Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: computer programs cannot see a very simple draw in a pawn ending

Author: Robert Hyatt

Date: 12:06:37 06/23/98

Go up one level in this thread


On June 23, 1998 at 12:41:39, Bruce Moreland wrote:

>
>On June 22, 1998 at 16:11:47, John Stanback wrote:
>
>>This is a nice position for testing/debugging transposition
>>tables and draw by 50 move rule.  I modified Zarkov to allow
>>it to search beyond 100 plies, but it didn't find the draw.
>>Then I added 1 line of code to store positions which were
>>scored by the 50 move rule or draw by repetition in the
>>transposition table before returning from search().  Now it
>>gets a score of 0 after a 101 ply search, which takes about
>>6.7 seconds.
>
>I don't store them.  I figure that the actual exact point where you find a
>50-move or rep draw is totally path dependent, so I figure it doesn't make sense
>to store the value there.
>
>Since mine works here, there is more to this, I think.
>
>bruce



if you don't store 50-move draws, then you shouldn't store 3-fold (2-fold)
repetitions either, as they are also *wrong*.  I accept the errors, since
I have two choices:  (1) be correct in hashing and miss or overlook draws,
or (2) be incorrect in hashing and find draws that might not be there.

When I have to error, I error on the side that makes me faster....

The correct thing is to never store a draw score, since they are *all*
path dependent except for insufficient material.



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.