Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: A Blast from the past - Feng Hsu - Part Two

Author: Robert Hyatt

Date: 15:09:32 04/23/05

Go up one level in this thread


On April 23, 2005 at 13:05:46, Rolf Tueschen wrote:

>On April 23, 2005 at 12:19:44, Rolf Tueschen wrote:
>
>>On April 22, 2005 at 21:06:04, Robert Hyatt wrote:
>>
>>>On April 22, 2005 at 18:03:10, Rolf Tueschen wrote:
>>>
>>>>On April 22, 2005 at 09:16:11, Robert Hyatt wrote:
>>>>
>>>>>On April 21, 2005 at 18:15:50, Rolf Tueschen wrote:
>>>>>
>>>>>>On April 20, 2005 at 12:50:51, Robert Hyatt wrote:
>>>>>>
>>>>>>>On April 20, 2005 at 12:30:55, chandler yergin wrote:
>>>>>>>
>>>>>>>>On April 20, 2005 at 08:51:38, Robert Hyatt wrote:
>>>>>>>>
>>>>>>>>>On April 19, 2005 at 21:23:56, chandler yergin wrote:
>>>>>>>>>
>>>>>>>>>>On April 19, 2005 at 19:14:29, Robert Hyatt wrote:
>>>>>>>>>>
>>>>>>>>>>>On April 19, 2005 at 12:05:58, chandler yergin wrote:
>>>>>>>>>>>
>>>>>>>>>>>>On April 18, 2005 at 12:17:13, Robert Hyatt wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>>On April 17, 2005 at 10:33:57, chandler yergin wrote:
>>>>>>>>>>>>>
>>>>>>>>>>>>>>On April 16, 2005 at 07:49:08, Rolf Tueschen wrote:
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>On April 15, 2005 at 20:51:07, Mike Byrne wrote:
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>Five years ago , Hsu's open letter to the world regarding a possible rematch
>>>>>>>>>>>>>>>>with Deep Blue.
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>http://www.chesscenter.com/twic/feng.html
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>Mike, the whole topic is uninteresting. The point Hsu didn't get five years ago
>>>>>>>>>>>>>>>and earlier in 1997, is the fact that he and his team (IBM involved this way or
>>>>>>>>>>>>>>>another) cheated on Kasparov during the process of the whole rematch in 1997.
>>>>>>>>>>>>>>>For me it's so basic that they offended their own (pretended or not) defined as
>>>>>>>>>>>>>>>science experiment. They wanted to show the class of DBII in its chess over the
>>>>>>>>>>>>>>>then best human chessplayer. But what they proved in effect was not the
>>>>>>>>>>>>>>>machine's superiority in chess but their success over Kasparov's psyche with
>>>>>>>>>>>>>>>classical tricks from psycho-wars. Kasparov will never agree with this
>>>>>>>>>>>>>>>interpretation because "complete control" is his obsession and he couldn't live
>>>>>>>>>>>>>>>with the truth that they "psyched" him "out". So he worked with the absurd claim
>>>>>>>>>>>>>>>that they did never prove their authentic output of the machine. But make no
>>>>>>>>>>>>>>>mistake, Kasparov wasn't responsible during that match - for NOT being
>>>>>>>>>>>>>>>vulnerable what psychology is concerned. Because he simply believed Hsu et al in
>>>>>>>>>>>>>>>advance that they - even if they wanted to win - wouldn't cheat him, what they
>>>>>>>>>>>>>>>did as a matter of fact.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>Hsu et al (plus IBM of course) cheated
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>a) on Kasparov as their human client for the experiment which alone is indecent
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>b) on their own science responsibility for the experiment, which didn't mean
>>>>>>>>>>>>>>>winning by all means but winning through the better chess
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>c) on their own interests, because they made all further experiments obsolete
>>>>>>>>>>>>>>>with their participation, because everyone would know by now that they would
>>>>>>>>>>>>>>>cheat on you with all tricks they could organize.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>d) on the silent contract for purposes of the massmedia: in 1997 it was clear
>>>>>>>>>>>>>>>that from a chess point even the strong machine DB II still wasn't able to play
>>>>>>>>>>>>>>>chess so that such a strong player as Kasparov normally could have been beaten.
>>>>>>>>>>>>>>>That was only possible with tricks which led to the development that Kasparov
>>>>>>>>>>>>>>>was psyched out or worse, that Kasparov was confused about the real strength of
>>>>>>>>>>>>>>>the machine.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>All of the resources available were used to specifically beat ONE Player,
>>>>>>>>>>>>>>Kasparov!  Feng-Hsu made specific Chip modifications.. GM Joel Benjamin
>>>>>>>>>>>>>>'tweaked' the Program after every game, changed the Opening Book, all
>>>>>>>>>>>>>>for Deep Blue to beat Kasparov. They knew that Kasparov used the Commercial
>>>>>>>>>>>>>>Programs during his analysis.. and thought Deep Blue used the Commercial
>>>>>>>>>>>>>>Opening Books. He was Naive.. didn't realize how he was being 'sandbagged'!
>>>>>>>>>>>>>>So there was human intervention. I call that cheating!
>>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>In that light, _all_ computer vs human games will have cheating in them.  Why?
>>>>>>>>>>>>>Last time I looked, _every_ program was developed by a human programmer (or team
>>>>>>>>>>>>>of human programmers).  Of course, I suppose it is perfectly OK for the human
>>>>>>>>>>>>>players to have assistants to do opening preparation for them?
>>>>>>>>>>>>>
>>>>>>>>>>>>>this is a red-herring that is way off the mark of sanity...
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>You miss the point, as usual!  You're the red herring here..
>>>>>>>>>>>>
>>>>>>>>>>>>Millions of dollars spent to beat one man; rather than just play chess.
>>>>>>>>>>>>That is a bit off the mark of sanity also...
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>I miss the point?  You _totally_ miss the point.  IBM didn't spend millions of
