Author: Robert Hyatt
Date: 20:34:50 02/18/03
Go up one level in this thread
On February 18, 2003 at 05:21:00, Sune Fischer wrote: >On February 18, 2003 at 04:38:32, Alastair Scott wrote: > >>On February 17, 2003 at 14:41:34, Anthony Cozzie wrote: >> >> >>>the elo system has no defined 0. results are only defined in terms of wins and >>>losses. For example, suppose one defined the average elo to be 1600, and placed >>>Kramnik, Kasparov, and Shirov in a room together and had them play 5000 games. >>>Kasparov's rating would be 1650 at best. Or we could define the 0 to be 0 - >>>Kasparov would have a rating of 1200, and some people would have negative >>>rating! The whole thing is just like potential energy in physics: only >>>differences in the rating system are meaningful. >> >>Excellent explanation, and there is also the Flynn effect (such rating systems >>tend to progressively inflate the numbers over time), which I believe has never >>been explained. > >How do you know they inflate if you can't compare them? > >-S. > One simple idea is to compare the "average" rating for the pool, over time. IE the average "IQ" is not going up, so the average rating should not go up since it is a relative measure among the pool members. If it does go up, it has to be inflation since not _everybody_ is going to improve steadily... >>Alastair
This page took 0.01 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.