Deep Blue vs Kasparov rigged?

[wit>trope]

I just thought of this and haven't researched it at all so please don't make fun of me if it's a stupid idea

I was thinking that maybe Kasparov was paid to lose the game. Could some commercial interests have done that to increase publicity for some company or project?

 

[Aphex_Twin]

I just thought of this and haven't researched it at all so please don't make fun of me if it's a stupid idea

I was thinking that maybe Kasparov was paid to lose the game. Could some commercial interests have done that to increase publicity for some company or project?

That's highly unlikely. First of all, there's Kasparov's EGO, second, IBM didn't get much more publicity over this. Third, the "project" was scrapped and Deep Blue disassembled (it's argued today that a personal computer running an engine like Shredder or Deep Fritz would take Deep Blue with it's software then). This isn't to say Deep Blue wasn't powerful (it was still 1000 times faster than a PC today), but Kasparov wasn't very well prepared.

 

[Richard Cribb]

I just thought of this and haven't researched it at all so please don't make fun of me if it's a stupid idea

I was thinking that maybe Kasparov was paid to lose the game. Could some commercial interests have done that to increase publicity for some company or project?

No it is not a stupid idea at all and while it is difficult to prove it makes sense, that last match for instance was ridiculous.
Of course what AphexTwin points out is quite strong arguments for the match not being rigged, but there still remains some doubt.

 

[crystal]

This isn't to say Deep Blue wasn't powerful (it was still 1000 times faster than a PC today), but Kasparov wasn't very well prepared.

A factor of 1000 is maybe a little overestimated. And it only applies to calculating chess minimax algorithm. Deep Blue used 256 special purpose chess processors in parallel. Today the top commercial programs can play at roughly the same level because of their superior pattern-based pruning algorithms.

What was really peculiar in the match was that Kasparov agreed to play the games on consecutive days. That puts a huge mentral stress on the human player while the computer does not even recognize the concept of tiredness.

 

[Aphex_Twin]

A factor of 1000 is maybe a little overestimated. And it only applies to calculating chess minimax algorithm. Deep Blue used 256 special purpose chess processors in parallel. Today the top commercial programs can play at roughly the same level because of their superior pattern-based pruning algorithms.

Deep Blue could top 1 billion nodes per second (and averaged around 1-200 million). Today's multi-processor computers don't get higher than a few million (<5) and my Intel Celeron @2800 doesn't seem to get past 700,000 (shutting down some memory-expensive XP processes might boost that by another 1-200,000). The point is DB was and still would be today something to be reckoned with in terms of hardware.

What was really peculiar in the match was that Kasparov agreed to play the games on consecutive days. That puts a huge mentral stress on the human player while the computer does not even recognize the concept of tiredness.

That adds weight to my argument that Garry was ill prepared. He probably expected an easy ride.

 

[Richard Cribb]

What was really peculiar in the match was that Kasparov agreed to play the games on consecutive days. That puts a huge mentral stress on the human player while the computer does not even recognize the concept of tiredness.

I can't really agree on that.
It was only a match of six games, and the strain could easily be bigger with resting days under such circumstances, since a lot of tension can be built up. Besides the psychological pressure is smaller when facing a computer than a human opponent.
Kasparov was and is in excellent physical shape and should not have any problems with that schedule, neither physically or mentally.
But as I already mentioned, I find the last game peculiar indeed. Kasparov chose an opening which, if my memory serves me correct, he hadn't played since around 1980, and on top of that a variation which is not only dubious but also ill-fitted to his style.
Supposing the match was not rigged, and having played through those games additionaly to reading an excellent article about it by Robert Hübner, the only possible explanation then seems to be that Kasparov simply wasn't so good to play computers as some others of the world's leading players.

 

[crystal]

luceafarul, but you must also take into account that the match got a huge amount of publicity and nearly the whole world was watching. Kasparov also had to "defend the humanity". I think the Deep Blue team just tried to out psych their opponent, so that Kasparov would blunder (which he did). The playing conditions were not entirely fair for the human player. Also Deep Blue could have got help from a human player, since no outsider were allowed to oversee that the moves actually came from the computer (I think that scenario is unlikely, but still a possibility).

 

[Richard Cribb]

luceafarul, but you must also take into account that the match got a huge amount of publicity and nearly the whole world was watching. Kasparov also had to "defend the humanity". I think the Deep Blue team just tried to out psych their opponent, so that Kasparov would blunder (which he did). The playing conditions were not entirely fair for the human player. Also Deep Blue could have got help from a human player, since no outsider were allowed to oversee that the moves actually came from the computer (I think that scenario is unlikely, but still a possibility).

Yes, that's quite good points!
I totally agree on this, but still I am unable to figure out why Kasparov chose that variation, and I am pretty sure he would never have tried it against virtually any human opponent.
I think I will look for that Hübner article before I post more.

 

[Aphex_Twin]

Interestingly, I just stumbled across an open letter by one of the Deep Blue programmers. It states that Kasparov did have a chance for a rematch with Deep Blue, only he backed out. Did the K-man "chicken out" as we may say?

The letter: http://www.chesscenter.com/twic/feng.html

 

[AceChilla]

I just watched a BBC documentary about Gary's fight against Deep Blue. In it was Kasparov's point of view of the match and he still has deep suspicions about it being rigged. One particular move in the second match astounded him because he felt it was not a move a computer would ever make. Also the fact that people were constantaly putting stuff into the computer and that he was never alowed to see the log files of deep blue makes him think it was a rigged game and the computer has had human input at that particular move in match two.

After match two Kasparov became a little paranoid and depressed about the whole project and he said that affected his game and was a factor why he played so bad in the next games, esspecially match five.

I really really don't think he lost on purpose, seeing how depressed he was after some of the matches and the fact he openly accused IBM of cheating.

By the way IMB stocks rose 15% after winning from Kasparov, it also provided loads of publicity.