Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Just a small question about chess program logic

Author: Ricardo Gibert

Date: 13:48:56 05/02/02

Go up one level in this thread


On May 02, 2002 at 12:26:35, Daniel Clausen wrote:

>On May 02, 2002 at 12:13:48, Jeroen van Dorp wrote:
>
>>[D]8/6p1/5pk1/8/2r5/4PP1P/R1p3PK/8 w - - 0 47
>>Pocket ChessPartner v1.0 - CENTURY 4.0
>>
>>Here Century pondered 47. Ra1 c1Q 48. Rxc1 for it's white opponent instead of
>>47. Rxc2 Rxc2.
>>
>>I wonder why: is it because Century might think (in a twisted computer chess
>>way) "swapping a rook for a queen is more valuable than swapping a rook for a
>>pawn", or because it pondered the chance that black makes a mistake, doesn't
>>promote, and white can play Rc1, disabling promotion for the time being?
>>
>>Just a small question about computer chess progam logic (I hope).
>>
>>Thanks.
>>
>>J.
>
>Not sure it's the reason here, but often it's this: If a side has a huge
>material advantage, the score increases steadily ply after ply. Since it takes
>two more plies to reach the "one rook up" advantage in the first variation,
>Century got a worse score for the 2nd variation and therefore chose variation 1.
>
>That or it's something different. :) [ie 1-ply-swindling-mode :)]
>
>Sargon

Interesting possible explanation. Good one!



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.