>>>>>>>>>>>dollars just to beat Kasparov.  IBM spent millions of dollars to get tens of
>>>>>>>>>>>millions of dollars of free PR.
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>THE GOAL WAS TO BEAT THE WORLD CHAMPION! ARE YOU DENSE?
>>>>>>>>>
>>>>>>>>>Perhaps I am, but clearly nowhere near as dense as you, apparently.
>>>>>>>>>
>>>>>>>>>The goal of the DB team was to beat Kasparov.
>>>>>>>>
>>>>>>>>YES! Paid for by IBM! There would NOT have been a Match unless IBM had
>>>>>>>>great confidence that the Deep Blue Team COULD Beat Kasparov!
>>>>>>>
>>>>>>>Wrong.  The consensus among experts in computer science and computer chess was
>>>>>>>that the second match _would_ be won by Kasparov, just like the first.  IBM only
>>>>>>>wanted the publicity from the matches, which was nearly priceless.  The result
>>>>>>>was not the important thing to the company.  It was important to the "team" that
>>>>>>>worked on the project of course.  But the "team" is _not_ "IBM".
>>>>>>>
>>>>>>>Until you can grasp that, you will continue to run around in circles, making
>>>>>>>lots of noise, and looking like an idiot.
>>>>>>
>>>>>>
>>>>>>Since you said bye-bye to science for this question you can't preach about
>>>>>>sanity and similar problems. The truth is simply that IBM lost interest in that
>>>>>>chess thing when they saw that their team couldn't win without cheating science
>>>>>>and Kasparov.
>>>>>
>>>>>
>>>>>
>>>>>I have no idea what that means.  The "science" in this was designing the
>>>>>hardware, developing the software, and so forth.  So DB itself was most
>>>>>definitely a product of and based on "science".
>>>>>
>>>>>The match was simply a demonstration of that scientific product.  I didn't see
>>>>>anyone at IBM say anything else..
>>>>>
>>>>
>>>>Ok, let's call it demonstration. So what can be demonstrated of your baby if you
>>>>leave a normally optimal client in his - let's assume for a moment -
>>>>self-induced confused state of mind? Is this a too difficult question? Where
>>>>then remains your chess question? Or do you make the statement that DBII was a
>>>>genial psychological weapon in the game of chess? Would surprise me.
>>>>
>>>>
>>>>
>>>
>>>
>>>Not sure what that means.  But one thing for certain, if I were playing a
>>>serious match against anyone, man or machine (using crafty) you can bet that the
>>>version they play against would not have been seen prior to the match.  That I
>>>would have tuned, prepared a special opening book, etc. would be taken for
>>>granted by anyone that knows me.
>>>
>>>I'd equally expect my opponent to have prepared some things based on
>>>observations made on earlier versions.  That would be perfectly fair IMHO.
>>
>>
>>Yes - and this was never a point in our debates! >But think for a moment about
>>an opponent who becomes suspicious for not at all clear reasons - for YOU! What
>>would you do THEN? Would you call him mad at the instance? Or do you try to make
>>a reasonable conversation? That is ALL what I expected from your side, you as a
>>scientist, well and also playing computerchess. Can't you see that you as a
>>scientist have MORE responsibilities than - say - just a normal chessplayer? Ok,
>>if YOU play normal chess against a similar strong player like yourself, of
>>course you play after the FIDE rules and if he gets mad about your moves, bad
>>luck for the guy. You shouldn't help him out of the misery. BUT, here in
>>computerchess, in machine vs human player events, isn't it your job to
>>communicate with the chessplayer? I mean - it's not you as Bob is playing, but
>>only your machine, no? I dont get why you dont understand me.
>
>
>***Here I hit the submit button by accident. Excuse me. Now follows the rest of
>my answer. ***
>
>
>
>
>
>
>
>
>>
>>
>>>
>>>
>>>
>>>
>>>>
>>>>>
>>>>>
>>>>>
>>>>>> Kasparov was on science while the team and Hsu were on
>>>>>>unscientifical dope. Proof, they simply should have answered Kasparov's
>>>>>>questions - in time.>
>>>>>
>>>>>
>>>>>How does one answer an accusation of cheating?  Anybody involved in such
>>>>>demonstrations would realize that there is _no_ way to prove they didn't cheat.
>>>>>The minute Kasparov made that claim, it became a "lose-lose" situation for IBM.
>>>>>If they said nothing, they lost.  If they said anything, they could not prove
>>>>>they didn't cheat, so they lost.
>>>>
>>>>I agree insofar if Kasparov had have plans before to spoil the whole party with
>>>>such a clame. But actually I believe that Kasparov was (probably for the wrong
>>>>reasons you always explained in r.g.c.c.!) honestly and seriously confused by
>>>>certain data and interpretations in his own team (Friedel!). My clame is that
>>>>Hsu and team should have tried to explain the situation to be able to continue
>>>>the experiment. But by simply reacting the way Campbell did react, it was bad
>>>>for their own goal. And here I dont mean winning no matter how, but through
>>>>chess. If your opponent is confused you don't win if you win by your chess
>>>>alone... All that is trivial, no?
>>>
>>>
>>>Too many personalities involved, too many unknown factors involved.  Trying to
>>>predict what "might have happened" or what "could have happened" is an exercise
>>>in speculation and/or futility...
>
>
>What I can say is this: if you let a client get crazy during your test or event,
>then you can't get a sane result. This is trivial. Now I repeat the question,
>was that the intended goal in this match? To see how the machine could play
>against a confused and de-motivated human chessplayer? Makes no sense in my
>eyes.
>
>

