Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: an idea how to generate a tactical test suite

Author: Jon Dart

Date: 07:00:20 09/08/99

Go up one level in this thread


I do this pretty routinely with games from my program. I usually use
Crafty for analysis, but sometimes Fritz.

I don't usually trust analysis that uses less than 10 minutes per move
on a 400 MHz machine. It sounds like a lot, but many moves that look good
or bad at shallow depths change score later on.

One problem with Crafty is that it will single out bad moves (blunders)
but it doesn't specially mark moves that are non-obvious but good. You
can look at the log files and see if a good move (actually made in the
game) takes a long time to be selected.

Also it is pretty common to not find a critical point where one of the
players went wrong. Sometimes the losing player's position just goes
slowly worse and there never really was a critical blunder.

--Jon

On September 08, 1999 at 06:31:41, blass uri wrote:

>I suggest to analyze comp-comp games from wccc(one hour per move)(it is possible
>to do it by Fritz engines).
>
>the tactical test suite can be to avoid all the blunders that the engines found.
>
>analyzing one game is some days for one computer and we have 106 games from
>wccc(I include the last draw between shredder and ferret).
>
>We need more than a year of computer time to generate this test.
>
>Uri



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.