Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Has Christophe's 0.01% chance actually occurred?

Author: Dann Corbit

Date: 12:45:55 09/25/02

Go up one level in this thread


On September 25, 2002 at 15:13:12, Hartmut Woldeit wrote:

>Hallo Dann,
>
>even the author of Ruffian 1.0.0 m  a y b e  created  Crafty/Fritz-clone of or a
>clone of another wellknown engine/program. What would it mean for the original
>author if the clone plays better than the original?

I think it can be done short term.  For instance, you could take 100 computers
and brute force test evaluation terms until you get a better set.  That modified
program will outperform the original for sure.  But, over time, the original
program will surpass the clone.  That is because the clone tweakers don't really
understand the plumbing.  They just turned a few handles on the pipes.  But when
the plumbing gets reworked, they will be lost.



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.