Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Fruit 2 and endgame play

Author: Dann Corbit

Date: 10:52:16 01/12/05

Go up one level in this thread


On January 12, 2005 at 11:15:22, Uri Blass wrote:

>On January 12, 2005 at 08:21:58, Norm Pollock wrote:
>
>>On January 12, 2005 at 07:43:57, Norm Pollock wrote:
>>
>>Then there is other matter of wear and tear on
>>>the disk drives.
>>
>>Let me clarify this. 5 man egtb starts disk accessing with as many as 14 pieces
>>still on the board in a blitz game.
>>
>>When there are only 5 pieces or less, then there is hardly any disk access and
>>egtbs are clearly better than searching.
>>
>>But for 6-14 pieces on the board, disk accessing is taking away time
>>(substantial time in blitz) from searching (and wearing out your disk drive). I
>>think egtbs are likely a disadvantage to a strong program when there are 6-14
>>pieces on the board, but are an advantage with 5 (or less) pieces on the board.
>
>Nobody force programs to access the disk when there are 6-14 pieces
>I expect programmers who are intelligent enough to make their program probe the
>tablebases to be also intelligent enough to not probe the tablebases if it is a
>disadvantage.
>
>I got permission to use the nalimov tablebases as long as movei is not
>commercial and my first implementation is probably going to be using it only
>when there are 5 or less pieces in the board.

I think it wise to test at least the pv also, when the pv has 5 men or less in
it.  Even if it is 100 plies long, that is only 100 probes and completely
insignificant for time.





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.