Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site fisher.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!allegra!princeton!astrovax!fisher!glickman
From: glickman@fisher.UUCP (Mark E. Glickman)
Newsgroups: net.games.chess
Subject: Re: Chess Ratings
Message-ID: <1354@fisher.UUCP>
Date: Wed, 11-Dec-85 15:02:18 EST
Article-I.D.: fisher.1354
Posted: Wed Dec 11 15:02:18 1985
Date-Received: Thu, 12-Dec-85 06:33:00 EST
References: <23@drutx.UUCP>
Organization: Princeton University.Mathematics
Lines: 12

> It is based on statistical probability theory, with
> ratings following a Normal distribution with a
> mean of 1500 and a standard deviation of 200.
 
Well, not quite.  The rating system is very much skewed to the 
left - there is a very high concentration of lower rated players
(that are incidentally, for the most part, inactive) that 
cause the distribution to demonstrate assymetry.  A player's
performance, however, is asymptotically normally distributed.

		      - Mark Glickman
		       (member of the USCF Rating Committee)