Author: Kolss
Date: 07:36:02 04/16/05
Go up one level in this thread
On April 16, 2005 at 09:32:14, Eelco de Groot wrote: > Hi Anson > >For a first order approximation, I always use 2812 - (50%-45%)*7 =2777, which is >increasingly inaccurate with greater score-difference. Hi, Yes, that is a very good approximation, 55 % corresponds to +35 Elo, 60 % to +70; starting around 65 %, things get off a bit (+108 Elo, still quite good), and above that, it becomes indeed increasingly inaccurate. Essentially, your approximation works perfectly for the most interesting range - after that, it becomes more difficult to estimate programs' or players' ratings accurately anyway... Best regards - Munjong.
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.