Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: gothmog's eval

Author: Sune Fischer

Date: 07:36:29 02/18/04

Go up one level in this thread


On February 18, 2004 at 09:33:59, martin fierz wrote:

>On February 18, 2004 at 09:29:33, Sune Fischer wrote:
>
>>On February 18, 2004 at 02:00:34, martin fierz wrote:
>>>
>>>sorry to have brought this up - gothmog is fine in this respect. my memory was
>>>faulty - it was a game against frenzee 146 where this happened, and frenzee
>>>thought it was winning with KRN-KR.
>>>
>>>cheers
>>>  martin
>>
>>The new version knows all these endgames.
>
>hi sune
>
>good to hear that. the problem is of course that if you put together a list of
>opponents to test against, you can't update them all the time as you would lose
>your comparison to earlier versions of your program. that's why i'm still
>testing with frenzee 146 and not with your latest.
>
>cheers
>  martin


Yeah I do the same :)

Just wanted to let you know the new one has become a lot more knowledgable
especially in these classicly drawn endgames. It was one of the things that
bothered me a lot so seeing it mentioned here I had to bring it up :)

Not all of it is 100% 'sound' because that would require a million different
rules, but most of the time it takes only a 1-2 ply search to avoid losing
material.

-S.



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.