English Chess Forum

Deep Blue v Garry Kasparov - Game 2 Position 36 White To Move -1997 Re-Match.

Garry said he could not understand how a Computer could play such a move as Deep Blue's BC2 – E4
in the above Position.

The truth of the matter is that Garry never knew how Advanced Deep Blue was! Computers with enough Knowledge can Replicate and even Improve on Human Knowledge ...as is what is now happening in many areas such as Science,Mathematics,Chemistry..Etc. as well as Chess.

To Prove a Point!...

The Best Chess Engine as of January 2011 is the Free UCI Engine ( Houdini 1.5 ) Programmed by Robert Houdart,which is even more Powerful than the famous ( Rybka 4 ) Programmed by Vasik Rajlich,which is Commercial Software now produced by Chess Base, and regarded as the Strongest Engine?.

In the above Position Houdini 1.5 Plays BC2 – E4 (Score +54) for White! at different time levels below 60 Seconds! Well within the Match time allowed...and also Plays the Strong QF2 – B6 (Score + 40) for White! Both moves are Winning.

So there we have it Garry what you never understood at the Time? Can now be proved to you...But you are and will always be Regarded as One of the Strongest Human Chess Players in History! even though I personally believe if Paul Morphy could have lived longer...instead of the his Tragic Death! he would have been the Strongest Player Ever!!

A.R.B

-----------------------------------------------------------------------------------------------------------------------------------------------------

Below Are Both The Games Using .pgn File Format.


[Event "Deep Blue v Garry Kasparov - Game 2 Position 36 White To Move -1997 Re-Match - B-E4"]
[White "Deep Blue - Houdini 1.5"]
[Black "Garry Kasparov - Houdini 1.5"]
[Result "1-0"]
[SetUp "1"]
[FEN "r1r1q1k1/6p1/3b1p1p/1p1PpP2/1Pp5/2P4P/R1B2QP1/R5K1 w - - 0 1"]
[Termination "adjudication"]
[WhiteType "program"]
[BlackType "program"]


1. Be4 Rcb8 2. Ra6 Qd8 3. Qa2 Rxa6 4. Qxa6 Kf8 5. g4 Bc7 6. h4 Rb6 7. Qa7 Kf7
8. Kg2 Qd6 9. Qa2 Ke7 10. g5 hxg5 11. hxg5 Kd7 12. Qa8 Rb8 13. Qc6+ Qxc6 14.
dxc6+ Ke7 15. gxf6+ gxf6 16. Rh1 Rg8+ 17. Kf3 Rf8 18. Bd5 e4+ 19. Ke3 Kd6 20.
Bxe4 Bb6+ 21. Kf4 Rf7 22. Kf3 Kc7 23. Rh6 Kd6 24. Rh8 Bc7 25. Re8 Re7 26. Rf8
Ke5 27. Rg8 Kd6 28. Rg1 Bb8 29. Rg6 Rf7 30. Rg8 Bc7 31. Rg2 Bb6 32. Rh2 Kc7 33.
Rh8 Bg1 34. Rh6 Kd6 35. Rh5 Bb6 36. Rh8 Kc7 37. Rg8 Ba7 38. Bd5 Re7 39. Be6
Kxc6 40. Rf8 Kd6 41. Rxf6 Ke5 42. Rf8 Rh7 43. Re8 Kf6 44. Ke4 Bf2 45. Rf8+ Ke7
46. Rb8 Rh3 47. Rxb5 Rxc3 48. Kd5 Rd3+ 49. Kxc4 Rd4+ 50. Kc3 Rg4 51. Rb7+ Kd6
52. Kb3 Rf4 53. Rb5 Kc7 54. Bf7 Rf3+ 55. Ka4 Rf4 56. Ka5 Be1 57. Be6 Kd6 58.
Rd5+ Kc6 59. Rc5+ Kd6 60. Bc4 Bc3 61. Ka4 Ke7 62. b5 Bd4 63. Rc7+ Kf6 64. Rf7+
Ke5 65. Bd3 Bb6+ 66. Kb3 Kd4 67. Rb7 Kc5 68. Rb8 Rf2 69. Rc8+ Kd4 70. Bc2 Kd5
71. Rc4 Rf1 72. Be4+ Ke5 73. Rc6 Kxe4 74. Rxb6 Kd5 75. Rf6 Kc5 76. b6 Rc1 77.
Ka3 Rc3+ 78. Ka4 Rc1 79. Kb3 Rg1 80. Kc3 Rg7 81. Rg6 Rh7 82. f6 Rf7 83. Kd3 Kd5
84. Kc2 Kd4 85. Kb3 Rb7 86. Ka4 Rxb6 87. f7 Rxg6 88. f8=Q Ke5 89. Kb5 Rg4 90.
Qe8+ Kf5 91. Kc5 Kf4 92. Kd5 Kf3 93. Qh5 Kf4 94. Kd4 Rg5 95. Qf7+ Rf5 96. Qg6
Ra5 97. Qd6+ Kf5 98. Qd7+ Kg6 99. Ke4 Ra6 1-0



