (idea ...Re8) 12.Rc1!? intending c2-c4 (Nunn)) 11...0-0-0 12.Nxe6 fxe6 13.Qg4 Qxd4 14.Qxe6+ Qd7 15.Qxd7+ Rxd7 16.Re8+ Rd8 17.Rxd8+ Kxd8 18.b3 Bc5 19.Bb2 Rf8 20.Rd1+ Kc8 21.Rd2 the resultant endgame, typical of the Ruy Lopez Exchange Variation, was better for White, there being no compensation for the doubled c-Pawns.; 6.d3 is possible refinement of White's plan, keeping open the option of castling long.]
6...Bg4 7.h3 Bh5
[7...Bxf3 submissive relinquishment of the Bishop pair hardly comes into the question, certainly for a two-Bishop man like Janowsky. 8.Qxf3 White would calmly prepare the pawn break f2-f4 as in the game.]
8.Qe2
[8.g4 "Winning the e-Pawn" with 8...Nxg4 (8...Bg6 9.Nxe5 does merit consideration, as the extra Pawn appears to outweigh the loosening of the castled King's defences, but Janowsky would undoubtedly have sacrificed a piece for two Pawns) 9.hxg4 Bxg4 Janowsky would be in his element pursuing the initiative against the exposed White King.]
8...Bd6 9.d3 Qe7
How should White proceed?
10.Nd1!
Once familiar with the Metger unpin manoeuvre, the purpose of this slightly surprising move will be quite clear. As the position is going to remain closed for some time yet, Capablanca needn't be in a hurry to develop the queen's-side, particularly when it is not yet clear where they belong.
[10.Bg5 attracted some support on the night, but, as usual, it is inadvisable to pin the opponent's King's Knight when we have castled but our opponent has not. In the absence of a threat to capture, the Bishop sortie would merely provoke ...h6 and ...g5-g4, with an initiative for Black against the White King.;
10.Be3 was also suggested. But what is the plan? A future d3-d4 might prove double-edged, opening the position for the Black Bishops and Rooks. Black might restrain the advance with ...c6-c5: although the d6 Bishop looks none too attractive, White has still to solve the problem of the pin on the Nf3.]
10...0-0-0
Janowsky optimistically seeks to hold on to the Bishop pair, but to no avail.
[10...Nd7 was proposed by Lasker to seek counterplay 11.Ne3 Bxf3 12.Qxf3 g6 with the idea of ...f7-f5]
11.Ne3 Bg6 12.Nh4 Rhg8 anticipating the attack on g7 by a Knight on f5
13.Nef5 Qe6
14.f4!
The tactical threat is 15. Nxd6+ Qxd6 16. f5 Bh5 17. g4 and Black's once-proud Bishop is caught in a Noah's Ark Trap. The strategical idea is to open the f-file, when White will have no doubt where the Rooks are headed.
14...Bxf5 15.Nxf5 exf4 16.Bxf4 Bc5+ 17.Be3 Bf8
The Bishop is passively placed but plays an important role in shoring up the King's-side. Capa elects to exchange it off.
18.Qf2 Rd7 19.Bc5 Bxc5 20.Qxc5
threat Qc5-a7-a8+
20...Kb8
The Pawn position is in White's favour, and he plans to turn it to account like this:
(a) to double Rooks on the f-file to induce the inevitable advance ...f7-f6;
(b) to exchange Queens and Knights,
(c) to advance the g-Pawn to g5, eliminate the f-Pawn and create a passed White e-Pawn. [If Black recaptures with the Pawn on f6, he will then acquire a permanent weakness; this minority attack by White has a subtle sting]
(d) then - well, the rest is a matter of endgame technique, which, even then, Capablanca had at his finger-tips.
21.Rf2
idea Ra1-f1 and Nxg7!, winning a Pawn
21...Ne8 22.Raf1 f6
sooner or later this move had to be made
[22...Qxa2? 23.Ne7 Rh8 24.Rxf7 lets the enemy into camp]
23.b3 Nd6 24.Rf4
Black now faces a critical decision - to stay in a middlegame or exchange into an ending. It's not unusual for errors to be made at this key moment...
24...Nxf5?
Janowsky undoubtedly recognised that he stood worse, but evidently felt he would be less badly off in the double-Rook ending, underestimating White's chances. Moral: studying the ending improves your handling of the middlegame. Recommended reading: MEDNIS, From the Middlegame to the Endgame.
