Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: The key of pawn hash tables is based on what exactly?

Author: Uri Blass

Date: 13:31:54 09/06/03

Go up one level in this thread


On September 06, 2003 at 15:44:03, Robert Hyatt wrote:

>On September 06, 2003 at 13:47:24, Uri Blass wrote:
>
>>pawns are obvious but the question is if you store also
>>1)side to move
>>2)king squares
>>
>>both 1 and 2 may be important for evaluation of unstoppable passed pawns.
>>
>>Uri
>
>
>The simple rule is this.  The signature must be based on the locations of
>_all_ pieces that influence the score.  If king position is important to
>your scoring, then king location has to be hashed in...

I understand but I prefer to store only information and not evaluation because I
think that I will change my evaluation to be not dependent only on the pawn
structure and I do not like to delete code that I write now later.

part of the information may be dependent on pawns but part of the information
may be also dependent on other factors like squares of the kings.

A possible idea is to have more than one hash and to have one hash for pawns
when I store information about pawns and one hash for pawn and king and side to
move when I store different information.

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.