The finish in
Wilterdink - MNb, corr NBC I-521, is a tactical gem. And the game is relevant for theory:
1.e4 d5 2.exd5 Nf6 3.d4 Nxd5 4.c4 Nb4 5.Qa4+!, says Bent Larsen in the first edition of
ECO B (1975).
5...N8c6 6.a3Another ! from Larsen.
6...Na6 7.Be3The alternative 7.Nf3 also earns a +/- from Larsen, today it is at best a += (Blatny). Another possibility is 7.d5 Nc5 8.Qd1.
7...Bd7 8.Qc2Here Larsen's variation ends with the assessment +/-. He is wrong.
8.Qd1 e5 9.d5 Nd4 (or 9...Ne7), and Black cannot be worse.
8...e5 9.dxe59.d5 Nd4 10.Lxd4 exd4 11.b4 c5! =, Lahi - Smolensky, corr 1990.
9....Qe7Maybe too ambitious. Black had a simpler option: 9...Bc5!, winning back the pawn without committing himself to castling long.
10.Nf3 0-0-0 11.Be2 Nxe5 12.h3?More critical is 12.Nc3!. Black has several options, yet I haven't found a convincing continuation.
12...Nxf3+Or 12...Qf6 13. 0-0 Bc5 14.Nc3 Bxe3 15.fxe3 Qh6, about equal.
13.Bxf3 Qf6 14.Nc3Eventually White will regret that he hasn't castled short. Safer was 14.0-0 Nc5 15.Nc3 Bf5 16.Bg4, about =.
14...Bf5! 15.Qb3? White is running into difficulties. He should have played 15.Qa4 c6 and then:
(a) 16.Rd1 Qg6 17.Ne2 Bb4+!! 18.axb4 Bc2 19.Bg4+ Kb8 20.Nf4 Bxa4 21.Nxg6 hxg6 22.Ra1 Bb3 23.b5 f5 24.Be2 Nb4 25.0-0 Nc2 26.Rxa7 f4 27.Bf3 fxe3 28.Rxb7+ Kc8 29.fxe3 Nxe3 30.Bc6 Rhf8 31.Ra1 Rd1+ 32.Rxd1 Bxd1 33.Ra7 Kd8 34.c5 Nc4 35.b6 cxb6 36.Rd7+ Kc8 37.Rxd1 bxc5 +=.
(b) 16.0-0-0 Nc5 17.Qb4 Rd4!!
18.Nd5 (18.Bxd4 Qxd4 19.Nb5 Qf4+ 20.Rd2 Be4! 21.Nxe4 Nxe4 22.Nxa7+ Kb8 23.Nc6+, drawn - eternal check) 18...Qe5 19.Bxd4 Qxd4 20. Rhe1 Bd6, and White has nothing better than 21.Ne7+ Kd7 22.Qb5+ c6 23.Bxc6+ Kd8 24.Re3 Bxe7 25.Be4 Bg5 26.Bxd3 Bxe3+ 27.Kb1 Nxd3 28.fxe3 Qe4 29.Qb3 Kc7 30.Qxd3 Re8, and Black holds the ending.
15...c6 16.Ne4 Qg6 17.Ng3An error in a bad position.
17...Rd3 18.Qa4 Rxe3+! 19.fxe3 Qxg3+ 20.Ke2 Bd3+! 21.Kd1 Bc5 22.Bxc6 Rd8 23.Bd5 Qf2White resigns. If 24.Re1 ...
24...Nb4!! 25.axb4 Bxb4 26.Bxb7+ Kc7 27.Qc6+ Kb8, and it's over.