Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Junior - Crafty NPS Challenge - a user experiment

Author: Russell Reagan

Date: 01:21:04 11/22/03

Go up one level in this thread


On November 22, 2003 at 03:52:25, Gian-Carlo Pascutto wrote:

>Nonsense. If he plays on till the score DIFFERENCE is 6 points,
>it's staticstically a very valid result (somewhere >90 or >95%
>certainty).
>
>This holds pretty much regardless of the actual amount of games
>that is played. Someone with more time can calculate the exact
>values but I'm pretty sure the above holds.

Hmmm. The program I have that calculates statistical reliability of results
disagrees with you (http://www.stevemaughan.com/whoisbetter.htm). According to
it...

16   - 10   -> 83%
106  - 100  -> 63%
1006 - 1000 -> 54%
5006 - 5000 -> 52%
and so on...

What you say doesn't make any sense. If we play a billion games and you're up 6
games on me, it means we're basically equal. You *might* be better with a 51%
statistical reliability. Do you really think a result of +1000006 -1000000
=997999994 means one player is better with a 90-95% statistical reliability?



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.