Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Blitz games aren't very significative...

Author: Uri Blass

Date: 05:37:57 11/08/01

Go up one level in this thread


On November 07, 2001 at 08:23:38, Terry McCracken wrote:

>On November 07, 2001 at 06:04:07, Jouni Uski wrote:
>
>>On November 07, 2001 at 05:51:44, Uri Blass wrote:
>>
>>>On November 07, 2001 at 05:14:29, Jouni Uski wrote:
>>>
>>>>On November 07, 2001 at 05:07:26, Otello Gnaramori wrote:
>>>>
>>>>>Hi Eduard,
>>>>>Could you please try with regular timings games (40/2h) , I think should be
>>>>>nicer for Fritz especially if running on slow machines that requires more
>>>>>pondering time.
>>>>>
>>>>
>>>>With longer games programs are worse actually! But of course it takes Eduard
>>>>some weeks/months to get one win against Fritz in 40/2h...
>>>>
>>>>Jouni
>>>
>>>1)With longer games program are better and not worse
>>>
>>
>>I must be more precise: with longer games programs are RELATIVELY weaker
>>than humans. Example Fritz got >2800 rating in blitz in Frankfurt, but only
>>about 2500 in long games in Holland and so on...
>>
>>Jouni
>
>I concur with that assessment Jouni, but the programme should have solved that
>combination, at longer T/C regardless.
>
>Actually I looked at it half asleep...sort of, it's still early and I'm still
>tired and I saw the mistake rather quickly!
>
>I'm surprised even on a so-called _slow_ machine even at G/5 or whatever missed
>the tactics of Qxh3??
>
>If I can see the danger quickly in a stupor, then Fritz 7 should under it's
>*adverse* situation should have also. Well, IMHO.
>
>Terry

I believe that the problem is the evaluation function
Fritz7 probably did not evaluated correctly the unbalanced material structure.

Deep Fritz's evaluation is better in this case.

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.