Author: Alessandro Morales
Date: 08:26:33 10/21/98
Go up one level in this thread
On October 20, 1998 at 18:48:08, john c cook wrote:
>how come he did not play fritz5 or the best program on the plane rebel10
>on a p2 450 or a amd 450 and talk to me thin i bet it would not have that
>many game it look to me that they set this up for best way for the human to get
>the best out of is game i have play a lot of master on my system and they did
>not do all that good but i have a pii 450 amd 250 meg memory
>play my computer and i bet you will not have all the win it got
HI John, Odell and Blass Uri!
The match was played one year ago and the goal was to verify the performance
of an OLD P133 (4 times slower than a PII 450) against a strong master.
We will soon begin a new match with the SAME PC, testing all the new programs,
but we expect the same results due to the fact that Marco Corvi has a better
"Learning function" than all the programs (and his rating is now 2320).
Then we will play a new match against a PII 450 Mhz 128 RAM, so we will have
more data to analyse..
Stay tuned on www.aladin.com/chess/news.html to see what will happen.. :-)
PS: Dr. Bob Hyatt will be probably right until the next generation of
Pentium / AMD running at 700+ Mhz.
Alessandro Morales (Aladin Chess Pages)
>
>On October 20, 1998 at 18:06:34, blass uri wrote:
>
>>
>>On October 20, 1998 at 17:09:31, odell hall wrote:
>>
>>>Hi All
>>>
>>>
>>> Yesterday I discovered a interesting computer chess site (aladin news) that
>>>hosted a match between fide master M.Corvi (elo 2306) and various strong
>>>programs including M-Chess 6, Rebel 8, Genius 5, Hiarcs 6 and Nimzo 3.5. The
>>>match was played at 40/2hrs and the machine that was used was a pentium 133 with
>>>16 megs of ram. The results so far is 6.5 to 4.5 in Favor of the computers. Does
>>>anyone know what performance rating this would give the computers?
>>
>>I think something near 2380.
>>
>>
>>> To be frank I
>>>was quite surprised at the outcome I assumed the programs would win nearly every
>>>game. In spite of the fact that they were running on a pent 133, according to
>>>ssdf they would be in the 2450-2490 range.
>>
>>The ssdf list is based only on game between computers programs and not on games
>>between programs and humans.
>>
>>The result is not surprising becuase Junior5 on better hardware has only 2500
>>performance.
>>
>>
>> When i see this kind of result it
>>>makes me wonder if Dr. Hyatt may be correct in saying that computers are not
>>>grandmasters, or even close. Is it possible that a strong international master
>>>would score a similiar result against a 2306 master?
>>
>>It is not impossible
>>7.5:3.5 is the expected result for 2450 player
>>and the difference is only 1 game.
>>
>>I think it is more difficult to prepare against humans than against computers
>>and I believe that a strong international master that learned the programs
>>weakneses can beat them.
>>
>>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.