Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: ELO & statistics question

Author: Gian-Carlo Pascutto

Date: 12:46:27 04/26/02

Go up one level in this thread


On April 26, 2002 at 15:15:56, Ratko V Tomic wrote:

>The natural error size would be +/-sqrt(N), where
>N is the number of games.

I presume you mean +- 1/sqrt(N) ?

>From that you can get the
>upper and lower values for probabilities (normalized
>scores).

Can you elborate?

--
GCP



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.