AylerKupp: <<Autoreparaturwerkbau> I agree resigning after somewhat timid pawn move 25.g6 does look unimpressive.> I agree with you and "your" Stockfish. I ran an analysis from the final position, although not quite as long nor as deep as yours and this is what I got with Houdini 6, Komodo 12.3, in addition to "my" Stockfish 10:
 click for larger view<Houdini 6>, d=25: 1. [+2.65]: 25. ... h6 26.gxf7+ Rxf7 27.Rg2 Kh8 28.Qg3 Nh7 29.Qg6 Qg8 30.Qxe6 b5 31.Qh3 Re8 32.Qh2 Bc8 33.a3 Rxf3 34.Bxh6 Re7 35.Rh4 Qf8 36.Bf4 Bf5 37.Be2 Rh3 38.Rxh3 Bxh3 39.Rf2 Bf5 40.Bd2 Qc8 41.Bd1 Rc7 2. [+3.92]: 25. ... Rc4 26.gxh7+ Kh8 27.Bxc4 dxc4 28.Rxc4 b5 29.Rf4 g6 30.Rd2 Bd5 31.Qg3 a5 32.b3 g5 33.Rg4 Nxh7 34.Rdd4 Bc6 35.Rxd7 Qxd7 36.Bxg5 Nxg5 37.Rxg5 Qd8 3. [+4.28]: 25. ... Kh8 26.gxh7 Rc4 27.Qg1 Rxd4 28.Bg5 f6 29.Bxf6 Rxd3 30.Bxd8 Rxf3 31.h6 Nxh7 32.b3 g5 33.Bxg5 Nxg5 34.Qxg5 Rf8 35.Qg4 Re7 36.Qb4 Rf1+ 37.Kb2 Rff7 38.Qxb6 Kh7 39.Qd8 d4 40.Qxd4 Rd7 41.Qb6 Rfe7 <Komodo 12.3>, d=29: 1. [+4.29]: 25. ... Rc4 26.gxf7+ Kxf7 27.Bxc4 dxc4 28.Rxc4 Kg8 29.Qg3 Rf7 30.h6 b5 31.Rf4 Re7 32.hxg7 Rxg7 33.Rg4 Rg6 34.b3 Qe7 35.Bg5 Qf7 36.Rh3 a5 37.Bh6 Rxg4 38.Qxg4+ Ng6 39.Qd4 Qc7 40.Qd6 Qc8 41.Bd2 a4 42.bxa4 bxa4 43.Rg3 2. [+4.81]: 25. ... Kh8 26.gxh7 Rc4 27.Bxc4 dxc4 28.Rxc4 b5 29.Rf4 Nxh7 30.Qg3 Qc7 31.Rg4 f6 32.h6 g5 33.exf6 Qxg3 34.Rxg3 Rf7 35.b3 Nxf6 36.Bxg5 b4 37.Bd2 a5 38.c4 e5 39.Kb2 Kh7 40.Rhg2 Nh5 41.Rg5 a4 42.Rxh5 Bxf3 43.Rgh2 a3+ 44.Kc2 Be4+ 45.Kd1 Bf3+ 46.Ke1 Bxh5 3. [+5.75]: 25. ... h6 26.gxf7+ Rxf7 27.Rg2 Kh8 28.Qg3 Nh7 29.Bxh7 Kxh7 30.Bxh6 Qe7 31.Bg5 Qc5 32.Be3 Qf8 33.Qg6+ Kh8 34.Bg5 Rc4 35.Rxc4 dxc4 36.h6 Rxf3 37.hxg7+ Qxg7 38.Qh5+ Kg8 39.Rg4 Bd5 40.Bc1 Qxg4 41.Qxg4+ Kf7 42.Qh5+ Kg8 43.Qg6+ Kf8 44.Bh6+ Ke7 45.Qg1 Kf7 46.Qxb6 Rf1+ 47.Bc1 Rh1 48.Qxa6 c3 49.bxc3 <Stockfish 10>, d=35:: 1. [+7.67]: 25. ... Rc4 26.gxh7+ Kh8 27.Bxc4 dxc4 28.Rxc4 Bd5 29.Rf4 Qb8 30.Rg2 Nxh7 31.Qg3 Qxe5 32.Rc4 Qxg3 33.Rc8+ Nf8 34.Rxf8+ Kh7 35.Rxg3 e5 36.b3 Rd6 37.c4 Be6 38.f4 exf4 39.Bxf4 Rc6 40.Rb8 Rc5 41.Rxb6 Rxh5 42.Rxa6 g6 43.Re3 Rf5 44.Bg3 Kg7 45.Kb2 Kf6 46.Rd6 Kg7 2. [+8.28]: 25. ... h6 26.gxf7+ Kxf7 27.Rg2 Ke8 28.Rdg4 Rcc7 29.Qg1 Qe7 30.Rxg7 Qxg7 31.Rxg7 Rxg7 32.Qxb6 d4 33.a4 Rc6 34.Qxd4 Rd7 35.Qg4 Rxd3 36.cxd3 Rxc1+ 37.Kxc1 Ba8 38.Qf4 Bc6 39.Kd2 Bd5 40.Qxh6 Bc6 41.Qf4 a5 42.Kc1 Nd7 43.h6 Nc5 3. [+9.77]: 25. ... Kh8 26.Rg4 Rc4 27.Rgg2 Qb8 28.gxh7 Nxh7 29.Qd2 Qxe5 30.Bxc4 Nf8 31.Bd3 f5 32.h6 g6 33.h7 Rf7 34.Qh6 Bc6 35.Qh4 Qc7 36.Bh6 Be8 37.Qb4 Nd7 38.Be3 Rg7 39.Qh4 Qd6 40.Bh6 Qe7 41.Qd4 e5 42.Qxd5 Rxh7 43.Bg5 Qf7 44.Qa8 Kg7 45.Rxh7+ Kxh7 46.Qxa6 And, the deeper the search, the more Black's position is considered hopeless for Black, although I suspect that his resignation was based on practical considerations, principles, historical precedents, and the reputation of his opponent (Anand had either won or would soon win the 2000 FIDE WCC tournament) rather than deep precise calculations; many games have been won by White in similar positions. Imagine having that position as Black against either the current of future WCC! And I think that it was Tal that popularized the lines with g5-g6 and then h5-h6 to destroy the pawn barrier around Black's king. Besides, with White's king secure on the q-side and well protected by his own pieces, there is effective no hope for Black's counterplay. As I'm sure you know but others might not, Stockfish considered 25...h6 to be Black's best move until , on my computer, d=29. Then it switched to considering 25...Rc4 to be Black's best move starting at d=30, with the evals between the 2 moves starting to widen. Eventually I'm sure that Houdini and Komodo would have agreed with that. |