Kasparov Vs Deep Blue

Few remember the first car to beat a racehorse.

Deep Blue, as with all other chess computers, isn't a chess player. They use brute force calculations to see the possible outcomes of all possible moves, finding the best line and applying it in a logical, broken-down process that the human brain cannot replicate. Their only chess 'skill' is the opening books and endgame theories that humans have compiled over hundreds of years and programmed in.

A.) any car can beat a racehorse, given the proper conditions... Racehorses run at a maximum speed of less than 40 MPH. Anyone who has played horses for a living knows that.

B.) Deep Blue should not even be in the conversation of "greatest chess players in history". In the rematch with Kasparov, it clearly had help from a team of GM's, enabling it to win the match. When Garry and his team asked the Deep Blue team to furnish the computer's 'thought process' about a particular move that seemed to be human-influenced, they never complied and have to this day never provided the data. In addition, immediately after the match, they dismantled the system so that no trace of their collusion could ever be found.

There was a book written about this match (which I own) called "A New Era: World Championship Chess in the Age of Deep Blue." It was written by Michael Khodarkovsky. I suggest you read it and form an opinion of Deep Blue's natural chess ability yourself.

No doubt that Deep Blue is a strong computer chess playing program, but I don't think it belongs in the conversation of "greatest chess players in history."

 That's the only part that isn't skill isn't it? The Opening books and tablebases imitate a humans memory. It's search depth and positional evaluation ability amongst other things define its "skill."

 In response to the first post:

 I think it will be remembered more as the last serious chess contest between man and machine and as the first time a computer beat a world champion in a match than greatest man v machine match ever.

 If deep blue was counted as a player, then I think it would be up there in the top 10 - 20 in history for sure but humans have built far stronger computers since; the last I heard (a while ago) Hydra was the strongest.

  I would call an engine a chess player simply because it plays chess and I don't think the dictionary excludes the definition of "player" to organic systems. But it has a separate category similar to men and women in tennis.

But in this case, the computer had help from a team of human GM's who influenced the course of the match from the outside... Doesn't that matter?!?

A team of GMs certainly prepared and tweaked it's opening repertoire and evaluation values but there's no proof they did any more than that.

Personally, I doubt there was any cheating, Kasparov just applied a faulty anti-computer strategy. I think Kaspy was better than deep blue and in a much longer match would have won.

BTW, I've read a book on it too, it was the one written by the leader of the deep blue team. I can't remember what it was called.

 There may be no proof of any wrongdoing, but why is it that the Deep Blue team refused (and passively, I might add) to provide any substance that supported a questionable move as having been computer-generated when Kasparov's team requested such proof? The match arbiter even intervened and couldn't make any headway... The 'proof' was conspicuously missing from the records when anyone asked for it.

They had proof of all the surrounding moves and provided all that documentation when asked... Why this one move? How come they could not provide the documentation for it as well???

Because it was entered for Deep Blue by the team of GM's - that's why! It doesn't take a genius to figure that out... Why else would they be so willing to show the documentation for all other moves but mysteriously fail to come up with documentation for the one mystery move that Kasparov and his team suspected was not legitimately made by the computer?

In a court of law, one does not need to prove their innocence - but this is not a court of law we are speaking of... The Deep Blue team should have gladly been able to provide the documentation supporting the mystery move and never has. Doesn't it strike you as odd that they physically tore the system apart after the match? Hmm... Did they have something to hide???

Don't get me wrong... I'm not a conspiracy theorist. I take most things at face value, as I do this event. It simply smacks of collusion - no two ways about it.

I don't know. To be honest I'm not informed enough to competently engage in speculation. I read the book years ago and don't remember a huge amount of details. As far as I knew they did provide transparency after the match.

I would point out two things though: since the Deep Blue match computers have unarguably become stronger than any human. So isn't it logical to reason that Deep Blue was the completion of the bridging phase thus was at least as strong as the best humans? And secondly; can you imagine the PR nightmare for IBM if the team was caught cheating? They knew their computer would eventually surpass the best humans so why would they cheat just to hasten this outcome? If Kaspy had won there would just have been another match, and another... The risk far outweighs the reward.

