Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Longest Checkmate in Chess from DTM Endgames?

Author: Ed Trice

Date: 19:58:06 02/04/04

Go up one level in this thread


Hello Guy,


>
>   - Checkers, 8-man EGTs done, 9-man in progress but not with DTC/M
>         ditto
>


Just a correction on the checkers data. The 10-piece databases are completed
(Schaeffer, 2003, Trice-Dodgen, 2003) see Advances in Computer Games 10.

These databases are "win-loss-draw" only.

Ed Gilbert has computed DTC for up to 9-pieces, but only stored positions with
DTC>= 30 ply, compressing it to a manageable 1.2 GB. If the DTC < 30 ply his
program just searches on the position.

Murray Cash has DTC for up to 8 pieces, but I do not know the size of these
databases.

Trice and Dodgen have DTM for up to 7 pieces that have been verified (about 21
billion positions). It is very hard to go beyond this for DTM for a variety of
reasons.

As far as we know, no other team has DTM for any checkers databases since it is
rather difficult to determine a "stopping condition" for DTM in checkers. In
checkers, win lengths can change as your computation progresses, since you don't
always have perfect information that correlates to your iteration depth.



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.