Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: 90% of (comp losses vs GM) are comp castle kingside (O-O)

Author: Gareth McCaughan

Date: 00:44:51 05/14/02

Go up one level in this thread


On May 13, 2002 at 16:52:20, Jonathan Lee wrote:

> Only 2 games (that I know of) has castled queenside and the comp. lost;
> they are (Piket vs. Junior 5.6 GHZ) & (Wiel vs. Fritz 300 MHZ).
>
> The rest computer has castled kingside (almost 91 percent).
>
> 73 percent of the lost games are both human & comp. castled kingside.

And, in other news, scientists announced that drinking milk is
hazardous to the health. "We found that over 99% of people who
died before the age of 35 had drunk milk at some time in their
lives", said Dr Dodge E. Statistic.

Perhaps 90% of *all* computer-v-grandmaster games feature
the computer castling kingside. If so, then it would be
no surprise that 90% of *lost* computer-v-grandmaster games
do so too.

It would be more interesting to know what proportion of
*all* computer-v-grandmaster games in which the computer
castles kingside are lost, and the same for games where
the computer castles queenside or not at all.

Note that even if it turns out that computers do lose more
games when they castle kingside, making them do so less isn't
necessarily going to help. The alternative to castling
kingside in any given position is likely to be not castling
at all...

--
g



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.