Author: Uri Blass
Date: 11:42:48 01/22/02
Go up one level in this thread
On January 22, 2002 at 14:27:28, David Rasmussen wrote: >On January 22, 2002 at 02:55:40, David Rasmussen wrote: > >>[D]8/r7/8/p2k4/R3p1K1/8/8/8 b - - 0 67 >> > >There are some programs (CM, Averno and Tiger, and maybe others) that "finds" >Kc5. But do they understand it? What is the score? What are the lines that are >being show? I don't think any of the programs understands what is happening. > >/David Chessmaster and tiger understand enough to play Kc5 and it means that their positional score is bigger than 1 pawn. If you do not use tablebase and use only material score you will find the following(I assume that you do not use hours for searching in this discussion): 1)Kc5 is losing a pawn. 2)Ke5 does not lose a pawn. The fact that programs can find Kc5 with no tablebases in few seconds prove that they have positional score that is bigger than a pawn. They understand that losing a pawn gives better probability to win based on their evaluation and it is better than what your program understood. It is better not to "understand" too much because in another position your "understanding" may lead you to trouble and this is the reason that when professional programs see a mistake they may change their evaluation to avoid the mistake but they are careful not to change it too much. Uri
This page took 0.01 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.