Author: Uri Blass
Date: 06:57:27 05/06/04
Go up one level in this thread
On May 06, 2004 at 09:43:43, Anthony Cozzie wrote: >On May 06, 2004 at 09:34:23, Uri Blass wrote: > >>On May 06, 2004 at 07:44:13, Martin Giepmans wrote: >> >>>On May 05, 2004 at 21:08:48, Will Singleton wrote: >>> >>>>[D]8/8/3Rr3/6Bk/7P/5K1P/8/8 w - - 0 74 >>>> >>>>This is from ICC today. The game is already won, but Bf4 wins quicker. Amateur >>>>plays it immediately, anybody else? (not Shredder, Ruffian, CM) >>>> >>>>13 1007 149 920346 Bf4 Rxd6 Bxd6 Kxh4 Kg2 Kh5 Be7 Kg6 Kg3 Kh5 >>>>14 1007 243 1496827 Bf4 Rxd6 Bxd6 Kxh4 Kg2 Kh5 Be7 Kg6 Kg3 Kh5 >>>>15 1048 501 3162742 Bf4 Rxd6 Bxd6 Kxh4 Kg2 Kh5 Kg3 Kg5 Be5 Kf5 >>> >>> >>>SpiderChess (without tablebases) also plays it (instantly). >>>I guess that some engines overlook it because of nullmove. >>> >>>Martin >> >>I do not see null move problem. >>It is evaluation problem and movei does not know to trade pieces when it has the >>advantage. >> >>It probably can solve it if I add a bonus in the case that the opponent has only >>a king but this also means that the wrong evaluation of KBP vs K will remain. >> >>Uri > >I don't think this is an evaluation problem, but rather a search problem. Movei >has neither tablebases nor hashtables, so it has trouble. > >anthony Movei does not use hash for pruning but it is wrong to say that it does not have hash tables. The evaluation of it is more than +6 before Bf4 when even after Bf4 Rxd6 Bxd6 it needs to see promotion of the pawn to see a bigger evaluation for white. 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.