Saturday, April 21, 2007

Nimzo-Indian Defense 4.Nf3

In between rounds at the 1988 U.S. Amateur Team East tournament, national master Peter Radomskyj and I analyzed this fascinating game of his. Although it had been played at a team championship 14 years earlier, the game was so fresh with tactical ideas that it might just as well have been played in a preceding round of the 1988 event. The beauty of a well-played chess game is its timelessness. [West]

This may be the best attacking game I have ever played against a master, which is why I am offering it here. It features a theme dear to the hearts of grandmaster and novice alike: the attack against the uncastled king. In fact, I am sure that a game like this is relatively rare in modern chess. Can you find a flaw in White's attack? At any rate, this sacrifice was very tempting over the board. And they say it is easier to attack than to defend. [Radomskyj]

Peter Radomskyj [Rutgers] - Bernard Hill [Queens Chess Club], U.S. Amateur Team Championship 1974

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Nf3


This move can sometimes lead to trappy play, but often it serves just to transpose into main lines. In 1974, the fianchetto of the white king bishop in conjunction with 4.Nf3 (as practiced today by Kasparov) was not played frequently. [Radomskyj]

4...c5 5.d5

This attempt to cramp Black may well be premature. Safer is transposing into the Rubenstein variation with 5.e3. After the game move, Black's most positionally sound continuation is probably 5...Bxc3+ 6.bxc3 d6. Instead he tries to refute White's move outright. [Radomskyj]

Batsford Chess Openings gives 5...Bxc3+ 6.bxc3 d6 7.e3 e5 8.Qc2 Qe7 9.Be2 O-O 10.Nd2 e4 11.h3 and 12.g4. [West]

5...b5


This has been played before: 5...b5 6.dxe6 fxe6 7.cxb5 d5 8.e3 Ne4 with an unclear position in Antoshin-Garcia, Sochi 1964. Another plan for Black is 5...d6 6.Bg5 h6 7.Bxf6 Qxf6 8.Rc1 O-O 9.e3 e5 10.Be2 e4 11.Nd2 Qg6 12.Kf1 Bxc3 13.Rxc3 Nd7!, again unclear in Larsen-Kaplan, San Juan 1969. Szabo recommends 5...Ne4 6.Qc2 Qf6 7.a3!? Bxc3+ 8.bxc3 exd5 9.Bb2 d6 10.cxd5 Bf5 11.c4 Qd8!. [West]

6.Bg5

Taking the pawn is probably too dangerous for White. So he tries to transpose into a kind of Spassky variation. Note that 5...b5 did not develop a piece. [Radomskyj]

6...bxc4

Black can probably equalize with 6...exd5 7.cxd5 h6, leading to a type of Benoni formation. [Radomskyj]

7.e4 d6 8.Bxc4 e5 9.O-O h6?


Black had to play 9...Bxc3 before making this move. [Radomskyj]

10.Bxf6 Qxf6 11.Nb5!

This knight will haunt Black for the rest of the game, if only in the notes! [Radomskyj]

11...Qd8 12.Qa4 Ke7


If 12...O-O, then 13.a3 Ba5 14.Nxd6 wins a pawn, e.g., 14...Bd7 15.Bb5. Looking at the board, I saw Black's uncastled king and undeveloped pieces. Just the blocked center stood in the way. And the only way to unblock it and open the game would be via a sacrifice. [Radomskyj]

13.Nxe5!


In all honesty, after analyzing this position, the only possible attempt that I see for refuting White's attack is 13...a6. If 14.Nd3, then 14...Bd7!, and White suffers from a pin that threatens to win two pieces for a rook. Then the question would be whether White can take advantage of the exposed position of Black's king, for instance with e5 at some point. Still, this seems rather nebulous. At least, I see no forced win for White, unlike the game continuation. Play might continue 13...a6 14.Nd3 Bd7! 15.Nxb4 cxb4 16.e5 axb5! 17.Qxa8 bxc4. Sure, White can threaten to expose the black king by 18.Rfe1 and can possibly win the hanging pawns on the queenside, giving him passed pawns in return. But this will take a little time, and Black may just be able to use that respite to consolidate. It looks pretty unclear to me. I remember that I was afraid of 13...a6 at the board. But since I was playing for a win, I took the chance. [Radomskyj]

In answer to 13...a6, White has the strong move 14.f4!. For example:

a) 14...dxe5 15.fxe5 axb5 16.d6+ Kf8 17.Rxf7+ Ke8 18.Qxb5+ Bd7 19.Qb7 Ra5 20.e6 +-;

b) 14...axb5 15.Qxa8 bxc4 16.Qxb8 dxe5 17.fxe5 c3 (17...Bg4 18.Qb5 Be2? 19.Qb7+ Qd7 20.Rxf7+ +-) 18.bxc3 Bxc3 19.Qa7+ Qd7 20.Qxc5+ and 21.Qxc3 +-. [West]

13...dxe5 14.d6+ Kf8 15.f4


Now the game resembles thematically a King's Gambit with the open f-file and pressure on the f7 square. If 15...exf4, then 16.Rxf4 f6 17.e5 when the threat of establishing connected passed pawns by e6 is a strong one, as well as the threat of further exposing the black king at some point by exf6. I admit that I could not see everything here. To a certain extent, I was going by intuition. White's attack would develop according to the particular defense that Black might choose. If after 15...exf4 16.Rxf4 Black tries 16...Qe8, then 17.Bxf7 Qxf7 18.Rxf7+ Kxf7 19.Nc7 should win; e.g., 19...Bb7 20.Nxa8 Bxa8 21.Qxa7+, etc. [Radomskyj]

15...Nd7 16.fxe5 Nxe5 17.Rxf7+!


Thematic and forced. It is the only way of continuing the attack. [Radomskyj]

17...Nxf7 18.Rf1

Both 18...Qd7 and 18...Qe8 lose to 19.Qb3. A fascinating alternative is 18...Kg8. Then White does not grab material on f7 but first plays 19.e5!, opening a line for his queen after Qc2. I don't see what Black can do. For instance, 19...Qg5 20.Rxf7 Kh7 21.Qc2+ Qg6 22.Bd3. [Radomskyj]

18...Qf6 19.Rxf6 gxf6 20.Nc7 Nxd6 21.Qc6!


White relentlessly pursues the attack. With Black's king so open and his pawns so weak, White's queen is particularly powerful in this sort of position. [Radomskyj]

21...Nxc4

This loses quickly, but alternatives do also. For example, 21...Ke7 22.Nd5+ Ke6 23.Nf4+ Ke5 24.Nd3+ Kd4 25.Qxd6+ Kxc4 26.Qd5+ followed by 27.Qxa8 is just one sample winning line. [Radomskyj]

22.Qxf6+ Kg8 23.Nd5


This knight, that Black failed to take off earlier, decides the game. The threat, of course, is mate in one. [Radomskyj]

23...Rh7 24.Qd8+ Kg7 25.Qe7+ Kg8


Or 25...Kg6 26.Nf4#. [Radomskyj]

26.Nf6+, Black resigns



{This article originally appeared in Atlantic Chess News in 1988}