Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Hiarcs modified vs Rybka after 193 games at 30min each

Author: George Speight

Date: 20:31:08 01/18/06

Go up one level in this thread


On January 18, 2006 at 16:15:35, robert flesher wrote:

>TINCAN, Blitz:30'  0
>
>
>1   Hiarcs 10                 +64/-51/=78 53.37  103.0/193
>2   Rybka 1.01 Beta 7 32-bit  +51/-64/=78 46.63   90.0/193


I personally think you did a very good job. Not a thing wrong with your results.
One bit of advice, though. From now on test full programs and not strictly
engines. Then you will be letting them use everything they have at their
disposal. Which is a fair deal in full program testing. I ran best of 30 with
book learning on and Rybka beat Hiarcs 10. I also ran a best of 30 with book
learning on and Fritz 9 beat Rybka..  Vas will indeed make improvements and
additions to make Rybka stronger, but in full program you never penalize one
program for what the other doesnt have. That will take care of all the
arguments. Best to you, George



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.