Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Writing a meta-language to describe eval function

Author: Vincent Diepeveen

Date: 17:44:42 07/02/03

Go up one level in this thread


On July 02, 2003 at 14:28:21, Daniel Clausen wrote:

>Hey, that must be the longest version of "It doesn't work because it didn't for
>me or because I thought about it for 2 minutes and didn't come up with a simple
>solution" so far.
>
>Congrats! =)
>
>Sargon

Don't write bullshit. If anyone knows something from evaluation then it's me.

Then we had Che and similar experiments.

In the draughts world there was such an experiment for years with several
draughts programs. They copied the idea from each other. It didn't work either
in the end.

DIEP version 1.2 or something had a very indirect way to formulate knowledge, it
took me a 6 months or so to find out how stupid that was. All you need is a few
good defines and not using bitboards.

It's of course only the bitboard guys that raise their hands for such
experiments.

Somehow some people keep reintroducing the same experiment over and over again
and they still do not see the weak points that need to get adressed first.

You for sure know shit here.



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.