Commentary for white move 32:

AUDIENCE MEMBER: About computer rankings, there's probably not just one computer ranking but there's a computer chess ranking and a human computer ranking, and I think that Garry has increased his computer chess ranking enormously since Philadelphia, and that this may make the difference, that although Deep Blue has increased its ranking, Garry's ability to play computers has increased considerably.

YASSER SEIRAWAN: All right. The gentleman explains that indeed, there is a world chess ranking, and we have somewhere in the neighborhood of 600 to 700 Grandmasters out of the many millions of people who play chess. And the Garry is the highest ranked human player in the world. I believe his ranking is 2820, which places him comfortably ahead of his nearest rival.

MIKE VALVO: That's an interested thought he introduced to us. Garry he's saying is better at playing computers than anyone? I think you're inferring there.

AUDIENCE MEMBER: And that he improved since Philadelphia.

MIKE VALVO: And that he's improved since Philadelphia.

YASSER SEIRAWAN: Well, also the idea that Deep Blue is ranked. There is the Swedish computer rankings, which are the most respected rankings, and Deep Blue is not ranked by the Swedish ranking because the /SWAOEDZ want to do their own independent tests, and of course they don't have a version of the Deep Blue parallel machine. But I think what you've suggested is that Garry is a better player because of the Philadelphia experience. We can certainly say that Deep Blue is a better player because of the Philadelphia experience. I know that Garry has adapted his style to the game today and to the match. So we're seeing a different Garry Kasparov. We're not seeing a Garry Kasparov that I would play against much he's clearly adapted his style. When he has adapted his style, does that make him a better player? I'm not sure. I don't know. Is Garry the best opponent for the computer? Again, that's questionable. That's questionable. Because it seems to me that there are particular players in the world, a friend by the name of Ulf Anderssen is a fantastic player, he's a python-like player, he just squeezes the opportunities for his opponent out of the position and I subpoena that -- suspect that Ulf would be a terror against computers.

MIKE VALVO: Probably would be the world champion human against computers.

YASSER SEIRAWAN: A phenomenal player. Yes, we have a question, please?

AUDIENCE MEMBER: Could you explain the rules regarding if Deep Blue is offered a draw, who gets to decide if he accepts? Last year the --

YASSER SEIRAWAN: That's a very good question, because last year there was a disruption regarding the draw offer. Mike?

MIKE VALVO: Yeah, I was the arbiter last year, and Garry offered Deep Blue a draw in the fifth game, and the problem was that a decision was made by the Deep Blue team and it took so long to make a decision that Deep Blue not knowing any of this decided to go ahead and make a move. (Audience laughter.)

YASSER SEIRAWAN: Wasn't waying the possibilities of the prize fund at the time.

MIKE VALVO: Which obviously creates a problem. And last year we had a special rule for the last game that if Garry had offered a draw in the last game, that they could think about it, and if Deep Blue decided to make a move, that they would be allowed to think about it and their clock would keep running, but they couldn't accept Garry's offer. They had to make the move and then offer Garry a draw in reply. That's what we did last year. I'm not sure what the final resolution is this year. I don't think we have a resolution. I think that's going to be decided real-time.

YASSER SEIRAWAN: Well, in fact I was a part of some of these earlier discussions just as we were starting the game yesterday, and Carol Jarecki, by the way, is the arbiter, and we will be ininviting Carol hopefully during the course of our presentation, to answer this question. But the idea is very simple. I offer you a draw, and if you've already determined to make your move, then the move is going to get played. And we must remember that Deep Blue's team declined Garry's draw offer because what they wanted to do was learn, so much as draw Garry Kasparov. A nice bee in your bonnet to be sure, but, you know, they wanted play. And interestingly enough, Garry hasn't offered a draw in this game, a game that probably at this moment stands slightly worse. But I would be curious to see the effect as everybody powwows and decides what would you do. We have a question. Yes, sir.

AUDIENCE MEMBER: This particular position here? It seems like the computer has just made a few moves that indicate that it doesn't have a clear strategy. I mean he moved the queen first to f1 and then also going to open the a-file, then moved back to f2, moved this bishop to this diagonal there and now is backing up the bishop again. I mean it seems like it's confused now. Do you think so?