25.Qxf5 Qxf5 26.Rxf5 Re8
White has several advantages (active Rf5, pressure on f6, potentially more active King which will run to e3 supporting the central Pawns).Black's Rooks are hampered by being defensive. [The activity of the Rooks is key in these endgames.]
The hobbled c-Pawns cannot be easily dissolved since ...b6 is met by b3-b4.Capa plans to create a passed e-pawn, which will be his winning trump. He will move the King to e3, advance the g-Pawn, and then advance the central Pawns. He will do this while restraining any attempt by Black to become active on the queen's-side.
27.g4!
The first step in the process. Note that with Queens still on the board, the minority attack would involve the liquidation of the White King's Pawn cover. This would probably mean that another plan would have to be found for White, unless the King could first be separated over to the Queen's-side. This same strategy - the minority attack - is also seen in the Queen's Gambit Exchange Variation. There the White b-Pawn advances to strike a target at c6. Black's preferred response would be a minority attack with ...f7-f5-f4, though, due to the positioning of the minor pieces, this is difficult to achieve.
Recommended reading:
Black holes on the Chess Board - Cougar Irons
The Tao of Chess - Frijtof Fischer
The Tao of Potboilers - R D Keene.
27...b6
hoping for something like ...c5, ...Kb7-c6, ...b5 and ...c4, liquidating the doubled Pawns and empowering the Rooks.
28.b4! Stopping any undoubling of the Pawns. Capablanca must expose his Pawns trying to restrain Black, and could make this move only after carefully calculating that this looseness could not be exploited. 28...Kb7 29.Kf2 b5 Planning now to open the a-file. 30.a4! Rd4
[30...bxa4 31.Ra1]
31.Rb1 What's this? The White Rook forced to be passive? Back at move 28 this is what White had to calculate: that, temporarily, White drops the Rook back, so the Black Rook can next be expelled, and the White Rook returned to active duty.
31...Re5 32.Ke3 Rd7
[32...Rxf5 33.gxf5 Rd7 34.d4 and White is winning]
33.a5!
fixes Black's Pawns on the Queen's-side, and lets White concentrate on the King's-side.
33...Re6
34.Rbf1
"I'm back!". Recommended reading for Black: The Book of Common Prayer.
34...Rde7 35.g5 fxg5
[35...Rf7 36.d4 with Rf1-f4 and e4-e5]
36.Rxg5
Capablanca says at this point that his chances of winning are excellent. He cites these advantages:
(a) he has a passed Pawn;
(b) his King position is ideal, eyeing d4, c5 and the King's-side
(c) his Rook commands the only open file.
36...Rh6 37.Rg3 Rhe6
restraining d3-d4
38.h4 g6 39.Rg5
planning h4-h5, fixing some sort of isolated weakness
39...h6
[39...Re5 40.Rf8 Re8 41.Rxe5 Rxe5 42.Rf4 is also easy sailing for White]
40.Rg4 Rg7
(else Rf1-f8-g8)
41.d4 Kc8 42.Rf8+ Kb7
[42...Kd7 43.Ra8]
43.e5 g5 44.Ke4 Ree7 45.hxg5 hxg5
[45...Rxg5 46.Rxg5 hxg5 47.Rg8 loses a Pawn without affording any relief]
46.Rf5 Kc8 47.Rgxg5 Rh7 48.Rh5 Kd7 49.Rxh7 Rxh7 50.Rf8 Rh4+ 51.Kd3 Rh3+ 52.Kd2 c5
[52...Rh4 53.Kc3 Rh3+ 54.Kb2 Rh4 55.c3 Rh2+ 56.Kb3]
53.bxc5 Ra3 54.d5 1-0
[54.d5 Rxa5 55.Rf7+ Kd8 56.c6 b4 57.Rf8+ Ke7 58.d6+! cxd6 (58...Kxf8 59.dxc7) 59.c7 winning]
"Capablanca's play was beyond reproach, and he executed his plan in a very able manner." said Lasker. It is also important to note that Black's unsuccessful attempt to generate play on the queen's-side in the end left an embarrassing set of weaknesses - c5,c6,a6 - each of which played its own role in his demise. Could Janowsky have improved his defence around moves 27-29?
1-0