Author: Rolf Tueschen
Date: 16:14:24 07/18/04
Go up one level in this thread
On July 18, 2004 at 18:57:55, George Tsavdaris wrote: >On July 18, 2004 at 15:35:07, Rolf Tueschen wrote: > >>On July 18, 2004 at 14:40:01, Rémi Coulom wrote: >> >>>Hi, >>> >>>I compiled a new version of the tool that I distributed yesterday. This one can >>>deal with larger numbers of games. I tested that it works for a 6000-game match, >>>so that should be enough for everyone. You can get it at this address: >>>http://remi.coulom.free.fr/WhoIsBest.zip >>> >>>The only difference with yesterday's program is that it is compiled with gcc >>>instead of msvc. the "long double" type is 80 bit in gcc and 64 in msvc. The >>>additional accuracy is enough to solve underflow problems. >>> >>>Concerning Joachim's problem, here is the output of the program: >>> >>> This program evaluates the likelihood that program A is better than >>>program B, based on the result of two matches played against the same >>>opponent (or set of opponents). The number of games played in each of >>>these matches does not have to be the same. If playing against a set >>>of opponents, the proportion of each opponent should be the same in each >>>match. >>> The likelihood is estimated by Bayesian inference, assuming an uniform >>>prior distribution of the probabilites of losing and winning. >>> The resulting integral is estimated with a Monte-Carlo method. It may >>>take a long time to converge when the number of games is large (>100). >>>The computation can be interrupted at any time with Ctrl-C. >>> >>>A wins = 197 >>>A losses = 108 >>>A draws = 95 >>> >>>B wins = 189 >>>B losses = 130 >>>B draws = 81 >>> >>>P(A>B) = 0.890434 (600000000 iterations) >>> >>>Rémi >> >> >>Rémi, >>excuse-moi stp, >>if you look at the two programs and their numbers, you can see without too much >>calculation that the p of A>B can never me .89... No way! To say that I don't >>need formulas. These numbers are much too small. > >Why not? Give some reasons..... Why 89% is too small? I can't understand. Not that value is small but that is way too big. I was talking about the other numbers, these numbers of losses and wins... > > >>I guess that you then let a >>routine program run through thousands and more iterations. Yes, but that is NOT >>chess, that is mathematics. And as you know mathematics never can construct or >>allow statements about reality of chess or any other field. > >What? Mathematics can't allow statements about reality on Chess or any other >field??? Of course it can. > > I can see the predicted iridium flares from my home with an accuracy of 1-2 >seconds, i can see a chess-engine announcing a VALID mate in 13, we can predict >the next eclipse of sun with amazing accuracy, thousants of people can be saved >by a prediction of a hurricane and BILLION things more depend on Mathematics. > Mathematics are everywhere in our world and due to them we can construct >statements that are VALID(verify experiments and observations), about >everything, if we use them properly. Yes, sure. But if you have say 300 games or 500, and with such a quantity you can't establish a significant probability for such statements about strength, you can't simply run a mathematical routine (formula) over say 120000 or some million "trials" so that you then get allegedly significant probabilities. You can't do it this way. But run so many games! Good luck. > >>Rolf
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.