Author: Dieter Buerssner
Date: 14:49:20 05/09/05
Go up one level in this thread
On May 09, 2005 at 02:18:32, Jarkko Pesonen wrote: >What is the confidence level that these +6 -6 are true? 99% or higher? AFAIK, Elostat reports 95% confidence level. As you can see, the last 1000 games, got a result outsidwe of the before estimated confidence level. This should happen once in 20 "experiments". So it is not far off, here. When Pavel's experiment has finished, I will try to do some statistical tests on the results. Perhaps by well known statistal methods, perhaps also by some "Monte Carlo simulations". Probably there was never a test with so many games between 2 opponents before. It might be worthwhile to use all those stastical methods. Dieter
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.