Author: Peter Berger
Date: 07:32:09 05/13/01
Go up one level in this thread
On May 13, 2001 at 10:02:56, Robert Hyatt wrote: >On May 13, 2001 at 09:24:35, Peter Berger wrote: > >>On May 13, 2001 at 09:13:02, Andrew Williams wrote: >> >>>Hello >>> >>>Here's a position from a blitz game between ChesterX and PostModernist on ICC. >>>PM is too hasty here, playing 49... Rxb3 On my Athlon 1200 it takes PM >>>34 seconds to choose g6 instead. However, this tranlates to more than two >>>minutes on the PIII 450 on which the game was played. How long does it take >>>other programs to avoid Rxb3? Please state hardware. >>> >>Bringer 1.8 on Athlon 1333 : >> >>0:00:00.1 ( 5/14) 19047 1.12 b8xb3 f3xh5 g7-g6 h5xg6 g3xg6 >>0:00:00.2 ( 6/17) 67407 0.97 b8xb3 f3xh5 f7-f8 f5-f2 >>0:00:00.7 ( 7/18) 180359 1.35 b8xb3 f3xc6 b3xc3 c6-d5 f7-f8 >>0:00:01.3 ( 8/23) 362162 0.95 b8xb3 f3xh5 >>0:00:02.0 ( 8/23) 495185 0.89 b8xb3 f3xh5 f7-e7 d2-h2 d8-g8 f5-g5 >>0:00:04.4 ( 9/25) 1094516 0.84 b8xb3 f3xh5 f7-f8 h5-f3 > > >This is a case of evaluation knowledge helping out. THe above move ends >up much worse than this.... > True - if I disable "SEE in QSearch" I get 0:00:00.1 ( 4/13) 17995 1.05 b8xb3 f3xh5 f7-e7 d1-g1 0:00:00.2 ( 5/14) 55407 1.06 b8xb3 f3xh5 f7-g8 f5-e6 g8-h8 0:00:00.3 ( 6/17) 124155 0.97 b8xb3 f3xh5 f7-f8 f5-f2 0:00:01.3 ( 7/18) 440499 1.23 b8xb3 f3xc6 b3xc3 0:00:02.6 ( 8/21) 988836 0.42 b8xb3 d1-g1 0:00:03.5 ( 8/22) 1297530 -0.13 g7-g6 0:00:08.8 ( 9/23) 3227558 0.67 g7-g6 0:00:09.8 ( 9/23) 3765624 0.89 g7-g6 f5-f4 0:00:11.6 (10/25) 4406650 0.93 g7-g6 f5-f4 g3xf4 e3xf4 b8xb3 0:00:15.2 (11/32) 5524729 1.12 g7-g6 f5-f4 g3xf4 e3xf4 b8xb3 d2xd6 0:00:18.7 (12/32) 7199627 1.14 g7-g6 f5-f4 g3xf4 e3xf4 b8xb3 0:00:30.3 (13/37) 11823534 1.12 g7-g6 f5-f4 g3xf4 e3xf4 b8xb3 d2-d3 d6-d5 h1-g2 f7-e6 d1-e1 e6-d6 e1-d1 So this would have been a little faster here and the score for Rxb3 must have been < 0.13 at ply 8 . Note : I am not the programmer of the engine so I don't really know what I am talking about :-) . I only know that disabling "SEE in Q-Search" is a safer setting for Bringer but it's usually slower in most testpositions and gives a worse branching factor overall. pete > > > >>0:00:05.3 ( 9/27) 1546776 0.85 g7-g6 >>0:00:06.2 (10/27) 1721445 0.86 g7-g6 f5-f4 g3xf4 e3xf4 b8xb3 f3xc6 >>b3xc3 h1-g2 f7-e6 c6-b5 >>0:00:11.6 (11/29) 3327446 1.12 g7-g6 f5-f4 g3xf4 e3xf4 b8xb3 f3xc6 >>0:00:15.1 (12/31) 4579675 1.20 g7-g6 f5-f4 g3xf4 e3xf4 b8xb3 d2-d3 >>d6-d5 >>0:00:22.2 (13/33) 6752077 1.36 g7-g6 f5-f4 g3xf4 e3xf4 b8xb3 d2-d3 >>d6-d5 c4xd5 c5-c4 d3-e3 c6xd5 d1-e1 d8-b8 >>0:00:33.7 (14/39) 10626823 1.21 g7-g6 f5-f4 g3xf4 e3xf4 b8xb3 d2-d3 >>d6-d5 c4xd5 c5-c4 d3-e3 c6xd5 d1-e1 d8-b8 e3-e7 f7-g8 >>0:01:24.5 (15/43) 27815802 1.15 g7-g6 f5-f4 g3xf4 e3xf4 b8xb3 d2-d3 >>d6-d5
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.