Author: Melvin S. Schwartz
Date: 20:56:23 06/13/99
Go up one level in this thread
On June 13, 1999 at 23:37:12, Melvin S. Schwartz wrote: > >On June 13, 1999 at 18:48:32, Tomas Casanovas Martinez wrote: > >>On June 13, 1999 at 17:48:50, Melvin S. Schwartz wrote: >> >>> >>>On June 13, 1999 at 16:23:17, Tomas Casanovas Martinez wrote: >>> >>>>I have been operating three computers simultaneusly -overnight- to test H.7.32 >>>>against the following programs: Fritz 5.32, Junior 5 and Crafty 16.6. All the >>>>games at tournement level: 40 moves 2 hours. Unfortunately my computers are >>>>different: >>> >>>What is the opening book you are using for Fritz? General ctg or F5 ctg? And >>>what is the Hash-table info for Fritz and Hiarcs? And only 5 games? >>> >>>Mel >>>> >>>>Pentium II 400 , 256 hash, 'Hiarcs 7.32 environment', against Fritz 5.32 >>>>Hiarcs 7.32 - Fritz 5.32 +2 -0 =3 3,5-1,5 >>>> >>>>Pentium 200, 64 hash, 'Junior 5 environment', against Junior 5. >>>>Hiarcs 7.32 - Junior 5 +4 -3 =1 4,5-3,5 >>>> >>>>Pentium 90, 32 hash, 'Fritz 5.32 environment', against Crafty 16.6. >>>>Hiarcs 7.32 - Crafty 16.6 +4 -3 =0 4,0-3,0 >>>> >>>>Although I know well that at least 200 games are necessary to establish tha >>>>rating of a new program, my feeling is that this extremely strong Hiarcs 7.32 >>>>will be rated around 2625, well above all the other commercial programmes. >>>> >>>>It is a bit surprising that whereas H 7.32 consistently wins against Fritz 5.32 >>>>in all the tests performed -Tania, et al.-, it has substantial problems to win >>>>against Junior 5 and Crafty 16.6, who have proven to be extremely strong rivals. >>>> >>>>Do you think that the different hardware and 'environments' have been decisive >>>>for the results obtained?. My computers are still working on... >>>> >>>>Tomas. >> >>Mel, I use Hiarcs.ctg both for Hiarcs and for Fritz. >> >Tomas, > >You should not being using Hiarcs ctg as an opening book for Fritz! Have you >read the CD insert about how the opening book for Hiarcs is unsuitable for >Fritz? You should be using the opening book designed for Fritz - F5 ctg. This is >on the Talk CD. Your testing is unfair unless you allow both sides to have their >proper opening book. How could you possibly use a Hiarcs opening book for Fritz? >That's like putting a Hyundai motor in a Cadillac. Don't Tomas, I should clarify the Hyundai - Cadillac anaogy: I didn't mean Hiarcs book is bad, just that it's bad for Fritz - and totally unfair. If you put a Hyundai motor in a Cadillac, it will not perform as if it had a Cadillac motor, right? Mel you want to test >programs on an equal basis - as equal as possible? > >Mel >>Tomas.
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.