Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: If Hiarcs 8 is improved by 100 rating points...

Author: Uri Blass

Date: 00:52:48 04/23/02

Go up one level in this thread


On April 23, 2002 at 03:45:58, Günther Simon wrote:

>On April 23, 2002 at 03:33:32, Uri Blass wrote:
>
>>On April 23, 2002 at 03:24:08, Jouni Uski wrote:
>>
>>>On April 23, 2002 at 03:12:31, Uri Blass wrote:
>>>
>>>>On April 23, 2002 at 01:35:29, Jouni Uski wrote:
>>>>
>>>>>...it's still behind Fritz/Tiger in SSDF!? So don't expect too much.
>>>>>
>>>>>Jouni
>>>>
>>>>And if it is improved by 200 rating points?
>>>>
>>>>Assuming as obvious that it does not improve by more than 100 points does not
>>>>make sense.
>>>>
>>>>Fritz improved by more than 100 elo points from the time that hiarcs7.32 came to
>>>>the market.
>>>>
>>>>I see no reason why hiarcs cannot do the same.
>>>>The fact that Hiarcs7.5 and Hiarcs7.8 were not released in the time of Fritz6
>>>>and Deep Fritz does not mean that these programs do not exist as private
>>>>versions of the programmer.
>>>>
>>>>Uri
>>>
>>>From SSDF:
>>>
>>>     THE SSDF RATING LIST 2002-03-27   85022 games
>>>                                          Rating
>>>                                          ------
>>>  1 Fritz 7.0 256MB Athlon 1200 MHz         2748
>>>  3 Deep Fritz 256MB Athlon 1200 MHz        2709
>>>  9 Deep Fritz  128MB K6-2 450 MHz          2654
>>> 15 Fritz 6.0  128MB K6-2 450 MHz           2623
>>> 23 Fritz 5.32  128MB K6-2 450 MHz          2554
>>> 38 Fritz 5.32  64MB P200 MMX               2481
>>> 44 Fritz 5.0 PB29%  67MB P200 MMX          2460
>>>
>>>Is there over 100 poits somewere?
>>>
>>>Jouni
>>
>>Yes
>>Fritz7 is 39 elo better than deep fritz.
>>Deep fritz is 100 elo better than Fritz5.32
>>
>>It means that Fritz7 is 139 elo better than Fritz5.32
>>
>>The difference between the release time of Hiarcs8 and the previous version is
>>similiar to the difference between the release time of Fritz7 and Fritz5.32
>>
>>I see no reason to assume that similiar impeovement is impossible.
>>We need to wait and see.
>>
>>Uri
>
>
>Sorry, but I have to disagree here as it seems you forgot to look at
>the hardware specs also. Only DF and F7 used the same speedy hardware
>and all other versions must be extrapolated to their processor speed
>which diminishes the differences between all versions a real lot.
>
>Regards,
>Günther

I looked at the hardware and this is the reason that I did not compare dierectly
between Fritz7 and Fritz5.32

Both fritz7 and Deep Fritz used the same hardware in my comparison(A1200) and
I got 2748-2709=39.
Both Deep Fritz and Fritz5.32 used the same hardware in my comparison(K6-450)
and I got 2654-2554=100 elo improvement.

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.