[Event "Deep Blue v Garry Kasparov - Game 2 Position 36 White To Move -1997 Re-Match - Q-B6"]
[White "Deep Blue - Houdini 1.5"]
[Black "Garry Kasparov - Houdini 1.5"]
[Result "1-0"]
[SetUp "1"]
[FEN "r1r1q1k1/6p1/3b1p1p/1p1PpP2/1Pp5/2P4P/R1B2QP1/R5K1 w - - 0 1"]
[Termination "adjudication"]
[WhiteType "program"]
[BlackType "program"]

1. Qb6 Rxa2 2. Rxa2 Ra8 3. Ra5 Qb8 4. Qxb8+ Rxb8 5. Ra6 Bc7 6. d6 Bb6+ 7. Kf1
Be3 8. Ke2 Bf4 9. Be4 Rd8 10. Ra8 Rxa8 11. Bxa8 Kf8 12. Bc6 Bh2 13. Kd2 Bg3 14.
Kd1 Bf4 15. Kc2 Bg3 16. Bxb5 e4 17. Bxc4 Ke8 18. Bd5 e3 19. Kd3 Bxd6 20. Kxe3
Ke7 21. Kd3 Kd8 22. Kc2 Kc7 23. Kb3 Kb6 24. c4 Be5 25. Be6 Bf4 26. c5+ Kb5 27.
Bd7+ Ka6 28. Kc4 Bd2 29. Bc8+ Ka7 30. b5 Kb8 31. Be6 Kb7 32. b6 Ka6 33. Bc8+
Ka5 34. Kd5 Bf4 35. Kc6 Kb4 36. Be6 Bh2 37. b7 Ka5 38. Bc4 Be5 39. Kd7 Bg3 1-0

Computer chess has advanced exponentially since 1997, so it's not surprising the best machines today find the move quickly.

Nonetheless Kasparov's argument that a computer couldn't have possibly played this move is pretty rubbish considering he is not a computer programmer and wouldn't have had access to the algorithms used by the Deep Blue team. As a trivial example, it's easy to add a strong weight to moves that sacrifice a pawn in the centre, and an engine would be biased towards these kind of moves. So to say that an engine could never play these kind of moves is simply wrong, it all depends on how the algorithms are set up.

I think that what has kept this debate alive for so long is that IBM could have rubbished Gary's claims so easily.

After the match, they could have allowed Garry (and journalists) into the computer room where Deep Blue was housed and embarassed him by re-creating the position and allowing the Computer to reproduce the move infront of his very eyes.

Given that Deep Blue was to be dismantled anyway and wasnt going to play anyone else - there were no trade secrets that were to be considered.

I really don't know if Deep Blue had human assistance or not, but what we all know for sure is that IBM could have brought an end to this speculation so easily - but yet chose not to.

The Very Important Point that Everyone Against IBM Misses is....

That (Garry Kasparov Allowed The position To Happen!)

What I mean by this is... if Garry was Playing against a Strong Human Opponent He would not have said anything,and just got on with the Game,it does not Matter whether His Opponent was Human! or Computer! He made the Critical Mistake and Allowed the Position to Happen!

A.R.B

I don't think anyone is missing that point.

Kasparov had been playing some kind of anti-computer strategy which goes some way towards explaining how he "allowed" that position to happen.