The question was "can the computer beat the WC in a standard time-control match?
 The answer was "yes" apparently.

>
>
>>>
>>>
>>>
>>>
>>>>
>>>>
>>>>>
>>>>>That is why there were no further matches.  Why would they violate the "fool me
>>>>>once, shame on you, fool me twice, shame on me" mantra???
>>>>>
>>>>>
>>>>>
>>>>>> Kasparov didn't insult with his fair questions to Hsu and
>>>>>>his team.
>>>>>
>>>>>Please.  He _directly_ accused them of cheating.  That wasn't a "fair question".
>>>>> It was a direct accusation of cheating, made in public and not in private,
>>>>>standing on a stage, in front of news media.
>>>>
>>>>
>>>>Please! This was from a man directly after the event. Did you ever coach someone
>>>>in sports and talked to him/her right after it? Wouldn't you be  careful in your
>>>>interpretation? Anyway, all that proves what I say that the team spoiled the own
>>>>thing by losing control over the intended factors. Chess as number one. I
>>>>suppose you forgot the details. They could well have talked to Kasparov even if
>>>>they had wanted to hide their output by all means. But they didn't talk to him.
>>>>Why?
>>>
>>>
>>>You speak of kasparov as if he acts like a normal person all the time.  That's
>>>hardly reality...
>
>
>No, but I say that if Kasparov is known and famous for his suspicions and
>beliefs in supernatural, the team should have prepared plan B to get Kasparov
>back on chess grounds again. If that then had failed then they could have
>continued the way they did in 1997. But then IBM and Hsu would have another
>status than right now. The whole world would know that Kasparov behaved
>unproperly. By not even talking to him, by insulting him of being nuts the Hsu
>team spoiled the whole event whose result now means nothing in the chess world.
>
>


That goes too far.  IBM was supposed to provide psycho-therapy for Kasparov if
he went off the deep end???



