Computer Chess Club Archives


Search

Terms

Messages

Subject: Pattern Recognition on Fritz 6; It may be helpful for Kasparov or Fritz

Author: Jonathan Lee

Date: 17:35:12 11/09/03


  Since the Kasparov match is near, I thought you want
to know some pattern recognition with Fritz losses.
   First of all, go to www.icdchess.com with the
Computer Chess Club search engine on "pattern
recognition" with Jonathan.  Most of what I wrote were
no brainers: (Unique) closed openings, complex middle
game & lose something, human wins as white 2/3 of the
time, and the original queen is used.

  The machines that are 1 GHZ or greater for Fritz
that lost had these things in common ALSO:
   1. Both machine and human castled kingside.
   2. Kingside pawn structure for the human:
   2.a Several grandmasters had usually white pawns on
h3, g2, & f4 and the king on g1.
   2.b Leko had white pawns on h4, g3, & f5 with the
king on h3 later.
   2.c Dreev was more creative: black pawns on h7, g5,
& f4 with a black knight on g6 and king on h6
eventually.
   3. Fritz piece or pieces were too much on the
queenside when it lost.

Jonathan
PS I know it was Fritz 6 and (not Fritz 8 or 9), but it could be helpful for
human or machine.



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.