Of course - with hindsight - his anti-computer strategy may have been a mistake, or possibly he misunderstood how strong Deep Blue in fact was.

There has been so much time that has passed that imo it is of no value to look at a modern engine and say "ah yes this result is found in 60 seconds so it proves xyz" - you'll never prove this one way or another. The only way to prove this 100% is to recreate the conditions - with the same hardware and software.

That opportunity was available after the match and for some reason IBM chose not to go down that path. Perhaps you might want to spend some of your energy explaining that one !

Let's be clear - if I was accused of cheating and I had the means at my disposal to show that was definitely not the case I would take that opportunity. Wouldn't you ? Wouldn't most people ? Of course neither does this prove anything.

Hi Peter Rhodes _________________
Chess Amateur.

The IBM Deep Blue team were a Professional Team! And after being basically accused of Cheating by Garry could not be bothered lowering themselves to His level.
They proved what they had to! And that was that...and Again it was because (Garry Kasparov Allowed The position To Happen!) whether it was some Anti-Computer Play made no Difference to the outcome!

A.R.B

Kasparov didn't exactly have the greatest reputation for sportsmanship - touch move violation against Polgar, extended rant about his loss to Radjabov at Linares being awarded the game prize solely because it was the only game he lost - so his accusation of cheating was probably just another case of sour grapes. He probably did underestimate how strong Deep Blue was, but in fairness he didn't have much to go on in that respect - he didn't have access to any of Deep Blue's previous games to prepare for playing it, whereas the IBM team had hundreds of Kasparov's games.

Michael Jones wrote:Kasparov didn't exactly have the greatest reputation for sportsmanship - touch move violation against Polgar, extended rant about his loss to Radjabov at Linares being awarded the game prize solely because it was the only game he lost - so his accusation of cheating was probably just another case of sour grapes. He probably did underestimate how strong Deep Blue was, but in fairness he didn't have much to go on in that respect - he didn't have access to any of Deep Blue's previous games to prepare for playing it, whereas the IBM team had hundreds of Kasparov's games.

Fair summary - and the computer didn't make him play one of the worst games of his life in Round 6

Still, computers are quite cheating machines by their nature. Try playing a strong one with just your own wits and no other form of help- you will find out what I mean.

Kasparov was too good for any computer in those days, and this arguement that it would be 'sinking to his level' to revel the facts does not hold water. Just imagine if a defendant decided to use that in court "oh, no, your honour. I am not going to present evidence for my innocece as it would be embarrasing for me to have to do so"

I say, guilty as charged!

I've no idea why they didn't publish the logs at the time, but they have done since so you can check the evidence for yourself: http://www.research.ibm.com/deepblue/watch/html/c.shtml

Michael Jones wrote:I've no idea why they didn't publish the logs at the time, but they have done since so you can check the evidence for yourself: http://www.research.ibm.com/deepblue/watch/html/c.shtml

Thanks for the link Michael.

Those who accuse Deep Blue of outside assistance are not going to be satisfied with the log because they will say "If the Deep Blue team were capable of cheating then they are certainly capable of modifying a log".

IBM dismantled Deep Blue as soon as they could, and with it they took away the only completely verifiable way for them to refute the accusations.

Peter Rhodes wrote:Those who accuse Deep Blue of outside assistance are not going to be satisfied with the log because they will say "If the Deep Blue team were capable of cheating then they are certainly capable of modifying a log".

The trouble is that if the computer hadn't been dismantled and was shown to have been able to find the move unassisted people (who wanted to) would still say, "Yes, but they fiddled it in some way". It becomes a conspiracy theory that no evidence can refute.

I'm not particularly interested in the 'truth' of whether IBM cheated or not, btw.

What I do know is that Game Over: Kasparov and the Machine is a marvellous film and well worth watching regardless of your point of view. It only costs a few quid too.

Jonathan Bryant wrote:The trouble is that if the computer hadn't been dismantled and was shown to have been able to find the move unassisted people (who wanted to) would still say, "Yes, but they fiddled it in some way". It becomes a conspiracy theory that no evidence can refute.

If the journalists had been allowed entry then I hope they would have provided some independence, but I'm sure you're right Jonathan. Kasparov would have found some context for disbelieving the evidence.

... and yes "Game Over" - that is a great film