Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: An Experiment that disproves Hyatt's 1000X NPS Theory

Author: Chan Rasjid

Date: 10:11:06 09/19/05

Go up one level in this thread


Ali -
"This is simply wrong ;I looked at the NPS for both programs"

On rigorous scientific( == commonsense)methodology, your experiments would not
be passed by any sane person.

1) The nps of 500 by chessTiger can only be known to Christopher Theron. GNU
chess's 500k, will be known to me if written by me and compiled by me. Reading
from lcd/screens has no scientific value here.

2) most importantly, GNU 4.11 may not be fitting for such experiments. Those
   who know how a chess program works know that some programs can be just not
   fit for the experiment you have in mind. GNU Chess given 1000x may mean
   nothing. Maybe given 1 year/move on a top current PC
   vs.Crafty with 1 day/move and no pondering :-

   Experiment of the Century
   =========================
   GNU Chess 4.11 - W
   Crafty 19.00 - B;
   Result 68 years later:-
   0 - 1 Checkmate.



Best Regards
Chan Rasjid
2005




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.