Author: Bruce Moreland
Date: 11:02:48 06/16/01
Go up one level in this thread
On June 16, 2001 at 05:19:11, Alberto Rezza wrote: >Obviously any comparison between the Elo ratings of, for instance, Capablanca >and Kasparov, obtained in this way, won't be very reliable. But the same is true >when comparing games from the same time: Kasparov and Mr. Nobody (1700 Elo), who >only ever played in his hometown, do *not* play in the same pool of players.. :) This is a real problem with people who don't play in the same pool. There is, for instance, the case that was a subject of wide discussion several years ago. There was a prison inmate with a USCF rating, he played a bunch of other prisoners with USCF ratings, and he ended up with one of the highest ratings in the United States. There was a lot of debate about how good he was, but it was hard to be sure since he was in prison, so he couldn't play anyone from the outside. That was very possibly a case of rating creep due to repeated play in a very small pool. I've often heard that ratings in different parts of the US are skewed, because the same players play each other, and end up all getting over- or under-rated somehow. bruce
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.