|Mar-24-05|| ||Karpova: impressive play by smirin! just when it seemed like tiger was winning, smirin got an edge. |
|Apr-29-05|| ||fgh: Good notes and game.|
|Sep-18-05|| ||lopium: I was speaking about that for me at least, every chess game is a story. With these comments, the feeling is stronger, don't you think?|
|Dec-03-05|| ||netlava: Are computers stronger positionally or tactically? Because sometimes I hear computers are strong at tactics.|
|Feb-11-06|| ||Catholic Bishop: brilliant game with some idiotic notes:
"A typical game against computers. They are positionally superior to us humans. but we still have the tactical edge."
|May-13-06|| ||DeepBlade: Gambit Tiger loses game because he accepted the gambitted piece on a1.|
oh god the <iron>y is killing me softly, like <rust>.
|Feb-23-08|| ||wolfking: ehmmm... i think it's the other way: Computers are strong tactically and humans positionally|
|Sep-15-08|| ||Katu: Huh, i like that the comments were from a chatroom...|
|Nov-29-08|| ||Xeroxx: annotated by NN!|
|Feb-20-09|| ||Brown: Smirin indeed was fortunate that Tiger did not recognize 29..Rxa4 as a nearly winning move for black. |
With this in mind, Smirin's only hope is the immediate exchange sac 29.Rxc3 dxc3 30.Rc1. Black's best seems to be 30..Rxa4 31.Rxc3 Qf7, but this runs into 32.Rc7 Qh5+ 33.Nh4! and black is surprisingly lost.
Never mind those lines, as after 29.Rxc3 dxc3 30.Rc1 Rxa4 31.Rxc3 the coolheaded 31..Rd8 keeps black in the driver's seat.
|Feb-20-09|| ||swarmoflocusts: <"A typical game against computers. They are positionally superior to us humans. but we still have the tactical edge."|
I can only hope that this is a joke reflecting the irony of a human catching a computer in a deep tactical net.
Usually, as others have said, computers play extremely accurately tactically, but humans hold an advantage in that we can conceive of deep positional elements, whereas computers (until Rybka) look only at what tactical ideas are available in a position. (This is why I find Rybka so scary -- it CAN understand positional elements. If we create a computer that understands every positional element at the level that Rybka understands space and can calculate tactically at the level common to today's computers (let alone the level at which computers will be able to calculate in the future), it will play chess like we have never seen it.)
|Feb-21-09|| ||Brown: No, there is no chess in the future that will look too different, and no computer will ever be flawless in every possible position.|
Addendum to my above post... White may look to improve his knight earlier on, say with 28.Nh4 followed by f4 and Nf3, or hop to Nf5 if black allows it. White is not worse, it seems. The a1 Rook is still verboten for black, and black's rooks are in danger of being out of play on the Q-side.
|Feb-22-09|| ||swarmoflocusts: <Brown: No, there is no chess in the future that will look too different, and no computer will ever be flawless in every possible position.>|
Forgive me, but do you have any reasoning at all behind this? Imagine a computer that could perfectly analyze every element of every position and from that deduce the best strategy, regardless of human considerations? Think of Rybka's brilliancies, then add perfect analysis of every other positional element. It would be amazing. As for your second statement, why do you assume we will never have computers that powerful? Computers are growing more and more powerful at an incredible rate.
|Feb-22-09|| ||Brown: You are forgiven.
Modern chess practice supports my views. Your argument suggest that I "imagine a computer..." Exactly. One can only imagine.
What you purport to be "perfect analysis" does not exist amongst humans in ALL positions, yet we are to expect humans to create a program that will have this understanding?
Chess has finite possibilities, and without rule changes, chess cannot look significantly different. Humans can play very accurately, so I don't see how computers will ever show us something extraordinarily different.
I think producing nearly-unbeatable chess programs is a worthy human interest, but nothing created will be perfect, and we're all the better for it.
|Feb-22-09|| ||ontocaustic: are your eyes brown?|
|Feb-24-09|| ||swarmoflocusts: The problem with what you call "accurate" human play is that humans, by our nature, are limited as to what we can see in different positions. Computers can play moves that humans would never consider. The true geniuses of chess are able to play counterintuitive moves in certain positions -- what if we created a computer that could play the correct positional move, whether counterintuitive or not, in every position?|
While I understand your doubt that such a computer could exist, I must insist that, frankly, we have absolutely no idea what computers in the future will be capable of.
|Feb-28-09|| ||Brown: <swarmoflocusts>
The best human players have been playing counterintuitive moves since before Morphy's time.
<we have absolutely no idea>. Well put, in defining how YOU have no idea what computers will be capable of in the future. If the greatest thing a program can do is make no mistakes in chess then it is truly a waste of creative effort.
There is a way, sort of, to make a perfect program: upload every single possible permutation of piece position on the board and program what the next move should be.
