Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Endgame code instead of Tablebases

Author: KarinsDad

Date: 13:48:14 04/16/99

Go up one level in this thread


On April 16, 1999 at 15:57:15, blass uri wrote:

>
>On April 16, 1999 at 14:04:52, KarinsDad wrote:
>
>
>
>>It depends on how many cases there are. Currently, there are several Gig of
>>tablebase files, so if 95% is just the listing of the positions and the
>>result(win, draw, or loss) and the other 5% is the best move to make from that
>>position, it would still require several gig to just have all of the positions
>>and the associated result. However, this could be pruned considerable. In the KR
>>vs. K solution, it is only a draw if the opposing king can take the rook
>>immediately, otherwise, it is a win.
>
>It is a draw also if it is a stalemate and the opposing side cannot take the
>rook.

Yes, how silly of me.

>
>I think that it is not practical to try to define rules to detemine  100% of the
>cases in complicate endgames like KR vs KN
>
>trying to define rules that are correct only for 99% of the cases and saving
>only the exceptions in the tablebases seems to be more practical.

Agreed.

>
>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.