Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Evaluation functions and databases

Author: Roberto Waldteufel

Date: 04:41:04 12/21/98

Go up one level in this thread



On December 21, 1998 at 01:49:14, Lawrence S. Tamarkin wrote:

>On December 20, 1998 at 23:35:02, Steve Lopez wrote:
>
>>On December 20, 1998 at 20:49:33, James Robertson wrote:
>>
>>>On December 20, 1998 at 19:14:34, Mario Petrilli wrote:
>>>
>>>>I've noticed that quite a few of the professional chess programs claim as a
>>>>positive feature of their software package the inclusion of a large games
>>>>database.  Why is this not a demerit of the evaluation function?  That is, if
>>>>the evaluation function is sufficiently strong, it should not require a
>>>>database.  Rather, it should perform an evaluation purely based on the >intrinsic properties of the chess position.
>>>>
>>>>Anyone with knowledge of evaluation functions please comment.
>>>
>>>The database is for you, not the program!!!
>>>
>>>James
>>
>>Exactly. While some playing programs will use the database as a reference in the
>>learning functions, the primary purpose of a large database of games is for the
>>user's enjoyment and education.
>
>
>Does Fritz5.32 have exactly the same 300,000 game database as the 5.16 version.
>If it has even more games I would be so happy...
>
>mrslug - the inkompetent chess software addict!

Hi Larry,

Games? You want more games? Well I don't know about Fritz, but you should go to
Dan Corbit's ftp site (see the CCC Resource centre for the URL) where you can
get games galore, all sorted by ECO code. This is a seriously huge collection,
so if you want it all, I hope you have a fast modem!

Best wishes,
Roberto



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.