But if you're looking for a computer with an evaluation function, then, no, nothing will be able to defeat a Kasparov, Karpov, Fischer or Capablanca on the top of their accurate, intuitive and counterintuitive form.
I have yet to see proof that any chess program evaluates every position perfectly. Assuming the programs are kept to a manageable size, I don't see that it happening...
|Mar-01-09|| ||swarmoflocusts: <<we have absolutely no idea>. Well put, in defining how YOU have no idea what computers will be capable of in the future.>|
Wait, so you DO know?
<But if you're looking for a computer with an evaluation function, then, no, nothing will be able to defeat a Kasparov, Karpov, Fischer or Capablanca on the top of their accurate, intuitive and counterintuitive form.>
Well, why not? This is speculation. You CANNOT deal in absolutes.
<I have yet to see proof that any chess program evaluates every position perfectly.>
That's 'cuz there isn't a program that can do that yet.
|Mar-01-09|| ||Brown: <swarmoflocusts>
I feel like a I have an idea of what computers are capable of in the chess world. This is why I emphasized, and disagreed with, your use of "we."
I can deal in absolutes if I wish. In fact, it seems I already have. For you to claim that I cannot deal in absolutes is to try (and fail) to force an absolute upon me.
<That's 'cuz there isn't a program that can do that yet>... yep.
|Mar-03-09|| ||swarmoflocusts: <I feel like a I have an idea of what computers are capable of in the chess world.>|
Very well. In that case, I have some questions for you (sincerely). Please: What is the highest rating a computer will ever achieve? How deep will computers be able to calculate by the time the Earth collides with the Sun (or all the humans kill each other, which appears much more likely to happen first)? How good will their positional analysis be? Why won't we eventually have computers capable of playing positional brilliancies on every other move? Or, if we will, how will this look at all similar to chess as it is played today? We have had only one Fischer, only one Kasparov, and they played brilliancies only rarely. A computer able to play them REGULARLY...!!
<I can deal in absolutes if I wish. In fact, it seems I already have.>
You CAN, but unless you back them up with some evidence we're going to have a very difficult discussion here. I am enjoying our discussion and I do not wish for it to become difficult; therefore, please humor me (or, I suppose, supply evidence, which might be difficult, as our discussion is entirely speculation).
<For you to claim that I cannot deal in absolutes is to try (and fail) to force an absolute upon me.>
The agnostic: "The only absolute is that we don't know anything for sure". I am not agnostic (well, if you look at it from a logical perspective, we're all agnostics, in that no one really knows, or at least those who do can't tell us, but this isn't the place for this discussion -- or perhaps it is?!); this does not mean I can't steal their arguments. How can you say that you DO know FOR SURE what will be in the future? Also, that you assume that my forcing an absolute upon you will fail wounds me. Let's give it a chance first.
|Mar-04-09|| ||Brown: Please... "I have an idea" is not "I know for sure."|
I disagree with you. Seems pretty obvious. We don't need to get too philosophical about it.
You need evidence, not me. I'm not claiming anything that is not currently regarded as fact right now. You are imagining a future, a future I don't believe in. Moreover, it's a future that won't ultimately mean much. It's CHESS, a game for pleasure and, for some, competition. Computers have not and will not change this.
A chess game does not avail itself to brilliancies "regularly," positional or otherwise. There is no evidence for it. There are positions that don't allow for it.
I don't think you are exaggerating the power of computers (as I once thought you were), rather than exaggerating the potential of chess. Yes, it's a great game, lots of possibilities but brilliancies at every move, every other move, etc... are not going to be found by computers. Any improvements now will come from humans using comps, or humans by themselves, well into the middlegame.
I guess what I am ultimately trying to say is...many humans are not so far away from mastering the game, and some have played as well as any uber-alien mind could. There is not much room for some human-made program to go...
|Mar-25-12|| ||Everett: <Addendum to my above post... White may look to improve his knight earlier on, say with 28.Nh4 followed by f4 and Nf3, or hop to Nf5 if black allows it. White is not worse, it seems. The a1 Rook is still verboten for black, and black's rooks are in danger of being out of play on the Q-side.>|
This seems to be correct. White must work to get his Nf3 and Bg2 into the game, and the only place he has play is the K-side. 29..Bxa1 was a blunder, and Smirin uses his control of the c-file and advanced h-pawn to good effect.
|Sep-30-12|| ||Naniwazu: GM Robin Smith has analyzed this game in his book 'Modern Chess Analysis' (Gambit, 2004, pp. 98-100). Even though he does repeat alot of the annotations above he gives a more objective view. For example to the move 15...Nb6! he writes <Tiger's move is actually quite good, as we will see> and 16...axb6! <There was a time when computers would almost invariably avoid double and/or isolated pawns. But here the open a-file gives Black's pieces excellent play.>|