Yes, I can. And so could they... Hence the destruction of any trace.

It's just too damn fishy that they could provide 'transparency' as you call it, of any move in the whole match except the one Kasparov and his team objected to and requested proof of... A move that, coincidentally, was the turning point of the match.

Let's not forget that there was a substantial sum of money at stake here too... For the Deep Blue team to have to split it amongst the GM advisers gives immediate incentive for them to win the match. They get more money with a win than a loss, and more total money means each person's share is worth more too.

They didn't think they were going to get found out, and Kasparov sniffed out their ploy immediately. They got caught red-handed, but knew there was no proof except for the computer's records. Therefore, they refused to provide the records by flat-out IGNORING the request for them, then claimed later that the particular records in question had mysteriously vanished through means unknown. Please...!

I wish I had the text available so that I could produce the quote here. I feel bad not being able to do so. Given time, I can find the book. I will provide the reference when I have it.

Yeah, I'd be interested in seeing a reference. I have been over the final game a few times and would be curious to know which move you're referring to. Also I would urge you to find and read the book written by the Deep Blue team leader. The explanation of why Deep Blue was so strong and where Kaspy went wrong is illustrated very convincingly.

I wouldn't call a couple of mil a substantial sum compared to IBMs net worth or what they stood to lose.

 You've peaked my interest now :) so I'll have a look into it when I get time. 

 I'm not referencing IBM's net worth or what they stood to lose at all... I'm talking about the team of human GM's they worked with.

Those guys all signed a contract to split any prize money that came from the match, win or lose. Sure, the developers of the engine get the lion's share, but that means the GM's have even less to split amongst them, win or lose. This gives the GM's incentive to feed the move in question to the machine, since they knew it was a winning move.

Incidentally, it was not in the final game of the match. It was during game 3. You might argue that a move in game 3 is hardly responsible for winning the match, but put yourself in Kasparov's shoes... A human - the World Champion, no less - who has everything to lose, is up against a machine he thinks he can beat... Then he suddenly finds out that he is not only up against a machine but also up against a team of fellow GM's working with the machine, which happens to be the strongest computer player known to man (at the time).

Demoralizing??? To say the least! Why even continue with the match? Most people wouldn't. It is a testament to Kasparov's will that he did, despite his certainty from that point on that the match was to be decided in Deep Blue's favor, no matter what he tried to do about it.

With the Deep Blue team providing no proof that they didn't assist the computer, who is to say that they won't do it again later in the match? This spectre hanging over a human player's head the whole rest of the match is too much for anyone to overcome. Even if he plays the best chess of his life he is already doomed to lose. No wonder he made some inferior moves at times throughout the rest of the match.

I will find the references and post them, as I promised to. Even if it is a month from now... I assure you, I know my history on this one and it doesn't bode well for the Deep Blue team.

No. I'm saying that Deep Blue had outside help. It wasn't "man vs. machine" but instead it was "man vs. machine-plus-men". 

Is playing chess ( level of mastery is irrelevant) enough to be called a chess player? In my opinion  chess is not just moving pieces on a wooden board, chess is much more, and thats what computer does it just moves pieces ( but it is doing it very well tough).

I think the book you mean is Behind Deep Blue by Feng-Hsiung Hsu. He doesn't describe the match as man vs machine but as man as performer (GK) vs man as toolmaker (Deep Blue), which is an interesting way of looking at it.

 If it was "just" moving pieces it would be a random number generator not a programme . No, I think it really is playing. I mean it's calculating and making decisions based on what it's been programmed to believe are the best moves. So am I when I play chess. I examine the position, decide what looks to me to be my strongest move then try to calculate whether it will work or not.

 Thanks. Yeah I recall reading him say that. I like that description.

 No, humans can still get draws and single wins occasionally against the top computers but no human is close to Hydras' (the strongest computer) strength in rating and even Anand would stand pretty much no chance in a match of any real length. I don't think it will be too long until the top computers can remain completely undefeated by humans but there might never be a stage where humans can't sometimes draw (especially with the white pieces).