Computer Chess Club Archives

Search

Terms

Messages

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

Author: Jonathan Lee

Date: 13:52:20 05/13/02

This seems to be a glaring statistic in pattern recognition on comp vs GM.
Softwares include Fritz, Junior, Chess Tiger, Shredder, Deeper Blue and Conners.

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.

Interesting to note here is there's a book called "How To Beat Your Dad in
Chess" by British GM Murray Chandler where it has about 47 deadly checkmates and
3 draws.  More than 80 percent (40 out of 50) involve castling kingside.
However, the softwares today are more sophisticated to understand these
positions.  The book will probably not help, but it's worth a try.
Jonathan (151 message)
PS When is Kramnik vs Fritz (if ever)?

Re: 90% of (comp losses vs GM) are comp castle kingside (O-O) Gareth McCaughan 00:44:51 05/14/02
Re: 90% of (comp losses vs GM) are comp castle kingside (O-O) Keith Ian Price 17:24:56 05/13/02
Re: 90% of (comp losses vs GM) are comp castle kingside (O-O) Robert Hyatt 15:10:36 05/13/02
Re: 90% of (comp losses vs GM) are comp castle kingside (O-O) Rafael Andrist 13:58:59 05/13/02
- Re: 90% of (comp losses vs GM) are comp castle kingside (O-O) Jonathan Lee 17:45:32 05/13/02
  - Re: 90% of (comp losses vs GM) are comp castle kingside (O-O) Rafael Andrist 01:39:56 05/14/02
  - Re: 90% of (comp losses vs GM) are comp castle kingside (O-O) Keith Ian Price 18:35:49 05/13/02
    - Re: 90% of (comp losses vs GM) are comp castle kingside (O-O) Jorge 07:30:38 05/14/02
- Re: 90% of (comp losses vs GM) are comp castle kingside (O-O) Miguel A. Ballicora 14:17:40 05/13/02

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.