< Earlier Kibitzing · PAGE 2 OF 2 ·
|May-12-07|| ||realbrob: I don't want to be exaggerated, but I think this shows computers are nothing without their databases. They can't really play openings, they know them "by heart". |
Here we have a strong human player playing a weird opening and a strong computer program understanding nothing of it.
It's amazing to see how Fritz castled kingside regardless of Black's g4 and h5 pawns.. Spontaneously putting his monarch in front of the firing squad!
|Aug-28-07|| ||dabearsrock1010: 10. O-O surely leads to a lost position for white|
|Aug-28-07|| ||CapablancaFan: 1...h5!? Goldsmith Defense? My God, at first look, thought to myself that this guy must have drank a little too much kool-aid before this game to open like this. Dosen't he know Fritz has a database of hundreds and hundreds of openings. Dosen't he...wait a minute I thought, he does, and that's the point! By opening with the Goldsmith, Pecci steers the game totally out of the books by move 5! Think about it, by 5...g4! Fritz is left to it's own devices. The problem Fritz had was that it was still attempting to play "traditional" moves in a game that has nothing traditional about it. You can see Fritz struggling to figure out how to meet the kingside pawn storm. The program decides it's "best hope" is a counter-attack(22Qxa7?) when the g & h files to it's king are opened like a can of tuna? Perhaps it may have already been too late anyway. After 27...hxg2! Fritz just gave up, LOL.|
|Aug-28-07|| ||CapablancaFan: <syracrophy: <chessgames> Please, we need better games for the GOTD. The recent games (W Roach vs A D Penman, 1992, Taimanov vs Yusupov, 1980, Lasker vs H Howe, 1902) are not properly for the greatness of a GOTD. And today's game is simply pathetic (No good game by Fritz). It's just a simple advice.> What are you looking for in the GOTD? I mean, does it have to involve a sacrifice of a queen everytime? A spectacular king walk or something? Really, what's your criteria? I'm glad <chessgames> gives us a wealth of games with a slew of various aspects for us to see.|
|Aug-28-07|| ||mrbiggs: Keep in mind that Pecci played Fritz 5 and 6. This kind of stuff is much less likely to work against today's Fritz on a good computer, much less Rybka or any other stronger engine.|
|Aug-28-07|| ||RandomVisitor: 5.h3 is strong for white, with the Rybka eval being about +1.0. 11.Bxc6 is also strong for white.|
22.Qxe5 is better for white. Rybka sees that 22.Qxa7? is losing in seconds.
|Aug-28-07|| ||Phony Benoni: It's interesting how players are so afraid of getting bad positions because of a computer's book knowledge will seek to avoid this by playing unusual variations which give them a bad position.|
|Aug-28-07|| ||Sneaky: The computer makes one mistake and suddenly it's strategically lost. |
All you have to do to repair White's play is to scratch 10.O-O? (castling into it, as they say) and replace with 10.Qd2 and 11.O-O-O. Now White is fine, if not better.
You could also make the argument to play 9.Qd2 and 10.O-O-O, since the 9.Bb5 Bd7 manuveur is fairly dubious on the face of it.
|Aug-28-07|| ||zaxcvd: What is Fritz's rating anyway?
The next step for chess computer programming would be to create opening moves on their own. I wonder how many grandmasters can beat a top level computer that plays without its opening book.
Pretty soon (maybe a few years) a computer will beat the world champion without an opening book.
Unfortunately for chess, opening moves have been exhausted. It may be time to move on, away from chess, to one of the many fascinating chess variants that have been created!
|Aug-28-07|| ||kevin86: The Goldstein Defense looks as good as gold. White seems to dare black to strike and he does with a vengence.|
The comp looked "on the fritz". It could have been worse-it could have been a WOPR-remember how it always begs for a game of chess?
