Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Deep blue chip-more than 70% of the voters think that it is not strong

Author: Robert Hyatt

Date: 14:26:34 06/01/99

Go up one level in this thread


On June 01, 1999 at 09:52:30, blass uri wrote:

>
>On June 01, 1999 at 08:51:05, Bruce Cleaver wrote:
>
>>I think most readers do not believe Hsu, because they forget that not only is
>>the hardware greatly stronger than a micro, but the eval function is also
>>greatly stronger than a micro (as Dr. Hyatt has pointed out) because Hsu can do
>>all sorts of cool eval things for free in hardware.
>
>I do not think that the readers forget it.
>I think that they simply do not believe that the eval function of DB is
>stronger.
>
>I think that the reason is that they were not impressed by the public games of
>DB or DB Jr.

If they don't believe it, where is the micro that has beaten Kasparov in a
6 game match?  Where is the micro that has beaten _any_ GM in a 6 game match
at 40/2hr?  Doesn't really matter what 'they' believe... it only matters how
'it' plays.  And there is a bit of evidence showing it is strong as all hell...



>
>  Also, as Larry Kaufman has
>>pointed out, comp-comp results tend to be exaggerated compared to comp-human
>>results.
>>
>>
>>  This brings up an interesting point - is it possible to fairly test a DB Jr.
>>with a micro?  Setting both to a figure of  X kn/sec doesn't cut it, because the
>>hardware eval function of DB Jr. is MUCH better than the micro's eval function -
>>even a highly developed one like Fritz or whatever....Any opinions??
>
>a highly developed one like Fritz???
>
>If you do the test of setting commercials to X kn/sec then I do not expect Fritz
>to have very good results.
>
>I think that Hiarcs7 or CM6000 are clearly better in this test.
>
>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.