Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Waltzing Matilda (was: statistics, 10 events tell us what ?

Author: fca

Date: 13:04:44 08/16/98

Go up one level in this thread


On August 16, 1998 at 15:05:24, Bruce Moreland wrote:

>On August 16, 1998 at 14:21:20, fca wrote:

>>O'Donnell(IM) v MCP7   0.5 - 9.5
>>O'Donnell(IM) v CG5    0.5 - 9.5

>I have seen programs beat grandmasters as many as 16 times in a row at 5 0.

Surely so (some of such games may be "punch drunk" syndrome stuff though, after
the first few losses - computers never see red, never get emmotional or
disheartened etc.).

But *not* vs a 166MHz Pentium and 10 0 and in a prepared match, I suggest. ;-)

Of course, the purpose of my quoting the results was to highlight that a simple
end-result does not tell the whole story.  The matches v CG5 and v MCP7 were
qualitatively very different.

Kind regards

fca

PS: I've just spotted your questions to me in the earlier post in the thread,
and have replied to them.  In that vein, may I mischievously point out that
there is a difference between a "match" and a "sequence".  Let a reasonable
program play against any GM for long enough, and somewhere in the sequence you
will find 16 consecutive wins for the program.  In the absence of evidence to
the contrary, such sequences are "normal" (this has nothing to do with normal
distributions, this is like Liouville's number etc.)... pi is normal, which is
why *I know* that somewhere in its decimal expansion are 58,124,760 consecutive
"zero" digits, exactly.  In fact, not just in one place, but in infinitely many
such places.



This page took 0.01 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.