Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Evidence the evaluation function is vital

Author: Chuck

Date: 04:08:13 12/12/04

Go up one level in this thread


On December 11, 2004 at 19:14:41, Anthony Cozzie wrote:

>On December 11, 2004 at 18:51:25, Dana Turnmire wrote:
>
>>In the following diagram the only correct move is Rxa1!  Genius finds it after
>>only searching 5 ply and yet the Fidelity Designer 2100 Display (6Mhz) still
>>insists on moving Ne5 after searching 8 ply and over two hours of thinking.
>>Is this not evidence of some flaw in the Designer 2100 evaluation?
>>[d]r1b1r1k1/1p3pbp/1qpp2p1/2n5/2PNP1n1/1PN3PP/3Q1PB1/BR2R1K1 b - -
>
>It mainly evidence that Genius has a good selective search.
>
>anthony

Yes, I believe the Designer 2100 was pretty much a full-width / brute-force
program, meaning that the Genius was quickly able to throw out some bad lines
while the the Designer 2100 wasted a lot of time considering every bad move.

Regards,

Chuck



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.