Author: Uri Blass
Date: 01:21:59 02/14/03
Go up one level in this thread
On February 14, 2003 at 00:43:35, Howard Exner wrote: >On February 13, 2003 at 08:47:36, Dan Andersson wrote: >Rebel test.eng on a 2400+ for maximum of 5 minutes. > >>[D]2br3r/p5k1/3b4/2pPp1pq/Pp2Q1PP/4NP2/1PB5/R2R2K1 b - - 0 1 >>Best move Qxh4 with a close game. Position is from an OTB game. > >Misses this after 5 minutes. It prefers Qh6 The same for latest Movei but after more time it found Qxh4(I did not save the analysis so I do not know how much time) > >>[D]r2q2k1/pp4b1/3np2r/3p1p2/3P1P2/2N1PBK1/PP5Q/2R4R w - - 0 1 >>Best move Qxh6. Finding it is not all that impossible. The thing is that I had >>to see it a couple of moves earlier. > >36 seconds The question is if the program find it with a positive score because I can imagine that programs can see it with a draw score faster if they evaluate black as better. Movei evaluates white as better(0.02 pawns) even without Qxh6 Latest Movei shows the following analysis on p850. depth=13 +0.03 h2h6 Nodes: 230665744 NPS: 156749 Time: 00:24:31.56 depth=13 +0.32 h2h6 Nodes: 264184766 NPS: 157804 Time: 00:27:54.13 depth=13 +0.79 h2h6 g7h6 h1h6 d8b6 c1h1 d6f7 c3d5 b6d6 h6g6 g8f8 h1h5 a8c8 h5f5 c8c2 g6f6 Nodes: 342839783 NPS: 159619 Time: 00:35:47.86 >>[D]r4rk1/5pp1/pq3n1p/5P2/1Q1ppP2/P7/1PP1B1PP/R3K2R b KQ - 0 1 >>Best move Qa7. Second best Qxb4. Team game on FICS. > >I like this position. I'm guessing most humans at 1800+ would play Qa7 just on >principle. Rooks can attack the Queen plus the white king can't immediately >castle. Test.eng misses this preferring Qxb4 same for Movei0079992 and probably something in the line Qa7 get pruned. depth=13 +0.18 b6b4 a3b4 f6d5 c2c3 d4d3 e2d1 d5e3 e1f2 e3f5 a1a5 f5d6 d1b3 f8c8 h1a1 c8c6 Nodes: 38927550 NPS: 169316 Time: 00:03:49.91 depth=14 +0.17 b6b4 a3b4 f6d5 c2c3 d4c3 b2c3 d5c3 e2a6 a8b8 e1g1 b8b4 a1c1 c3d5 c1c5 d5e3 f1e1 Nodes: 95873948 NPS: 172500 Time: 00:09:15.79 depth=14 +0.17 b6b4 a3b4 f6d5 c2c3 d4c3 b2c3 d5c3 e2a6 a8b8 e1g1 b8b4 a1c1 c3d5 c1c5 d5e3 f1e1 Nodes: 146720884 NPS: 169187 Time: 00:14:27.21 After Qa7 depth=10 +0.00 a1d1 f8d8 g2g3 a8b8 b4a5 a7e7 a5e5 e7a7 e5a5 Nodes: 5047361 NPS: 149197 Time: 00:00:33.83 depth=10 +0.00 a1d1 f8d8 g2g3 a8b8 b4a5 a7e7 a5e5 e7a7 e5a5 Nodes: 13458886 NPS: 148520 Time: 00:01:30.62 depth=11 -0.08 a1d1 f8d8 b4c4 a6a5 c4a4 a7b8 d1d4 b8b2 d4d8 a8d8 a4a5 b2c1 e1f2 c1f4 f2g1 f4e3 g1f1 Nodes: 18006912 NPS: 146887 Time: 00:02:02.59 depth=11 -0.08 a1d1 f8d8 b4c4 a6a5 c4a4 a7b8 d1d4 b8b2 d4d8 a8d8 a4a5 b2c1 e1f2 c1f4 f2g1 f4e3 g1f1 Nodes: 36754058 NPS: 148046 Time: 00:04:08.26 depth=12 -0.23 a1d1 f8d8 b4c4 a7b8 d1d4 b8f4 c4c3 d8c8 c3b3 c8b8 b3g3 f4g3 h2g3 b8b2 Nodes: 42668709 NPS: 147944 Time: 00:04:48.41 depth=12 -0.23 a1d1 f8d8 b4c4 a7b8 d1d4 b8f4 c4c3 d8c8 c3b3 c8b8 b3g3 f4g3 h2g3 b8b2 Nodes: 70086600 NPS: 150565 Time: 00:07:45.49 depth=13 -0.43 a1d1 f8d8 b4a4 a7b8 e1g1 b8b2 e2c4 b2b6 a4b4 b6b4 a3b4 e4e3 d1c1 Nodes: 90568825 NPS: 150451 Time: 00:10:01.98 depth=13 -0.43 a1d1 f8d8 b4a4 a7b8 e1g1 b8b2 e2c4 b2b6 a4b4 b6b4 a3b4 e4e3 d1c1 Nodes: 203782392 NPS: 151893 Time: 00:22:21.61 > >>[D]8/4k3/1p6/p1pPPp1p/2P1p3/4P2P/P1K5/8 b - - 0 1 >>[D]8/4k3/1p6/p1pPPp2/2P1p2p/4P2P/P1K5/8 w - - 0 1 >>[D]8/4k3/1p6/p1pPP1p1/2P1p2p/4P2P/P1K5/8 w - - 0 1 >>Analysis of the REBEL-Web game. And variations of the position. All drawn. The >>eval of programs are usually way off. > >Rebel is -1.45 after 5 minutes starting with h4. Movei's weakest point is the endgame so I even did not try. > >>[D]r6r/4pk2/ppp3pB/q3P3/2pP4/1bN3Q1/1R4P1/6K1 w - - 0 33 >>Best move e6+. How long to play. And does it see that black is worse? Game >>against GT. > >played afgter 3 seconds. eval of 0.49 after 1:13 Played after less than 3 seconds eval start as negative but increase to 0.00 after a short time depth=8 +0.00 e5e6 f7e6 b2e2 e6d7 e2e7 d7e7 g3c7 e7e6 c7c6 e6e7 c6c7 e7e8 c7c6 e8d8 c6d6 d8c8 d6c6 c8d8 Nodes: 3911879 NPS: 183140 Time: 00:00:21.36 depth=9 +0.00 e5e6 f7e6 b2e2 e6d7 e2e7 d7e7 g3c7 e7e6 c7c6 e6e7 c6c7 e7e8 c7c6 e8d8 c6d6 d8c8 d6e6 c8c7 e6e7 c7c6 e7e6 c6c7 Nodes: 9179483 NPS: 186954 Time: 00:00:49.10 depth=10 +0.00 e5e6 f7e6 b2e2 e6d7 e2e7 d7e7 g3c7 e7e6 c7c6 e6e7 c6c7 e7e8 c7c6 e8d8 c6d6 d8c8 d6e6 c8d8 e6d6 Nodes: 28193105 NPS: 188292 Time: 00:02:29.73 depth=10 +0.00 e5e6 f7e6 b2e2 e6d7 e2e7 d7e7 g3c7 e7e6 c7c6 e6e7 c6c7 e7e8 c7c6 e8d8 c6d6 d8c8 d6e6 c8d8 e6d6 Nodes: 30540156 NPS: 187789 Time: 00:02:42.63 Uri
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.