>
>>>
>>>
>>>
>>>
>>>>
>>>>
>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> But scientist Hsu believed in getting away with such a misbehaviour
>>>>>>because he believed that Kasparov would prefer to sack the money without opening
>>>>>>his mouth. Hsu lost that game! Do you really believe that IBM wouldn't have sued
>>>>>>Kasparov if they had known that Kasparov was plain wrong with the allegations?
>>>>>>You bet.
>>>>>
>>>>>
>>>>>Sure they could have.  And what would that have accomplished?  "big bad company
>>>>>sues disgruntled world chess champion over cheating claim?"  They already had
>>>>>enough bad P/R.  Why put the cheating claim in every newspaper, magazine, TV
>>>>>news broadcast, etc???
>>>>>
>>>>
>>>>
>>>>You agree that it was BAD P/R? Thanks. That is what I'm saying. But you always
>>>>said that the only thing what mattered was that Kasparov was a poor loser...
>>>>
>>>
>>>
>>>Let's make sure we are on the same page.  Kasparov _was_ a poor loser.  And his
>>>cheating claim turned this into a giant wad of "bad publicity"...
>
>
>Ask real chess experts what they think about Kasparov's loss in 1997. Nobody
>will answer that DBII was stronger than Kasparov. And all would tell you that
>Kasparov played a poor chess and that he was confused and beat himself by his
>own suspicions. This is by far what IBM should have needed for a P/R. And all
>only because Hsu et all didn't want to talk to Kasparov. Isn't that clear?


I believe that on the week in question, DB was better.  I believe that because
the evidence supports it.  Whether DB was better overall is a completely
different question that has not been (and never will be) answered.



>
>
>
>>>
>>>
>>>
>>>
>>>
>>>>
>>>>
>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> -  Hsu was simply naive believing that he could treat Kasparov in such
>>>>>>a distasteful manner. - You still didn't understand the main point. You knew
>>>>>>already at the time (and explained this straight and fair to me and others) that
>>>>>>Kasparovs question couldn't be answered in a judicially relevant style. So, if
>>>>>>Hsu, who must had known this too, had told Kasparov exactly this - Kasparov
>>>>>>could have found a new playing motivation - for the best of the event. That was
>>>>>>the least the science responsible of the show should have given his client in
>>>>>>the experiment. But no, Hsu and in special Campbell thought that they could get
>>>>>>away with denouncing Kasparov's quests as nuts. Hsu got the bill for that
>>>>>>unbelievably stupid offense 6 years ago, when Kasparov didn't even answer him -
>>>>>>did you hear soemething from Hsu since that time?
>>>>>
>>>>>On a few occasions, yes.  He's moved on to other VLSI projects.  But for
>>>>>clarity, Hsu was right.  The claim was "nuts".
>>>>
>>>>
>>>>No, even if the claims were what you think, then Hsu should have clarified the
>>>>situation by TALKING.
>>>
>>>Why?  Someone says "you are cheating".  There is little to be gained by talking
>>>about it.  Just more accusations and such.  Far better had Kasparov not made the
>>>claim in public, and handled it privately in a less confrontational way.  Who
>>>knows how that would have worked out?  We'll certainly never know since he
>>>didn't take that route.
>
>Let's end this debate with the verdict that you don't accept the higher
>responsibility of scientists on the computer's side. That is enough for the ones
>who can read...
>
>


This was a chess match.  Both sides had the responsibility to show up ready to
play.  Anything beyond that is really wishful thinking.  I've seen Karpov lose
matches due to mental/physical fatigue.  Is that his opponent's problem?  Not
according to the rules of chess. I have had opponents fail to show up due to
being ill.  Is that my fault or responsibility?  No, he forfeits the game and
the tournament goes on...




>
>
>>>
>>>
>>>
>>>
>>>
>>>>
>>>>
>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>You are Naive! If the Team LOST, the Publicity would be Horrible!
>>>>>>>>IBM would be the laughing stock of the Century!
>>>>>>>
>>>>>>>
>>>>>>>So after 1996 when they lost the first match, IBM became the laughing stock of
>>>>>>>the century?  Do you now see why it is impossible for anyone to give any serious
>>>>>>>consideration to your statements?  You are not firmly grounded in reality, or
>>>>>>>you would have remembered that they had _already_ lost a major match to
>>>>>>>Kasparov, yet the project continued, and IBM was promoting DB to the hilt.  And
>>>>>>>had they lost in 1997, we would have seen chapter 3 the next year.  Losing was
>>>>>>>_not_ a problem.  At least to those of us that understand what was going on...
>>>>>>>
>>>>>>
>>>>>>Did you ever hear of the fatal consequence of an unjustified win? I doubt that
>>>>>>Americans can understand that. Because it's opposite to all the rules of
>>>>>>practice Americans believe in. As I said, chess has a different ranking of honor
>>>>>>than the Americans believe in. Here I assist to Chandler. Why IBM/Hsu didn't
>>>>>>simply play their chess and tried to improve it - the way you do it with your
>>>>>>Crafty? Because they confused winning ugly with winning at chess. Something you
>>>>>>never did, Bob. So why do you defend Hsu and IBM? Why?
>>>>>
>>>>>We played a similar match against Levy in 1984.  We prepared the same way, by
>>>>>preparing a special book, by tuning the program to avoid blocked positions.  We
>>>>>still lost.  But we did the same exact sort of preparation.  There was nothing
>>>>>dishonest about it, because David also admitted that he had studied computers
>>>>>for many years and had specific plans to beat both us and chess 4.x, which he
>>>>>did.
>>>>>
>>>>>The only difference was that in 1997 DB won...
>>>>
>>>>Because Kasparov was beaten in a psycho war.
>>>
>>>
>>>perhaps so.  After all chess is a mental exercise above all else...
>
>
>Yes, I agree. But in machine vs human chessplayers the mental tricks should be
>induced by the machine itself and NOT the team of operators. In a higher sense
>that was also forbidden by the rules where a distraction of Kasparov was
>forbidden. ;)

