Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Correspondance chess computerchess match, 30' per move

Author: Uri Blass

Date: 05:58:15 12/18/02

Go up one level in this thread


On December 18, 2002 at 07:48:32, Thorsten Czub wrote:

>http://f50.parsimony.net/forum200318/messages/1587.htm
>
>we are doing a match between a few interesting chess programs.
>all programs get 30' per move on a 1200 mhz machine.
>
>instead of using BIG BOOKS we use 1.a3 a6 as the beginning position.
>all computer programs are out of book after these forced moves and have
>to compute from move 2.
>
>Maybe we will add some more interesting programs.
>
>If you have ideas which one, please tell us.
>
>(some programs use new styling!)
>
>thank you,
>Thorsten Czub

Questions:
1)Do the programs know that they get exactly 30 minutes per move?

There are ideas that can be important if the program knows that it is going to
get exactly 30 minutes.

suppose that a program fail low after 29 minutes and find a score of -7.00 pawns
against itself.
In normal condition it may try to finish the iteration but in conditions of
exactly 30 minutes per move it may be better to calculate the value of second
best move at reduced depth and play it because it may have no time to find
another move at the same depth.

2)How do you use Junior7

Do you use it in analysis mode or in a specific time control when you stop it
after 30 minutes(for example 1 hour/1 move+1 hour/1 move+24 hours per game).


I do not know which is the best way to use Junior at that time control but you
may get different results because Junior does different jumps in the depth when
it assumes a specific time control.

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.