Author: Albert Silver
Date: 12:13:25 12/27/05
Go up one level in this thread
On December 27, 2005 at 13:55:25, Stuart Cracraft wrote: >On December 26, 2005 at 08:36:35, Albert Silver wrote: > >>This position and combination took place in a match against Junior 8: >> >>[D]2r3k1/2qb4/p3prp1/3p2b1/1B1P2P1/1Q6/1P2NPK1/3RR3 b - - 0 32 >> >>As a human, being told "Black to play and win" makes finding the key move very >>easy, so all I can suggest is to try and work out the correct follow-up. >> >>Some engines shuld have more trouble than others on this all the same. At least >>according to some preliminary checking I did. On my computer, Rybka spotted it >>in 5 seconds BTW. >> >> Albert > >My program is too poor to "get it": > >Alpha=-394 Beta=406 Maxdepth=9999999 MaxTime=99900 xboard=0 >Itr/Max Mv Time Score Nodes PV > 1/ 4 a6a5 6 0.00 55 a6a5 > 2/ 5 c7c4 6 0.00 547 c7c4 b3c4 d5c4 > 3/ 9 c7c2 6 0.04 3650 c7c2 b3c2 c8c2 d1b1 > 4/16 c8b8 6 0.10 18583 c8b8 d1d3 > 5/16 c8b8 6 0.20 44271 c8b8 e1f1 c7c4 b3c4 d5c4 d1b1 > 6/19 g5h4 20 1.54 370363 g5h4 e1f1 c7b6 d1b1 c8b8 f2f4 > 7/21> g5h4 420 1.94 465288 g5h4 e1f1 c7b6 d1b1 c8b8 f2f4 a6a5 > 7/28 g5h4 980 5.33 1246768 g5h4 e2g3 f6f4 b3d3 f4g4 d3a6 h4g3 f2g3 g4g3 g2f2 > 8/28 d7b5 1018 20.99 5050579 d7b5 b3g3 c7g3 g2g3 c8c2 e2g1 f6f2 g1f3 f2g2 >g3h3 g5f4 b4c3 > 9/29 d7b5 1032 45.42 10996834 d7b5 b3g3 c7g3 e2g3 c8c2 g3h1 b5e2 b4e7 e2d1 >e7f6 >10/33 d7b5 1006 128.85 31370163 d7b5 b3g3 c7g3 e2g3 c8c2 g3h1 b5e2 d1b1 f6f4 >h1g3 e2f3 g2g1 f3g4 >11/ Some engines are faster than others. I'd give it more time. 10 plies is probably not deep enough anyhow. Here are some comparative solution times and depths on an Athlon64 3500+ with 512 Mb hash: Analysis by Toga II 1.1: 32...Bb5 33.f4 Bxf4 34.Bc5 Bg5 35.Qg3 Bc4 36.Ra1 Qh7 37.Rh1 Qb7 38.Qe5 Qxb2 -+ (-1.64) Depth: 13/42 00:00:14 12750kN 32...Rxf2+ 33.Kxf2 Qh2+ 34.Kf1 Qh1+ 35.Ng1 Bb5+ 36.Rd3 Rc7 37.Rxe6 Rf7+ 38.Ke2 Qxg1 39.Bd2 Bxd3+ 40.Kxd3 Rf3+ 41.Re3 Rxe3+ 42.Bxe3 Bxe3 43.Qxd5+ Kh7 44.Qf7+ Kh6 -+ (-2.16) Depth: 13/51 00:00:28 25927kN 32...Rxf2+ 33.Kxf2 Qh2+ 34.Kf1 Qh1+ 35.Ng1 Bb5+ 36.Rd3 Rc7 37.Re2 Rf7+ 38.Rf3 Bxe2+ 39.Kxe2 Qg2+ 40.Kd3 Qxg1 41.Rxf7 Kxf7 42.Ba5 Qe3+ 43.Kc2 Qxd4 44.Qb7+ Be7 45.Bc3 Qe4+ 46.Kc1 -+ (-2.52) Depth: 14/56 00:01:06 62572kN Analysis by Fruit 2.2.1: 32...Bb5 33.Qg3 Qxg3+ 34.Nxg3 Rc2 35.Nh1 Kf7 36.Bd6 Be2 37.Ra1 Bxg4 38.Rxa6 Rxb2 39.Ra7+ Kg8 40.Kg3 Be2 -+ (-1.67) Depth: 14/46 00:00:33 26865kN 32...Rxf2+ 33.Kxf2 Qh2+ 34.Kf1 Qh1+ 35.Ng1 Bb5+ 36.Rd3 Rc7 37.Re2 Rf7+ 38.Rf3 Bxe2+ 39.Kxe2 Qg2+ 40.Kd3 Qxg1 41.Rxf7 Kxf7 42.Ba5 Qe3+ 43.Kc2 Qxd4 44.Qb7+ Be7 45.Bc3 Qe4+ 46.Kc1 -+ (-2.53) Depth: 14/54 00:01:45 88611kN Analysis by Junior 8: 32...Bb5 33.Qh3 Qc2 34.Be7 Bxe2 35.Bxf6 Bxf6 36.g5 Rc6 37.gxf6 Bxd1 38.Rxe6 Be2 39.Re8+ Kf7 ยต (-1.26) Depth: 15 00:00:11 22224kN 32...Bh4 33.Rf1 Qc2 34.Qe3 Rxf2+ 35.Rxf2 Bxf2 36.Qd2 Bh4 37.Nc3 Qxd2+ 38.Rxd2 Bg5 39.Re2 Bf6 40.Bc5 Kf7 -+ (-1.58) Depth: 15 00:00:50 101134kN 32...Bh4 33.Rf1 Qc2 34.Qe3 Rxf2+ 35.Rxf2 Bxf2 36.Qd2 Bh4 37.Qxc2 Rxc2 38.Rd2 Rc6 39.Rd3 Rc4 40.Bd6 Ba4 -+ (-1.61) Depth: 16 00:02:02 230283kN 32...Rxf2+! -+ (-1.88) Depth: 16 00:02:27 278729kN 32...Rxf2+ 33.Kxf2 Qh2+ 34.Kf1 Qh1+ 35.Ng1 Bb5+ 36.Rd3 Rc7 37.Re2 Rf7+ 38.Rf3 Bxe2+ 39.Kxe2 Qg2+ 40.Kd3 Rxf3+ 41.Nxf3 Qxf3+ -+ (-2.79) Depth: 16 00:02:49 324962kN Analysis by Deep Fritz 8: 32...Bb5 33.f4 Bxe2 34.Rxe2 Qxf4 35.Qg3 Rc2 36.Qxf4 Rxe2+ -+ (-1.98) Depth: 11/33 00:00:06 7566kN 32...Bb5-- -+ (-1.70) Depth: 12/33 00:00:14 19654kN 32...Bb5 33.Qh3 Qc2 34.Be7 Rxf2+ 35.Kxf2 Bxe2 36.Rc1 Bxg4+ 37.Rxc2 Rxc2+ 38.Kg3 Bxh3 39.Bxg5 Bf5 40.Ra1 Bd3 41.b3 -+ (-1.49) Depth: 12/42 00:00:17 24271kN 32...Rxf2+! -+ (-1.49) Depth: 12/42 00:00:33 45753kN 32...Rxf2+! -+ (-1.65) Depth: 12/42 00:00:33 45948kN 32...Rxf2+! -+ (-1.96) Depth: 12/42 00:00:33 46226kN 32...Rxf2+ 33.Kxf2 Qh2+ 34.Kf1 Qh1+ 35.Ng1 Bb5+ 36.Rd3 Bh4 37.Re2 Rc1+ 38.Be1 -+ (-2.30) Depth: 12/42 00:00:35 48633kN 32...Rxf2+! -+ (-2.59) Depth: 13/37 00:00:39 53683kN
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.