Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: For Pablo Restrepo and for Programmers too (Text german)

Author: Steve Glanzfeld

Date: 10:00:06 04/25/05

Go up one level in this thread


On April 25, 2005 at 12:50:23, Kurt Utzinger wrote:

>      An interesting idea you have presented here. On the
>      other hand: such problems (time losses) if one plays
>      with the Fischer clock modud (4m+2s).

Hi. I don't think Fischer tc should be a problem. The key is: A program must
understand that it should "unbalance" the position and create a chance to
increase it's evals, even from a somewhat lower level, rather than to lose on
time with an endless number of "perfect" moves which don't result in significant
evaluation improvement.

It is difficult, because it means that the program should choose a move which
leads to a lower evaluation by "conservative" methods. I'm not sure if programs
can ever be able to understand moves like RxN (because, NOT playing RxN is based
on deep calculation too!).

I hope that programmers like Prof. Hyatt will tell us their opinion about this.

Steve



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.