Computer Chess Club Archives


Search

Terms

Messages

Subject: Bilbao and statistically meaningful results

Author: Roger D Davis

Date: 16:33:22 10/09/04


Are the Bilbao results likely to be statistically meaningful, or is it just
another case of too few games being played to make a statistically meaningful
result?

Shouldn't the number of games played in a tournament be determined by the
anticipated effect size? In other words, if we believe the parties are nearly
equal, with one side winning only 55% of the time, then shouldn't that determine
the number of rounds? Fewer games and the outcome is statistically likely to be
a fluke. We've seen it happen time and time again in comp-comp matches...one
side pulls ahead in the first 50 games, only to lose that advantage in the
second set of 50 games.

And who wants bragging rights to a fluke, even if they did win.

Can anyone say whether the results are likely to be meaningful from a
statistical point of view given the number of rounds in Bilbao? I don't want to
fall into the trap of believing that Bilbao is more meaningful than it really
is, or even meaningful at all.

Roger









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.