What "mental tricks" were induced by the DB "team"???


>
>
>
>
>>>
>>>
>>>
>>>
>>>
>>>>
>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>>  The goal of IBM was to take
>>>>>>>>>advantage of the free publicity of such a match, win lose or draw.
>>>>>>>>
>>>>>>>>
>>>>>>>>  You notice
>>>>>>>>>that after the first loss, they didn't fold their tent and run.  The PR was too
>>>>>>>>>good to walk away from.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>>AT THE TIME IT WAS KASPAROV!
>>>>>>>>>>THE PR WAS OF COURSE EXPECTED!
>>>>>>>>>>>
>>>>>>>>>>>That isn't so hard to grasp, is it?  Do you think Sonic pays those two morons
>>>>>>>>>>>lots of money to look stupid?  Or to bring attention to their fast-food chain?
>>>>>>>>>>>
>>>>>>>>>>>It was _never_ about "beating Kasparov".  That was a goal that I had, that
>>>>>>>>>>>Thompson had, that Slate had, that Hsu had, that every commercial program author
>>>>>>>>>>>had, etc.  But IBM didn't have that as a goal.
>>>>>>>>
>>>>>>>>
>>>>>>>>OF course it did!
>>>>>>>
>>>>>>>Of course it didn't.
>>>>>>
>>>>>>
>>>>>>Just for clarification:
>>>>>>
>>>>>>you and Thompson believe in machines playing sound chess and this way winning at
>>>>>>the moment when machines could play "better" practical chess against humans. But
>>>>>>IBM and Hsu believed anno 1997 that the time was ripe to win by cheating their
>>>>>>own science basics. Because winning ugly in chess isn't winning in chess. It's
>>>>>>more a character defect or insanity. Don't you see that when you yourself follow
>>>>>>that moral and logic in your own practice as a chess programmer!? Why then do
>>>>>>you defend the misbehaviour of Hsu and IBM?
>>>>>
>>>>>I simply don't see any "misbehavior".  For any contest between two competitors,
>>>>>you first form a set of rules, then you hold the contest, and you verify that
>>>>>the rules are followed.  Can you cite any rule that the DB guys didn't follow???
>>>>
>>>>Yes of course. They violated number one rule of science. If you want to find out
>>>>about x (chess of the machine), dont test factor y, you can't control! Isnt that
>>>>trivial?
>>>>
>>>>
>>>>
>>>>>
>>>>>
>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>THe Heck they didn't!
>>>>>>>>>
>>>>>>>>>Glad you agree with me...
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>  IBM's goal is to make money,
>>>>>>>>>>>make stock dividend payments, and keep the stockholders happy.  Nothing more,
>>>>>>>>>>>nothing less.
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>To make $$$$$$$$$ YES!
>>>>>>>>>>
>>>>>>>>>>By Beating the World Champion, they expected to make a BUNDLE, and they did.
>>>>>>>>>
>>>>>>>>>You do realize they "made a bundle" after the _first_ match?  How did that first
>>>>>>>>>match end?  Oh yes, a loss.  It was the _playing_ of the match that generated
>>>>>>>>>the world-wide interest.  Winning made it even better, but had they lost, and
>>>>>>>>>Kasparov kept the cheating claim in his hat, there would have been a third
>>>>>>>>>match.  And a fourth.  But not after the insult hit the street...
>>>>>>
>>>>>>How a fair questioning the scientifical details could be insultive? Arent you a
>>>>>>scientist yourself?
>>>>>
>>>>>
>>>>>Ever heard of "praise in public, chastise in private?"  The correct approach
>>>>>would have been to ask Carol to check on this specific move, privately.  I'd bet
>>>>>that Hsu/Campbell/Hoane/Tan/etc would have provided the data with no questions
>>>>>asked.  But he did it in public.  In an attempt to divert attention from his
>>>>>poor play and onto the DB group with an accusation that could not possibly be
>>>>>defended.
>>>>>
>>>>>"When did you stop beating your wife?"
>>>>>
>>>>>How to answer that?
>>>>
>>>>
>>>>Not exactly that similar. How they should have reacted? Easy one. The declared
>>>>in public "we have a serious problem of faith on the side of our client Kasparov
>>>>but we will try to settle that in every thinkable manner that could help to
>>>>solve the problem for Kasparov - as soon as possible before the next game has
>>>>started!" The rest in private. But not the way Campbell replied: somehing like
>>>>"he must be out of his mind..."  That was NOT a university seminar but a real
>>>>life situation where utmost care from the science side should have been applied.
>>>>Not to give Kasparov a bully but to save science. It's so trivial.
>>>>
>>>>
>>>
>>>
>>>way too little way too late.  Once the claim was made, it was front-page news.
>>>any resolution would be too late.
>
>I see. You, as a scientist let yourself be influenced by anything a journal or
>newspaper is writing, and that more so because you are obliged to forget about
>your science, if a it's been told to you by mass media. This is self-evident.
>NOT.

