Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Question for Theron or Utzinger

Author: Christophe Theron

Date: 14:50:51 03/30/04

Go up one level in this thread


On March 30, 2004 at 13:28:36, Pierre Chevalier wrote:

>An important subject on this site has been the number of games necessary
>for a result to be statistically meaningful.  My question is whether
>Kasparov and Karpov have played enough games against one another or if
>Kasparov's margin of victory has been large enough to be statistically
>meaningful?  Most chessplayers only play one another maybe ten times
>but K and K played each other in 144 championship level games.  Anyone
>can answer but Theron and Utzinger I know have expertise on this matter.



Here are the tables that I use. Grab a calculator and draw your own conclusion.

(sorry about the bad formatting, the tables have been formatted to look good on
proportional fonts on my Palm)


Reliability of chess matches
(assuming each opponent has 1/3 chances to win, 1/3 to loose and 1/3 to draw)

90% confidence
Games    %err+/-    elo+/-
     10    20        140pts
     20    15        105pts
     25    14          98pts
     30    12          63pts
     40    10          70pts
     50      9          56pts
   100      6.5          35pts
   200      4.72      33pts
   400      3.34        23pts
   600      2.66      19pts
   800      2.39      17pts
1000      2.12      15pts
1200      2.00      14pts
1400      1.81      13pts
1600      1.66      12pts
2000   ~1.50      11pts

80% confidence
Games    %err+/-    elo+/-
     10    15        105pts
     20    11          77pts
     25    10          70pts
     30      9          63pts
     40      8          56pts
     50      7          49pts
   100      5.0          35pts
   200      3.75      26pts
   400      2.60      18pts
   600      2.15      15pts
   800      1.86      13pts
1000      1.66      12pts
1200      1.46      10pts
1400      1.40      10pts
1600      1.34        9pts

70% confidence
Games    %err+/-    elo+/-
     10    15        105pts
     20    10          70pts
     25      8          56pts
     30      8          56pts
     40      6.3          44pts
     50      6.0          42pts
   100      4.0          28pts
   200      3.0          21pts
   400      2.2          15pts
   600      1.7          12pts
   800      1.5          11pts
1000      1.3            9pts
1200      1.24        9pts
1400      1.14        8pts
1600      1.04        7pts

400 games:
err=+/-2%    conf=68%    elo=+/-14
err=+/-1%    conf=38%    elo=+/-7
err=+/-0.5%    conf=22%    elo=+/-4

800 games:
err=+/-2%    conf=84%    elo=+/-14
err=+/-1.5%    conf=71%    elo=+/-11
err=+/-1%    conf=52%    elo=+/-7
err=+/-0.5%    conf=28%    elo=+/-4

General rule: when the number of games is multiplied by 4, the error margin is
divided by 2.

Official elo formula:
  elodiff=-400*log10(1/wp-1)
(wp is the winning percentage)



    Christophe



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.