YASSER SEIRAWAN: Well, the question that you're asking, is Deep Blue losing the thread of the position, has he not taken the most of his advantages. For example, the last move, this rook on -- the previous move, I beg your pardon, move 31 H3-a2. Let me just show you the position. This move. It looks a little bit like a wasted tempo like what the heck is that all about. But the idea is white wants to prepare the move axb5 but he's side stepping the tactical nuance Nxb5. So I don't think Ra3-a2 was a bad move. In these kinds of positions it's very hard to just slam an advantage home, what Joel had mentioned earlier, this idea of opening up a front on the king-side to go along with some of your maneuverings on the king-side is quite true. It's very difficult to say this should be the way to win it, and go ahead and win it. So Deep Blue is groping a bit. I think that's all right.

MIKE VALVO: And there's another factor there, too. What does black have to do in the meantime? Can white afford to make many meaningless moves like this to prepare the situation? And can black do anything about his position in the meantime? I think black is pretty much got as close to an optimum position that he could have with his knight on d6. What has he got to do?

YASSER SEIRAWAN: Right.

MIKE VALVO: Constructively.

YASSER SEIRAWAN: Indeed. One of the things that he ought to be looking at is that this bishop on d7, he might try to just drop back with his queen, with the idea that he could drop his bop on d7 back to the square e8, and perhaps with the idea of f7-f6 as well, this bishop could come out, if you will. White's dilemma -- and I do believe that white has the advantage -- is this c2 bishop. It's actually a nice bishop, but it's impeded by this pawn at e4. I'm sure it would be to white's advantage to try to be able to trade bishops, a line like Bd1-g4 would in the long run ensure that the control of the f5 square is made. So I think it would be in white's favor to try to force the exchange of the white squared bishop. Still, this is one of those kind of maneuvering positions, style. We have another question in the audience? Sir?

AUDIENCE MEMBER: Yes. It seems to me that given that black has very little to do to further his further, it seems the white position ought to be to adopt a slow, methodical strategy, seems that he could put a lot more weight in a relatively short period of time on the f7 pawn than black could possibly bring in support of it. So the idea would be to start bringing all of the rooks openly to lean on the f7 pawn, reposition the bishop to h5, leaning on the f7 pawn, gradually basically trying to pile everything possible on that, given that black has very little, relatively speaking, that it can get over in support of that pawn.

YASSER SEIRAWAN: The question is if the f7 pawn really that weak. As we look at the actual game position, black has three supporters: The queen on e8, the king on g8, the knight on d6. So to reasonably expect that white could crash through on the f7 square I think is an unreasonable expectation.

AUDIENCE MEMBER: Well, white can eliminate one at any time by trading the bishop for the knight.

YASSER SEIRAWAN: That is true.

AUDIENCE MEMBER: And he can bring -- black has a limit to what he can bring in support in the long Ryan whereas white has no limit to what he can bring. He can bring every piece in support to an attack on f7 except for his black bishop.

YASSER SEIRAWAN: Well, it would be a major repositioning of white's /PORSZ. As you say the bishop would have to come to the h5 square. The rooks themselves would have to swing over to the f-file --

AUDIENCE MEMBER: But black has nothing to do short-term.

YASSER SEIRAWAN: I understand that. But, again, I think it would be an unreasonable expectation that you would crash through on the f7 square.

MIKE VALVO: It's the kind of position that you threaten here, you threaten there, you threaten here, you threaten there, you don't expect to actually capture does, -- something, but you expect concessions of one sort or another.

YASSER SEIRAWAN: If I was to try to focus on white's position, again, I would want to do it on the queen-side. I would want to resolve the queen-side conflict with an ab and ab. Really prior to doing that I would like to trade this bishop off and I would like to play like Bd1-g4. I don't expect that Deep Blue would do that. I think, quite frankly, that that isn't an move or maneuver it would make, but yet I think it's the best maneuver.

MIKE VALVO: Yeah. And probably for a couple of reasons. One it's a long, involved process, and secondly it would involve doubling this pawn here which the computer wouldn't like very much at all. This is where a human has the advantage because it sees "Well, gee, that bishop isn't doing anything, let's get it off the board and get its counterpart." The computer doesn't think that way. It's just adding one position on top of the other to come up with a move.

YASSER SEIRAWAN: We're going to take one more question and then I want to finally finish the review of the opening. One more question. Do we have one?

MIKE VALVO: We have one over here.

AUDIENCE MEMBER: Yes, hello.

