Author: Robert Hyatt
Date: 13:28:16 04/18/05
Go up one level in this thread
On April 18, 2005 at 14:04:03, Rolf Tueschen wrote: >On April 18, 2005 at 12:13:42, Robert Hyatt wrote: > >>I don't know why we keep re-hashing this. > >Because Hsu six years ago and now Mike could hav implied that it were Kasparov >who lost a good possibility to make money. I tried to show that it was Hsu who >lost his future in 1997. > This is wrong. Kasparov made the claim about cheating. From that instant, the probability of any further matches with him dropped to exactly 0.00, for reasons others have already supplied. Once the claim was made, that ended IBM's interest in having anything to do with him, or probably with computer chess, period. > >>But two key points. >> >>1. The cheating claim is simply false. Why? Because _other_ programs have >>reproduced _every_ single move that Kasparov questioned. And when I say _every_ >>I do mean every, from avoiding Qxb6 and playing the Be4 move (shredder 8/9 will >>play this after a couple of minutes thinking or less) to other moves. > > >I know this, but since 1997 I pointed at the lack of validity if other progs >reproduced the moves AFTER the gamescores had been made public. You know better >than me, how easy things could be implemented. See the tricks with test >positions. Also - the cheat wasn't necessarily the play of moves no computer >couldn't have found but a) the circumstances how the crucial moves were played, >b) how the team reacted on Kasparov and c) the deconstruction of the machine >before anything could have ben checked (I don't claim that something could have >ben proven this way). Why would someone want to make their program play a move just because deep blue did it? In fact, right after the match, the move by DB (h5) was reproduced by Junior, as another example. This was produced by a commercial version of Junior sold prior to the match, so there was no way this could happen. Crafty found the Be4 move after a very long time (on far slower hardware at that time than today's boxes)... > > >> >>2. The log files have been public for years, yet he keeps saying they were >>never made available. I have them on my laptop. Others have them. I presume >>they are still available on the IBM web site although I have not looked since I >>originally down-loaded them years ago... > > >This is all true but we do also know that Hsu et al didn't fulfill their promiss >to prove and answer everything in the aftermath. The best example is the >muteness of Joel Benjamin. You once told me that output as such wouldn't & >couldn't "prove" anything beyond reasonable doubts. So, the fact that >"something" is lying somwhere dosn't prove that much. We should better research >the team's reactions during the event. We come to my favorite point of the >psycho war. Which then leads forcedly to the invalidity of the match result. The >important point is always, that even the possible different interpretations you >could give (and you gave many in the past), that the burden of "proof" lies on >the science team and nobody else's shoulders. That's trivial for science. > As I said back at the time, it is _impossible_ to prove they did not cheat. On the other hand, Kasparov could have easily proved that they did cheat, had they done so. But he offered no proof. No computer analysis to show that this move is the only move that can be played by a computer, etc. But the basic premise here is that you can't prove that something never happened, you can only prove that it did. > >> >>In addition, remember that Kasparov signed the contract to play, under the >>conditions that were actually used. No one held a gun to his head. In fact, >>all they did was hold money out and he bit. And lost. And then started a >>years-long series of complaints and excuses. The old "a lack of planning on his >>part does not constitute an emergency on IBM's part" fits pretty well. He had >>his eye on the money and let the details slide. That's his problem, not IBM's >>or anyone else's. > > >Of course!! > >But it would look better if such arguments didn't come from a scientist. It is >and always was beyond my mind why you always relied on such a factual & >suprficial description, Bob. Don't you see, that with that tale you could >justify even cheats of mafia-like gigantism. Nobody ever doubted that Hsu and >his team didn't make a splendid job of winning by all means. But critics like me >only had to point out that science proofs and experiments can never be based on >tricks and cheats. In special the out-manouvring of its own client! ><sigh> > >You always replied that this never was about science. I can only ask you to go >figure the whole press and declarations before the event. It always was about >the playing strength of the machine and NOT about the possible tricks to psyche >out a human opponent through naked impolitness. But you can well claim that the >exclusion of impoliteness and other psycho tricks wasn't especially part of the >contract - and you were correct again! But is this the level we must discuss >this? > >You ask why this has to be repeated all over the years? Because out of justice >and fairness towards Kasparov who sure isn't a faultless figure in chess but who >wasn't outplayed at all by a superior chess machine. Admit at least this, >because you never were famous for irrational claims about the alleged >sensational strength of chess automats. You won't deminor the success of Hsu and >other friends in the scene! But as a scientist you must obey to international >standards and not sports values in the USA alone. With all the necessary >excuses. Good evening to Birmingham. Kasparov has sunk his own ship here. He'll never have any credibility with anyone on this particular topic. That's as it should be... whether or not he is the greatest chess player of all time notwithstanding...
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.