IBM
Skip to main content
 
Search IBM Research
     Home  |  Products & services  |  Support & downloads  |  My account
 Select a country
 IBM Research Home
Deep Blue
Overview/Home
The Match
 ·Post-Match Images
 ·Last year's match
 ·The rules of the rematch
 ·The Commentators
The Players
The Technology
The Community

Related Links
 Press room
 Chess conference
 Site guide
 Search Research
 Feedback
 
 


Deep Blue game 6: May 11 @ 3:00PM EDT | 19:00PM GMT        kasparov 2.5 deep blue 3.5


White: Deep Blue
Black: Kasparov
1. e4
c6
2. d4
d6
3. Nf3
Nf6
4. Nc3
Bg4
5. h3
Bh5
6. Bd3
e6
7. Qe2
d5
8. Bg5
Be7
9. e5
Nfd7
10. Bxe7
Qxe7
11. g4
Bg6
12. Bxg6
hxg6
13. h4
Na6
14. O-O-O
O-O-O
15. Rdg1
Nc7
16. Kb1
f6
17. exf6
Qxf6
18. Rg3
Rde8
19. Re1
Rhf8
20. Nd1
e5
21. dxe5
Qf4
22. a3
Ne6
23. Nc3
Ndc5
24. b4
Nd7
25. Qd3
Qf7
26. b5
Ndc5
27. Qe3
Qf4
28. bxc6
bxc6
29. Rd1
Kc7
30. Ka1
Qxe3
31. fxe3
Rf7
32. Rh3
Ref8
33. Nd4
Rf2
34. Rb1
Rg2
35. Nce2
Rxg4
36. Nxe6+
Nxe6
37. Nd4
Nxd4
38. exd4
Rxd4
39. Rg1
Rc4
40. Rxg6
Rxc2
41. Rxg7+
Kb6
42. Rb3+
Kc5
43. Rxa7
Rf1+
44. Rb1
Rff2
45. Rb4
Rc1+
46. Rb1
Rcc2
47. Rb4
Rc1+
48. Rb1
Rxb1+
49. Kxb1
Re2
50. Re7
Rh2
51. Rh7
Kc4
52. Rc7
c5
53. e6
Rxh4
54. e7
Re4
55. a4
Kb3
56. Kc1
draw!


Game 4, white
19.Re1

Commentary for white move 19:

AUDIENCE MEMBER: If they had said -- or you said that they were programmed -- they would program Deep Blue to play against Kasparov's style. Now, if -- you said that they would have played -- that they would have programmed Deep Blue to play against Kasparov's style. Now, after the first match, wouldn't they have seen that Kasparov plays differently against the computer and therefore tried to program the computer to not only play against his normal style, but against the style that he plays against computers?

DB MOVE: 19 Re1 circumstances before I answer that question, I'll give myself a pat on the back for guessing the last move. The move Rh1-e1 has been played.

MIKE VALVO: There were little hints that they would do that. They imagined all the wild, unimaginable things that Kasparov might try for this match and they tried to program for them all. But they didn't get them all. You know, if you consider what happened in the previous game, yesterday's game, the computer was on its own early. Now, in this game the computer was on its own early, but went back into book. Garry wasn't completely successful in getting back into book. But yesterday they were out of book for about three of moves, they were out of the book for the rest of the game. And the computer did things it didn't actually do, too, it wanted to play d5 but it played d6 ultimately. It could have taken advantage of Garry's opening by doing something different by fianchettoing the king-side but it didn't. He was aware of that, they weren't able to counter it because they didn't April this particular order of moves. So this is going to go on. If there's another match, this is going to continue, a cat-and-house kind of thing. Garry is going to try to anticipate where are the positions the computer plays badly. And how does he know that? He has various microcomputers, and he tests them out. He wants to know what are computers going to do. And this you could do at home with your own computers. If I did something strange how is the computer going to reply. That's all he did.

AUDIENCE MEMBER: Eventually doesn't it just become -- to have it that you're playing so far off normal it's not worth doing anymore and it might be better to go back to the normal style that you're used to and play strongest that way.

MIKE VALVO: From Garry's point of view or the computer's?

AUDIENCE MEMBER: From Garry's point of view.

MIKE VALVO: I would /KPH-P that to be happening. I expected him to play c5 today, I didn't expect him to play what he did. And that caused a quandary by the way for the journalists upstairs. We told them it was the Pribyl Opening and didn't want to call it the Pribyl, they wanted to call it something more familiar like the Pirc, and I said, "But it's the Pribyl. That's what it is. It's not something else." There's lots of confusion upstairs about the things that Garry is doing. They're trying to recognize a pattern in Garry's play. They're not seeing Garry's -- the old Garry and they're not seeing it.

MAURICE ASHLEY: The question is, can the computer defeat the human. Can the human keep off the computer. And the one argument is that the computer -- if Garry were to play like the human Garry Kasparov, the natural Garry Kasparov clearly his feeling is that he'd be at a disadvantage. So he's taken on another human strategy, that is, flexibility, and said, "I can do something different. I can shoal you how human I am. I'm not fixed the way you are, a program. I can change this, I can do something different, something so alien to me that you would not have guessed, and still beat you." Isn't this sort of the whole point that makes Garry's point so good, that that's what he should be doing?

MIKE VALVO: That's what he's doing. He's trying to emphasize computer strengths -- human strengths, rather. Human strengths are recognizing patterns. Computers are not good at recognizing the patterns in this kind of situation. Generally speaking, yes, they are. But in chess there's so many different patterns that they don't foresee way on down the line. And that's what he's doing in this game. He's threatening a lot of different breaks. I didn't expect the f6 break, but he's threatening the c5 break. He's got a solid kind of position. He wants to play the computer where the computer is weak. I was hoping to see Garry play where he is strong, very creative and swashbuckling kind of chess. But he's saying I tried that in game one of last year and I didn't do well. I don't want to do that again. I want to see if I can beat him in other ways. They've got to prove to me that they can play these /POLGSZ before I'll come out and play him that way."

Real-time text commentary is made possible by LiveNote, Inc. and Vincent Varallo Associates




  


Related Information

      Program of events:

 
      Chess links:

 
  About IBM  |  Privacy  |  Legal  |  Contact