Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Question regarding GS 2930 test suite position #13

Author: Robert Hyatt

Date: 08:09:49 05/28/01

Go up one level in this thread


On May 28, 2001 at 02:25:22, Uri Blass wrote:

>On May 27, 2001 at 22:35:31, Robert Hyatt wrote:
>
>>On May 27, 2001 at 09:40:32, Bruce Moreland wrote:
>>
>>>On May 26, 2001 at 21:11:48, Robert Hyatt wrote:
>>>
>>>>
>>>>More data.
>>>>
>>>>summary:
>>>>
>>>>After 2 hours per position, crafty gets 8 of 14 right.
>>>>
>>>>after 6 hours per position it gets 9 of 14 right.
>>>>
>>>>that is still 2 behind DB Jr.  and when you factor in the fact that DB2 is
>>>>32x faster than DB Jr, DB2 would get 11 of 14 right in < 1 minute per position,
>>>>which is 1/360 the time I need to get 9 of 14 right.
>>>>
>>>>I think the machine is amazing, _still_.
>>>>
>>>>This seems to be ample evidence of it.
>>>
>>>The shadows on the cave wall have always looked pretty interesting.
>>>
>>>bruce
>>
>>
>>I don't follow.  This thread was _started_ by a DB team member asking someone
>>here about position 13.  That doesn't seem like "shadows on the cave wall" to
>>me...  seems like very recent activity being reported directly by the source.
>
>I think that bruce means that games are the real things when test suites are
>only the shadows.
>
>The interesting acticity is games of Deep Blue Junior against the top programs
>of today and unfortunately we can get only the shadows.
>
>Uri


With a bright enough light, and enough study, even the shadows reveal a lot of
facts...

IE if I know the distance from you to the wall, and from you to the light, I
can exactly determine your height.  And if you turn sideways for a sec, I can
make a pretty good cut at determining your weight.

That is better than "unknown"...

I'm interested in anything I can find out about DB/DBjr...



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.