Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Question about iterative deepening

Author: Robert Pope

Date: 07:41:55 07/30/03

Go up one level in this thread


On July 29, 2003 at 23:15:38, Bruce Moreland wrote:

>On July 29, 2003 at 21:11:23, Mathieu Pagé wrote:
>
>>Hi,
>>
>>I have this question about iterative deepening:
>>
>>When I have search a position at, let say, 8 ply and i had searched half of the
>>moves at ply 9 and none of those move return a better score than the best move
>>at 8 ply wich move should I choose ? The best at ply 8 or the best at ply 9 even
>>if I have not search all moves ?
>>
>>someone have an hint ?
>>
>>Thanks for your help.
>>
>>Mathieu Pagé
>
>Moves at ply 9 are not in competition with moves at ply 8.
>
>When you search all of ply 8, you know what the best move is at ply 8, so if
>your time runs out here, you make the best ply 8 move.
>
>When you start ply 9, your search can end before you finish searching the first
>move in ply 9.  If this happens, you make the best move from ply 8.  If you've
>finished one or more moves in ply 9, you make the one of those with the best
>score.
>
>If something weird happens, like the first move in ply 9 returns a hugely low
>score, you might want to search more moves, even if your time management system
>thinks you really should move now.  It's better to think a little longer than it
>is to walk into a loss.
>
>bruce

Don't forget that when you abort a depth 9 search because you ran out of
allotted time, the move you were in the middle of examining can't be used as the
winner of the search.  Since its alpha-beta search got cut short, the value that
gets return at depth 1 is meaningless (but can easily be higher than any of the
prior fully searched moves).



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.