Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: An Old idea for chess programs

Author: blass uri

Date: 22:44:02 10/20/98

Go up one level in this thread



On October 20, 1998 at 20:05:24, William H Rogers wrote:

>I did not mean to imply that getting a faster machine was the only thing that
>influenced people, although when a programmer comes to a mental block for new
>ideas, that may be his only solution.
>In the freeware testing of games Rebel beat MChess without ever leaving his
>books. That turned Rebel in that game only, as a device for looking up moves.
>The real engine of the program did not ever come into play.
>Almost all chess programs give you the option of turning off the books and I
>think that the real strength of the program then comes into view. That is all
>that I was suggesting.  MChess is not a simple minded program, but it fought
>against a proven line of game logic and lost before the other program started to
>use its logical engine.
>If you wrote a program that only had a limited book and had to compete with
>other programs with unlimited books, the advantage would be thiers, reguardless
>of your programs playing strength.

Only if they have a good book.
Sometimes a book can be counter productive.
I saw cases when a program left book with a lost position and this is what
happened to Mchess.

Programs should not use their books without checking it.
Crafty does it and use 10% of its time for analyzing when it is on book.

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.