Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Fritz5 smokes Hiarcs 6 engine in 30' fixed match on PII 233

Author: Enrique Irazoqui

Date: 14:16:37 10/17/97

Go up one level in this thread



On October 17, 1997 at 15:18:11, Thorsten Czub wrote:

>On October 17, 1997 at 12:54:02, Enrique Irazoqui wrote:
>
>>
>>On October 17, 1997 at 11:32:00, Thorsten Czub wrote:
>>
>>>It is designed to do so !
>>>Instead of Blitzgames I would try games with time per move >60 seconds !
>>>Hiarcs is and was a tournament program. It is designed to play 40 in
>>>120' games.
>>
>>I agree that Hiarcs is tuned for slow time controls. Still, at 1 minute
>>per move on a P200MMX/64, Fritz 5 beats Hiarcs 6 by 67 to 33 in the
>>games I played. It beats H6 by more at faster games, and I assume H6
>>would win at 40 in 2, but not by much at all.
>>
>>Enrique
>
>My friend and I we have together 6 machines.
>We have tried it out in different time-levels from 5' Blitz to 60/60 and
>at least 40/120.

Well... I have only 2 machines, but I have over 100 games played between
Fritz 5 and Hiarcs 6. Plus many others played by F5 against several
opponents.

>With enough time Hiarcs overtakes Fritz.
>Or overtakes Genius.

I think so too, as I said before. H6 should win, like Rebel 9, Not by
much, though.

>This was the same 1993 and the problem has not changed much.

Here I disagree. The problem is very different now. Fritz 3 and Fritz 5
are different animals. F5 knows much  more.

>All you measure is - what you said yourself - that Hiarcs is not coming
>deep enough. Mark has worked a lot on making Hiarcs coming deeper.
>From my point of view he has done a good job in giving Hiarcs some extra
>plies.
>ALL knowledged based chess programs have problem beating fast-search
>chess programs , but I have never seen a fast-search program leading the
>SSDF-list !

You are still talking of Fritz 3, not about Fritz 5.

>In the end they have not enough knowledge to survive. But maybe this
>championship will change it ??
>We will see !!

We'll see nothing really. 11 rounds won't prove much one way or another.
500 games per program played by SSDF mean much more. In there F5 will
not be so far from top. 50 points at the most if it runs with enough
RAM.

Enrique



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.