We are talking about "public perception".  Not "expert in computer chess
perception."  Those are two completely different things.

>
>
>
>
>>>
>>>
>>>
>>>>
>>>>>
>>>>>
>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>I notice you won't respond to the point about the first match.  Which simply
>>>>>>>shows that facts have no place in your arguments, just nonsense.
>>>>>>
>>>>>>Why he should respect the facts if you as a scientist is denying them too?! I
>>>>>>answered you your question. IBM was interested as long as "they" were believing
>>>>>>that this worked on scientific grounds. When they saw how Hsu et al spoiled the
>>>>>>whole myth of a scientifical challenge IBM was forced to leave it in their own
>>>>>>interest. Because the PR had turned against them. Winning ugly, cheating science
>>>>>>(and their client Kasparov), that would have negative PR... That's the simple
>>>>>>answer. It is true that a simple loss to Kasparov would have been positive PR!
>>>>>>And so the first match was no problem for IBM. If Hsu et al wouldn't have played
>>>>>>dirty a loss in the second match wouldn't have bothered IBM neither! Answer good
>>>>>>enough for you?
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>>Not even close
>>>>>
>>>>>The two matches were played the same way.  Same kind of preparation by both
>>>>>sides. The only difference was the final result the second time around.
>>>>
>>>>By chess or what?
>>>
>>>
>>>by _result_.  The thing everyone was looking at...
>
>
>You mean the mentally confused human chessplayer playing listless chess against
>a machine, the machine winning, as if that had anything valid to say to the
>world. You miss the human factor in the whole thing. If the human chessplayer
>listlessly plays on just to sack the money, your match, your decision who's
>stronger, is over, out, and you, as a scientist are busted. And you are still
>busted in the chess world if then you get the so and so prize in the USA. But
>the bad consequences are still there today.
>

Certainly it shows _something_.  Machines don't get flustered.  They don't lose
their ability to play chess by getting swamped with ideas that "I am being
cheated here..."




