Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: If 75 Games are not considered a Statistical proof, neither is the SSDF.

Author: Bruce Moreland

Date: 22:17:14 01/30/01

Go up one level in this thread


On January 30, 2001 at 21:53:55, Dann Corbit wrote:

>On January 30, 2001 at 21:24:13, Robert Hyatt wrote:
>[snip]
>>I think that 1000 game match might well be way off from 50%.  IE it is not
>>unlikely that one side will pull ahead a significant amount, and then they
>>start playing equally for the rest of the match.  But there is nothing in
>>statistics that says after you flip a coin and get 100 consecutive heads,
>>that sometime later you will get 100 consecutive tails to offset them.  It
>>is more likely that this series will simply end with heads being ahead 100
>>counts...
>
>You can have a bad run anywhere.  For instance, for two perfectly even engines,
>one can win ten in a row [including against itself, for example].  The
>probability of this happening right off the bat is identical to it happening one
>million games into a very long trial.  Over a very long run, the bad runs will
>cancel each other out (on average).  And if you have one after one million
>games, 10/1000000 is a very small percentage so it won't hurt much.

All that happens if you have a run of good or bad luck is that you shift your
location within the normal distribution to the left or to the right a little
bit, which is kind of the point of the normal distribution.  If you do 1000
tests you probably do not end up with +500 -500, but all the silliness
*approximately* cancels out, and most of the time you end up pretty close to the
mid point, and almost all the time you end up within a reasonable distance from
it.

bruce



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.