Author: Christophe Theron
Date: 08:41:59 03/14/98
Go up one level in this thread
On March 14, 1998 at 01:10:56, vincent dichiacchio wrote:
>an lct test of OWL chess gave it a 2045 elo on a p100. during the test,
>i noticed that the times occasionally varied for no apparent reason. in
>particular, positional test 12 could range from 8 seconds to >600. that
>was the case even if i booted up the game just before running that
>positional test. can anyone explain why that would happen?
>incidentally, the 2045 was with the 8 seconds.
If you are talking about the "OWL Chess" that is provided with the
Borland C compiler, what I can say is that it uses a small random number
(between 0 and 3) in its evaluation.
This can explain the inconsistencies you noticed.
BTW, this program seems to be a translation in C of an old pascal
program from Borland, "Turbo Chess", that was sold in 1985 in a "Turbo
Gameworks" package.
Does anyone know who is the author of this program? The way the code is
written shows that it has been done by an experienced chess programmer,
especially if you consider it was released in 1985. It uses 0x88 data
structures, fast move generation and ordering, a little bit of
selection, static piece-square tables, simple pawn structure eval, check
extensions and push-passed-7 extensions.
I suspect this program could have been written by the Spracklens, but I
think Sargon didn't use extensions in 1985.
Could it be the ancestor of Fritz?
Maybe the author of this nice program reads CCC?
Christophe
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.