>
>
>>>
>>>
>>>>
>>>>
>>>>
>>>>>
>>>>>
>>>>>>
>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>>>
>>>>>>>>>>Kasparov was "Sandbagged" every step of the way!
>>>>>>>>>>If you don't believe that you are Naive!
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>One of us is "something". And it isn't "naive" either...
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>IBM got a Billion dollars worth of publicity from that, so it was obviously
>>>>>>>>>>>>worth it.
>>>>>>>>>>>
>>>>>>>>>>>They got _exactly_ what they paid for, yes.  Nothing more, nothing less.
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>Since the Program was specifically tuned to Kasparov's evaluation & Openings,
>>>>>>>>>>>>other GM's with a different style would probably  have Beaten Deep Blue easily.
>>>>>>>>>>>
>>>>>>>>>>>Impossible to say.  No way to tune a program _specifically_ to beat one player.
>>>>>>>>>>
>>>>>>>>>>That's NONSENSE and you KNOW IT!
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>Nope.  But then again I have actually written a couple of chess programs.  I'm
>>>>>>>>>not guessing.
>>>>>>
>>>>>>
>>>>>>You're not telling the whole truth because of course you can tune a machine for
>>>>>>a three games event (for both colours).
>>>>>
>>>>>Perhaps you can.  I can't.  The chess tree is simply too large.  I can't even
>>>>>prepare a book that is safe for 6 rounds against the same opponent.
>>>>
>>>>
>>>>You're not a chess GM like Benjamin. Of course I'm even less than you.
>>>>
>>>>
>>>>
>>>>>
>>>>>
>>>>>>It's also a gamble. You do know that you
>>>>>>dont do this in your experiments against GM on the net. Because it would suck.
>>>>>>You know quite well that only chess counts over longer periods. But Hsu could
>>>>>>gamble. That was the job of Benjamin.
>>>>>
>>>>>
>>>>>I see GM players lose against Crafty all the time, even though they out-play it
>>>>>on occasion, because of the clock.  Is that cheating or dishonest?  I see Crafty
>>>>>win or draw due to endgame tables all the time.  Cheating?  I see Crafty lose
>>>>>and I tweak something so that it doesn't lose that way again.  Cheating?  Sounds
>>>>>like exactly what the DB guys were doing to me...
>>>>
>>>>You filter your own story about your very distinctful manner to handle a
>>>>conflict for Dzin. All what I want is that you get it what I'm intending. That
>>>>the guys should have avoided playing dirty. For the sake of the whole event. And
>>>>possibly future continuations.
>>>>
>>>>
>>>>
>>>>
>>>>>
>>>>>
>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>Programs are specifically tuned to beat other Programs...
>>>>>>>>>
>>>>>>>>>"other programs" != "humans"
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>Changing the static positional evaluation is simple and easy!
>>>>>>>>>>
>>>>>>>>>>You are being dishonest here!
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>You are being ignorant, since you haven't done this.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>They knew that Kasparov valued the two Bishops more than the Knights, and
>>>>>>>>>>also how he evaluated Rooks, and GM Joel Benjamin 'tweaked' the Program weights
>>>>>>>>>>for these  & other factors!
>>>>>>>>>>SPECIFICALLY FOR KASPAROV!
>>>>>>>>>
>>>>>>>>>Horsecrap.  _everybody_ values two bishops higher than knights.  It is mentioned
>>>>>>>>>in every chess book ever written.
>>>>>>>>
>>>>>>>>ONLY in certain Positions are Bishops better than Knights!
>>>>>>>>Even a Beginners Book tells you that!
>>>>>>>>The more advanced Books tell you why!
>>>>>>>>
>>>>>>>>I suggest you review a few...
>>>>>>>
>>>>>>>I'll wager I know more about good and bad bishops than you do.  And in 99% of
>>>>>>>the positions the bishop pair is better.  In certain blocked pawn structures the
>>>>>>>knights are better.  Good players preserve the bishop pair until they see an
>>>>>>>unfavorable pawn structure, because it is not possible to anticipate the final
>>>>>>>pawn structure at move 10 in many openings.
>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>>You might prepare openings, but it was pretty obvious in this event that any
>>>>>>>>>>>opening preparation was not going to work since Kasparov played things he had
>>>>>>>>>>>not played before.
>>>>>>>>>>
>>>>>>>>>>He used "Anti-Computer" play as part of his plan...
>>>>>>>>>>It worked; perhaps he should have stuck to it!
>>>>>>>>>
>>>>>>>>>He did and he lost because of it.
>>>>>>>>
>>>>>>>>NO! He gambled and lost... HSU in his book explains that!
>>>>>>>>
>>>>>>>>He did not continue his anti-computer strategy.
>>>>>>>
>>>>>>>Then you watched a different match than I did.  He even gambled on the last
>>>>>>>round by playing an opening that commercial programs of that time-frame could
>>>>>>>not win from the white side.
>>>>>>
>>>>>>
>>>>>>Bob, you dont address the Benjamin issue. Why did Kasparov play that horrible
>>>>>>variant in the Spanish Opening. It's a losing choice. Why did he play that?
>>>>>>Because he wanted to prove how weak DB II really was? What's your opinion?
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>>I believe that the answer is one of the two following ideas:
>>>>>
>>>>>1.  He just screwed up by playing an opening he was unfamiliar with, he
>>>>>transposed two moves, and lost as a result.
>>>>>
>>>>>2.  He had tried that opening as black against Fritz, and won easily, and
>>>>>thought the trap would work against DB.  It didn't.
>>>>>
>>>>>Which is true doesn't matter.  In neither case is DB at fault.  You can blame
>>>>>idea 1 on Kasparov's preparation and decision to play an opening he didn't play
>>>>>much.  you can blame idea 2 on his chessbase advisors.  But he picked them.  He
>>>>>listened to them.  It blew up on him...
>>>>
>>>>See the correction by Uri and me. (The opening was played in game 2)
>>>>
>>>
>>>
>>>I am talking about the opening played in game 6...  with the h6 move that
>>>exploded.
>
>I know but I meant game 2 with the weak opening, as a famous super GM wrote to
>me in 1997. Kasparov played intentiously a losing, because too passive, line and
>wanted to see what the machine could do. And when the machine came into
>positions where it was cracking, where Kasparov tried to find ingenious
>solutions, the machine apparently was treated along Rule 13 of the contract, see
>Chandler's message. The hardware was influenced and whoopie it played the better
>move and Kasparov saw that he wasn't just playing a machine. Of course I agree
>that he shouldn't have accepted Rule 13!
>



Unfortunately for this argument, he chose to accept the rules, that was not our
decision to make...



