Author: Christoph Fieberg
Date: 12:19:15 07/03/01
Go up one level in this thread
On July 03, 2001 at 10:48:27, Peter Berger wrote: >On July 03, 2001 at 10:40:33, Christoph Fieberg wrote: >> >>After this surprising result I investigated the Kasparov-Kramnik match. What is >>the likelihood that a 2849 rated player does not win a single game against a >>2770 rated player and even is defeated 2-0 after 15 games? >> >>The outcome was shocking! The chance is only 1:1439 (likelihood 0.0695%!). > >I think for a match beween humans your model won't work at all. > >It can be argued if the results of the single games are statistically >independent in a comp-comp match ( I think they are still independent enough >that your model should work well ). > >But in a human match they definitely aren't. > >pete > I found out that for a human tournament the program works very well. But for human matches it could be argued that they require specific paramenters - also in spite of the fact the Bobby Fischer won 18 games against the world elite in a row (unforgetten his 6-0 wins against Larsen and Petrosjan). Unfortunately there not enough games to find out the specific parameters. I think psychological matters play a huge role in the human matches (fortunately computers are not effected, are they? :-)). Best regards, Christoph P.S. See the description of my program at: http://www.icdchess.com/forums/1/message.shtml?178041 >> >>This results changes my view on the outcome of the Fritz-Junior match. >> >>Nevertheless it is really astonishing what happened in these matches. >> >>Christoph
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.