YASSER SEIRAWAN: Hello.

AUDIENCE MEMBER: Actually, just a very quick question, one on the f7 square. If a well timed f7-f6 unreasonable as a possible defense? I'm just throwing that out even though the light squares look weak, there's still the light squared bishop on them, and regrouping to solve that problem. And number two, the main question I want to ask is, it's kind of paradoxical because it looks to me like white's great E advantage lies in the ending. Yet computers don't tend to play endings as well. So does Garry continue to play for a possible ending that where he could be somewhat inferior but have winning chances in? I mean -- I don't understand, I guess.

YASSER SEIRAWAN: Okay. A multiquestion there. Is it in black's interests to play f7-f6 at some point in the future, and I think that the answer is inequivocally yes, for some of the reasons that you mentioned, the half open file, black is potentially vulnerable on the f7 square, so it makes sense to play the move f6 which would support the e5 pawn. But at the moment -- at the moment black might be able to scoot his pieces out over on the light squares on the king-side. And in fact I mentioned that after a move like Qd8, the bishop could drop back to e8, as he played. I just wanted to, finally talk to -- shall I try? -- maybe we'll see a move shortly, or else I'll never get through this opening. Okay, we're in the Spanish. An ancient opening. And we get to this very-well-known position. Black has just played the move b5 after Bb3.

YASSER SEIRAWAN: Hey, I'm never going to get through this,

MAURICE is back, international master

MAURICE Ashley.

MAURICE ASHLEY: And

YASSER, if it's okay, we'll give him a rest. Thank you very much. We'll be seeing you later.

MAURICE ASHLEY: And I would like to introduce former two-time U.S. champion, the author of "The complete idiot's guide to chess," International Grandmaster Patrick Wolff. (Audience applause.)

MAURICE ASHLEY: So, Patrick, I'm going to play the idiot. Tell me what's going on?

PATRICK WOLFF: Well, we're all the idiots compared to Garry and Deep Blue, but so far as I can tell, the computer's been playing very well. Has a nice, solid space advantage, has some pressure on the queen-side position. As I'm sure that Mike and

YASSER have been explaining to you, it's a slow game. The super is closed up. There aren't a lot of tactics that are going to be happening right away. But white has some pressure on the queen-side. Black has some weaknesses over there to take care of, and white has more space in the middle of the board. That means that white's pieces, generally speaking, are going to be more powerful than black's pieces.

MAURICE ASHLEY: Are you surprised by the way the computer has played this position, sort of like Karpov, almost, you know, very positionlyal -- positional, very slow, playing for the points, playing for the space. Are you surprised by this kind of method by the computer?

PATRICK WOLFF: Well, you know, there's an old joke or have enough -- old puzzle, if you have enough monkeys typing in the room, you get all of Shakespeare's works. And I think what you're seeing here if you get the calculations growing with the right kind of algorithms you can get the style of Anatoly Karpov as

MAURICE has been saying. Anatoly Karpov is a great chess player, former world champion, famous for his positional play, deadly position, take all the play out of the position, suck the blood like a vampire, and then he'll kill you, unlike Kasparov who comes up, whack. I think the computer is playing very well. I am surprised. I think the computer has managed to hold onto a solid advantage. I haven't liked a few of the decisions that Deep Blue has made, but overall, no complaints, good game.

MAURICE ASHLEY: Aim not sure how the programmers of Deep Blue would appreciate that monkey comment. But --

PATRICK WOLFF: It's only the computer, by the way. It takes very smart people to line all the calculations up so that you can have them all typing away.

MAURICE ASHLEY: Well, what, in your opinion, has been the -- have been moves that the computer has played, in your field, are not quite the best? You may want to take a look at the announcer's board, the kind of moves so far that you would critique.

PATRICK WOLFF: Okay. Well, if we go back a little bit,

MAURICE, let's go back to where --

MAURICE ASHLEY: Before we do that --

PATRICK WOLFF: Bf8 I believe is the move?

MAURICE ASHLEY: Kasparov has retreated his bishop back.

PATRICK WOLFF: I think it's important to talk a little bit about what's been happening with this dance of the bishops. It's probably a bit mysterious, and a lot of times it may seem as if nothing is going on in a position like this. You know, compared to yesterday's game where rooks and bishops and pawns are going all over the place, it may seem like somehow, you know, well, just play anything and it'll do.

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