>
>
>
>>>
>>>
>>>>
>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>No wonder they didn't want a Re-Match! Kasparov had learned from his games.
>>>>>>>>>>>>
>>>>>>>>>>>>This match did not prove machine superiority over a human!
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>It proved DB's superiority over Kasparov for a week back in 1997.
>>>>>>>>
>>>>>>>>Total NONSENSE! It proved NOTHING!
>>>>>>>>Kasparov gambled and lost..
>>>>>>>>
>>>>>>>>
>>>>>>>>You consider a Match of 6 games significant in any respect?
>>>>>>>
>>>>>>>
>>>>>>>Yes, considering that no world champion had lost any sort of match at long time
>>>>>>>controls to a computer prior to that event.  So it definitely has significance.
>>>>>>>At least to most of us.
>>>>>>
>>>>>>
>>>>>>Not in chess circles! Every expert in the sciene knew that Kasparov wasn't
>>>>>>playing as Kasparov. If one can call this gambling?
>>>>>
>>>>>
>>>>>All I can say, is that this has gone down in history as the first time that a
>>>>>world champion (human) lost a match to a computer, played under tournament time
>>>>>controls.  Nothing more, nothing less.
>>>>
>>>>Under what circumstances? Against a sane opponent or a confused one? What did
>>>>the result mean in truth? - All important questions. Also in chess. And you know
>>>>that!
>>>
>>>
>>>You miss the point.  The general public perception is "machine beat man, end of
>>>the story."
>
>
>As a scientist I don't get impressed by floods of people and their perceptions.
>This is not politics or commercials on TV.  ;)
>
>Thanks for the exchange so far, Bob. I wish you a nice weekend.
>
>
>
>
>>>
>>>
>>>
>>>>
>>>>
>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>>A loss of the last game to a stupid Computer is not significant!
>>>>>>
>>>>>>
>>>>>>Chan, it kicked IBM out of the field... So it was a significant ugly winning the
>>>>>>match for IBM. <g>
>>>>>>
>>>>>>
>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>  Nothing more,
>>>>>>>>>>>nothing less.
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>That is not my personal opinion but the verdict of several commentators out of
>>>>>>>>>>>>>>>the GM scene. People who can read the game and its problems. In Germany at first
>>>>>>>>>>>>>>>GM Unzicker criticised the match for its chess content and later GM Hübner
>>>>>>>>>>>>>>>showed where Kasparov played out of fear. So that scientifically, I conclude,
>>>>>>>>>>>>>>>the whole event didn't prove anything about the 1997 strength of a chess machine
>>>>>>>>>>>>>>>in a meaning of superiority over human race. Its chess simply was too bad. With
>>>>>>>>>>>>>>>the exception of the game two, where God's hand might have come into play... or
>>>>>>>>>>>>>>>human interventions.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>Hsu's and DB II's main defender here in CCC and usenet, Prof. Hyatt, did always
>>>>>>>>>>>>>>>point out that this was a match about winning. And this way it was in accordance
>>>>>>>>>>>>>>>with all what makes American sports and spirit for fight so lovable. Bob always
>>>>>>>>>>>>>>>explained that this wasn't about science, alone because of the leadership of IBM
>>>>>>>>>>>>>>>that mainly had commercial interests. But we in the World of chess we do know
>>>>>>>>>>>>>>>what we read and saw in the massmedia and we defend our hero Kasparov, about
>>>>>>>>>>>>>>>whom we did well know that he was easily to irritate by suspicious details. That
>>>>>>>>>>>>>>>was the only weakness he had. But therefore winning against him by such
>>>>>>>>>>>>>>>hokuspokus disturbances did NOT decide who was the stronger chessplayer, since
>>>>>>>>>>>>>>>all the pychotricks didn't come from DB II but from the ingenious team around
>>>>>>>>>>>>>>>Feng Hsu. And therefore it's over for Hsu. He should challenge FRITZ, SHREDDER
>>>>>>>>>>>>>>>or DEEP JUNIOR! But no more human chessplayers. Period.



This page took 0.01 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.