Author: Miguel A. Ballicora
Date: 11:04:24 02/28/02
Go up one level in this thread
On February 28, 2002 at 11:38:16, José Antônio Fabiano Mendes wrote: >On February 27, 2002 at 23:35:42, Robert Hyatt wrote: > >>On February 27, 2002 at 17:03:08, Chris Carson wrote: >> >>>On February 27, 2002 at 16:00:18, Robert Hyatt wrote: >>> >>>>On February 27, 2002 at 15:19:12, Chris Carson wrote: >>>> >>>>>On February 27, 2002 at 12:34:58, ALI MIRAFZALI wrote: >>>>> >>>>>>In the 1999 article in the IEEE journal;it was reported by Hsu that a single >>>>>>chip version of Deep Blue won a 10 game match against the top 2 commercial >>>>>>programs (of 1997) 10-0 ;and that Deep Blue Jr had a plus 2700 elo performance >>>>>>against the GMs working on the project .Have these games ever been Published? >>>>>>What were these 2 commercial programs? >>>>> >>>>>No the games were never published. The name/version of the programs were never >>>>>published. The HW the programs were running on was never published. >>>>> >>>>>There has been a lot of talk on in this forum about this match and the results, >>>>>but it is all speculation since the above facts have never been published. >>>> >>>> >>>>actually the _names_ of the programs were published. And the hardware used >>>>was published. The games were not published for the games vs the micro >>>>programs. The games against the human grandmasters are certainly public >>>>record as they were played at various public events and tournaments around the >>>>world... I don't personally have them, but I have seen the pgn versions around >>>>somewhere. >>>> >>>>If you look back thru the CCC archives you will find information about the >>>>DB(one chip) vs the micros. More than one person asked them about it at the >>>>various public presentations they made after the kasparov match, and then they >>>>reported a summary here... >>> >>>I do not want to get into a debate on this topic. If you will tell me where >>>they (DB team) published, I would like to read the article. >>> >>>I am talking about information published by HSU or another member of the DB >>>team, I do not consider a discussion in this forum by people (not on the DB >>>team) to be valid. If HSU or a member of the DB team discussed it here, then >>>point me to that post, but it must be HSU or a DB team member. >>> >>>If you can point me to a published reference by HSU or a team member, that >>>covers this information, that would be great! >> >> >>Hsu hasn't posted here. However, several (including Bruce I believe) have >>attended talks he or Murray have given, and _they_ have reported on those >>talks here. >> > A summary of one such talk by Hsu can be found here: > http://www.stanford.edu/class/ee380/9798sum/lect03.html Thanks for the link! I quote "... Finally, in May 1997, 48 years after Shannon's lecture, Kasparov, the human World Champion, lost the rematch to Deep Blue and thus ended the quest for one of the oldest holy grails in computer science" In other words, the quest was waiting the human to screw up badly in a short match and declare the issue solved. This is more like a talk from a marketing person or a sport coach. Please, lets not call this science. It is embarrassing. Everybody is talking about statistical significance but when it comes to declare the issued solved, 6 games is enough. Not to mention that this match was decided by two stupid mistakes from Kasparov. With this criterium, cancer could have been declared cured for ever 3 decades ago. I have a very high respect for the DB team, but this is not science. At least, it is not basic science. Regards, Miguel > >>If you choose not to believe what they posted, that's a decision you have to >>make, of course...
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.