BTW,what is the difference between a "good game of chess" and "a nice game of chess"? (for fans of WARGAMES)
|Aug-28-07|| ||Chessmensch: In addition to what's been said, white goes off chasing irrelevant queenside pawns and giving meaningless checks while his kingside/king position is crumbling. That's fiddling while Rome burns.|
|Aug-28-07|| ||RookFile: This is all nice, but today's version of Fritz is probably 10 times as powerful as the 2001 version.|
|Aug-28-07|| ||cn1ght: zaxcvd, people have been talking about the end of chess for a long time. It's not about to die, although I share your concerns.
computer programs see tactics not positional chess, or thats what I've read the last however many times I've ever checked...If the computer thinks it can win a pawn because there are no mate threats at the end of its calculating systems ya it will play stuff that to humans is considered suicide... Of course like everyone's said the programs are better now and people are trying to havee some positional factors... but as far as I can tell computers are still second best at positional, or any non-tactical chess. As to the game... I do kinda agree this one is junky but still cool to see that humans have beaten games even as late as fritz...|
|Aug-28-07|| ||Whitehat1963: <zaxcvd> and <cn1ght> (Interesting names by the way), you might enjoy reading this article about computers, chess, games and "artificial intelligence" in general:|
|Aug-28-07|| ||zaxcvd: cn1ght,
I am not cheering on the end of chess. In fact, I love chess and it is the most enjoyable game for me. However, I do not believe in studying chess, and I think far too many people waste time studying chess openings so that they can presumably increase the rating 50 to a 100 points.
2300+ rated players however must memorize excessive amounts of opening theory to stay in the game. IN my opinion, the game has lost some of its former glory for this reason. (Not that it affects my play since I am not 2300!)
I just think if you really enjoy playing chess, you should enjoy quite a lot of chess variants out there - some of which are quite 'chess-like'.
Don't forget Anand just finished 2nd in Chess960 tournament. Grandmasters are looking for alternatives too. There are quite a lot of 'chess-like' variants out there with no excessive opening theory to taint the game.
Take a look at http://www.omegachess.com - its endorsed by GMs!
An interesting one on chess variants:
|Aug-28-07|| ||xKinGKooLx: This opening sort of reminds me of the Borg defense by move 6. White's 10. O-O was a mistake because it allows Black to make the most of his advanced queenside pawns. I might try this opening sometime!|
|Aug-28-07|| ||r00723r0: <xKinGKooLx>, you mean kingside pawns.|
|Aug-28-07|| ||xKinGKooLx: <r00723r0: you mean kingside pawns.>|
D'oh, yes, of course I do. Thanks.
|Aug-28-07|| ||apple pi: How did Pecci do in the whole match?
|Aug-28-07|| ||pawnofdoom: <apple pi> I assume he won. Pecci has played 22 games against different versions of fritz in this database and he won every single one. This was probably when Fritz wasn't so strong|
|Aug-28-07|| ||sfm: <zaxcvd:...I think far too many people waste time studying chess openings...>
Well, there's also something charming about bringing your knowledge to the battle. Studying is a part of most things in life, so why not chess. It is all about making better moves, however we can learn to do it.|
At the same time - I agree! I actually like Fischerrandom where you are forced to think from your very first move - just like it was when you were a chess beginner.
But should it get wildly more popular theory would develop here as well.
|Aug-28-07|| ||sfm: <10. O-O surely leads to a lost position for white...>
That is maybe too much to say, but as many suggest, it looks more safe with Qd2 (maybe Qe2 is better) and 0-0-0.
What really screws up the game for the computer is that it does not know a golden rule: Closing the centre if your opponent is further ahead with attack on a wing than you are is usually a bad idea. 11.d5, followed up by planless play Qd1-c1-d1 and a4, losing 3 moves. It's over.|
I have beaten some computers by:
(a) managing to close the centre, reducing tactics where they are smarter than me
(b) castling the opposite side and hoping they didn't realize that there's no time to waste.
It works - sometimes...
|Dec-28-07|| ||timhortons: we dont know what version of fritz these is....and how strong is the computer that runs it......but once you got it by the horn it plays like a broken record...im sure these guy loss by hundreds of hundreds game to fritz...its fun playing with computers...very much different is the approach in playing with them|
|Dec-09-08|| ||spreadsanity: I stepped through this game in Shredder 11, and it thought that 10.O-O wasn't the greatest idea, and that 22.Qxa7 was a very bad move (it preferred exchanging queens on move 22). It would be interesting to see how Pecci's tactics work against the current generation of chess computers.|
|Aug-10-15|| ||Zhbugnoimt: Pecci would lose to any modern engine within 25 moves.|
< Earlier Kibitzing · PAGE 2 OF 2 ·