Author: Rolf Tueschen
Date: 14:41:10 01/13/03
Go up one level in this thread
On January 13, 2003 at 16:43:15, Frank Phillips wrote: >On January 13, 2003 at 16:03:18, Rolf Tueschen wrote: > >>On January 13, 2003 at 15:10:30, Frank Phillips wrote: >> >>>On January 13, 2003 at 00:25:44, Robert Hyatt wrote: >>> >>>>On January 12, 2003 at 18:02:55, Frank Phillips wrote: >>>> >>>>>On January 12, 2003 at 16:42:13, Vincent Diepeveen wrote: >>>>> >>>>>>On January 11, 2003 at 09:24:34, Frank Phillips wrote: >>>>>> >>>>>>look to analysis from Seirawan. June 1997. they crack down deep blue everywhere >>>>>>from technical viewpoint. And Kasparov even more. >>>>>> >>>>> >>>>>Have you got a link. The best I found was a book was $46. >>>>> >>>>>Your contention is that these errors were so gross (even though apparantly not >>>>>punished by the best player in history, probably) that no modern micro make >>>>>them? Interesting: I am not able to judge but would be interested in seeing the >>>>>link. >>>>> >>>>>Frank >>>> >>>> >>>>When they criticize arguable the best player on the planet even _more_ than >>>>they criticize DB, I think you can make an intelligent guess at how accurate >>>>the "criticism" really is... >>> >>> >>>What I find interesting is this: >>> >>>if Deep Blue was very good and did not make losing positional errors; >>>if Deep Blue is better, or no worse, than Junior or Fritz (ie Junior has no >>>magic formula to offset Deep Blue's speed advantage); >>>and Fritz held Kramnik; and >>>Junior performs well against Kasparov; >> >>Are you stating this as facts? Fritz held Kramnik?? Of course the truth is >>different. Also I don't understand how you could take such exhibitions to answer >>such questions. This is not possible. > >October 2-22nd 2002 score 4-4. And Kramnik had the exact program in advance. >No other data. Seemed pretty serious preparation, if taken at face value. Impossible to imagine! > >> >> >> >>> >>>then perhaps computer chess is closer to besting mankind than I had previously >>>believed and we do not need a mega-monster processor to be able to do it. >> >>You seem to be unaware of the many different implications of your terminology. >>"Besting mankind". What does this mean? It was said that even the close win of >>DB2 was NOT a proof for "besting". Of course besting 99,99% of mankind. But we >>talk about the best possible human chess. And there you are again in the problem >>of exhibitions. To make it short. Human chess tradition is not based on Swiss >>System tournament chess where you have to play unknown opponents. > >Not sure I understand your point and your use of the word proof, which I did not >use My prejudice has been (and still is) that we need something with the power >and complexity of Deep Blue to be better (on average) than the best human chess >players at over the board chess at standard time controls. Recent events have >challenged that notion for me. Not justified. > >As to Kasparov playing an unknown opponent, this is a reasonable point. But I >doubt that in other matches his opponents had to promise to play as they >normally did - as in Kramnik versus Fritz for example. Layman's view of GM chess. Since it's a psychological war between almost equally strong players, the most important thing to know is the personality of the opponent. And this is the constant in one's life. Therefore - among other reasons - it's so problematic when operaters manipulate personalities of machines in very short events. Then it's a bit of gambling. No longer real chess. > Even if this was a real >advantage, I think it still demonstrates that the gap in performance between DB >and GK was not massive over those six games played. Whether that would have >been true over further games we will never know. > >> >> >>> >>>Unfortunately I find Vincent's points ambiguous in this respect, since he seems >>>to claim that Deep Blue played crap yet Kasparov played worse - I think he might >>>have even implied deliberately, but I may have misunderstood him. >>> >>>I guess we will see at the end of January....or not. >>> >>>Frank >> >>No, it could well be that Vincent is quite right and still DB2 was better than >>all of the actual commercial progs. BTW Vincent isn't alone with this opinion. >>Other GM said the same. Another aspect is the question if Kasparov was sure to >>play a third match. But I don't want to speak in Bob's voice. > >Yes it could be that they both played crap; and DB is better or worse than >commercials. The 100x speed is a real advantage. (I kept it short because my first and longer version had been destroyed when I lost contact to the internet.) Rolf Tueschen > >> >>Rolf Tueschen
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.