Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: What went wrong with P.Conners and Zugzwang in WCCC?????

Author: Dann Corbit

Date: 15:02:04 06/29/99

Go up one level in this thread


On June 29, 1999 at 01:28:28, Jouni Uski wrote:

>These superfast programs only got 3 resp 3.5 points! Specially P.Conners was
>one of the favourites before WCCC. I hope someone can give explanations.

Let's suppose that one of the programs has an ELO 500 points above all the other
programs.  Did you know that the odds of winning all seven games against those
opponents would be only 68%?

Since the win expectancy for a 500 point difference is 0.0532402, the superior
opponent has 0.9467598 as their expectancy. The odds of an output of 1 seven
times in a row is therefore ( 0.9467598 )^7 =  0.6818340171542.

In short, from a tiny match like this, where the real strength of each and every
opponent is a true unknown, we really have no idea whatsoever which program is
best.  We do have a champion, and it also seems probable that the top ten
finishers are among the best programs.

P.Conners and Zugzwang may be the among the strongest programs in the world or
even the strongest for all we know.  Or it could be Mini, for that matter.

People seem to think that a simple contest like this shows us which is the best
program.  It most emphatically does not accomplish that goal.  What is achieved
is a championship, which is something altogether different.



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.