Mig 
Greengard's ChessNinja.com

Kramnik-Fritz 06 g4

| Permalink | 128 comments

Today is game four of this six-game match between the world champ and the silicon kid. Deep Fritz leads 2-1, although Kramnik controlled play in the first two games and wasn't in any real danger in the third. The difference, as usual, and in extremis here, is human blunders. Kramnik has black today and in the final sixth game. The live game will be here on the official site, and with commentary at the ICC and Playchess. Start time is 1500 local, 9am EST, 6am PST.

That's also 9am in the Bahamas, where I'll be headed by the time this game starts. I tilt my Kalik in your general direction.

128 Comments

This is going to be another Slav. People only play the Slav today.

"This is going to be another Slav. People only play the Slav today."

Except when they play the Catalan.

Fritz will not play the Catalan ;)

This is a great opportunity for Fritz to play 1.e4, forcing Kramnik to either play for a win with 1. ... c5, or agree to a tacit draw with his usual peaceful Petroff or Berlin Defence.

Topalov should have given him the same treatment when he was ahead in the score and Kramnik had to play for a win...

Not sure Topalov was ever both in the lead and playing White, was he?

It's entirely possible to beat the machine with the Berlin. Probably, even DF10 doesn't understand the Berlin endgame very well.

Well, I lose my money. I'd have bet for sure on a Berlin rather than a Petroff against the machine. Is this sort of open position really a good idea?

I suppose Kramnik has said that he thinks anti-computer strategies are dead and you just have to play what you think are the very best openings against them....


Kramnik offered to trade Qs, just another opening gift to Kramnik by Chessbase or a cruel joke to mock Kramnik ?
13.Qg4 is still book of Fritz F-K (118min-99m)

I was just thinking of you; Ovidiu - looks like a corker of a game for you today. Curious opening choice from the programmers.

IM Irina Krush is giving audio commentary on WCN.

Larry Christiansen scheduled for game 5 on Sunday

Position looks painly drwaish after 14 moves.

Unfortunate draw for Irina K! You wouldn't want to spend four hours commenting on this position. Looking on the bright side, no-one'll be listening.

At least you get paid for it.

What is the URL for Seirawan's commentary? The official site gives the wrong URL.

Thanks,
Kapalik

What's the URL for Irina's commentary?

Kramnick has played 17....Kf8 and I bet, given the choice, only a few would rather play the black side.

Remember Kasparov going on about a middlegame strategy, splitting the board in half - Fritz's Re1 is doing this here.

Looks like Kramnik made some bad moves. His position really gets worse.

Black's got some problems, no?

24 rd8+ followed by 25 Rd7 looks good

Re8+ then Rxh8 is better I think

Looks quite nasty, doesn't it, after Rae1. Can White do anything in particular after ...Rd7? That looks forced to me - or is ...Ng7 an alternative, preparing ...Re8 and meeting Re7 with ...Ne6?

Re8+ and Re7 looks like it wins a pawn and
at least keeps the 7th rank

I find these games very interesting. Kramnik shows how to defend slightly worse endgames. Somehow he seems to strive for slightly dubious endgame positions, exactly where he can show true mastery in mastering them.

h3. I didn't see that one,
gives a luft for the king
but still gives time for Re8+
and Re7?

24.h3 ... already in zugzwang ?

Just not ..Ng7 please :)

Does it, tg? ...Rd7; Re8+ Kg7; Re7 Rhd8 and now if Bxf7?? then Rxe7; Rxe7 Rd1 mate, and otherwise Black's next move is ...Kf8.

But it looks wrong to spend a tempo forcing the king to g7 and freeing the rook if you don't have to. Maybe 23...Rd7 just 24 g3 or something?

rdh, yeah excuse me,
Black does seem OK

24...Ng7 25.g4 h5 26.f3 f5 27.Kg2

Maybe the pawns on a4 and b4 will tell for black later on! Kramnik lured them to those squares

Kapalik, http://wcn.tentonhammer.com to download playing software, http://audio.worldchessnetwork.com:8000 to hear the audio commentary (access is not restricted)

Irina is handling broadcast pretty well, she never gets bored

You don't quite mean that fogof - the rook's en prise after 26...f5. But yeah, predicting 25 g4: I guess that was why h3 and not g3. Kramnik'll do very well to get out of this. Bet he wishes now he'd played the Berlin.

I thimk, now Kramnik is trying to win that game. He wanted to provoke g4 with Ng7

25 Re5 surprises me, now ..Nf5?

I'm rooting for Kramnik here!
Pawnds a4 and b4 will tell!

Space being good vs overexpanding being bad.
Kramnik would be the one to know the
subtle differences, for sure. But somehow
I don't think he wished for his present position.

I'm rooting for Kramnik now
Pawns a4 and b4 will tell

Me too, tg, very much. Knight seems to have a square on d6 all of a sudden instead of being ludicrously fianchettoed on g7.

Danger seems to have passed; looks drawish again

NOW it plays g4 - hasn't the horse bolted?

I suppose it's still possible the knight could find itself a bit cramped; it doesn't have any terribly great squares after c4-c5. But then Gerhard's dream might come into play along the fourth rank one day....

Black can't do much, so I guess
the computer is just going to keep
on increasing space. I admit I
don't know if that's good or bad
here.

Bit reminiscent of the famous Fischer-Taimanov game. Maybe the machine's played really well; it's nicely coordinated now. But perhaps Black can play ...Nc8 and ....Rd4 - keep Gerhard happy!

Once again a model game from Kramnik

Thanks Freitag and Goran for the commentary links.

Game looks interesting. Good recovery by Kramnik.

Kapalik

Yeah, rdh, i'm still hoping Kramnik will punish those pawn moves

blind machine. First it seemed to have a plan with doubling the rooks, but now it shows it was just random.

32 ... Kf8 33 Kf3

Kramnik's king goes back, Deep Fritz's king goes forward.

Fritz in time trouble now. Only fifteen minutes for five moves.

You can see the beads of molten silicon rolling down the side of the box....

.....gxf4! I find that a very surprising move. Here was I thinking much of the purpose of ...g5 was to secure an outpost on f4 for the knight. Anyone got an explanation? I have a nasty feeling that if we understood the first thing about what was going on this would be a very instructive game...

Seirawan is saying that 36 ... g5?! was problematic for Kramnik, he preferred 36 ... Ne7, activating the offside knight.

What happened to Kramnik punishing Deep Fritz's pawn pushes? There is a -fine- line between gaining space and overextending. Looks like Deep Fritz gained space.

Seirawan fearful of 42 Ke4 but Deep Fritz has just played 42 Be4, which he calls "not a bad move at all".

This is exactly the type of endgame in which a bishop shines over a knight: pawns on both sides of the board and the knight not dominantly placed in the center.

Anyone have a Rybka eval for the position after 42 Be4? Not a fun position to play for Black.

Rybka 2.1c suggests 42... Be6 or 42...b6 with very slight advantage for White.

ok I meant 42...Ke6 there

what's the approximate depth of fritz's search at this point (move 44)?

Black can now play f5. There's no way White can make real progress in this position.

nine more moves and Kram has the right to adjourn the game.

Is the bishop really giving Fritz that much of an advantage? It's behind enemy lines and the pawn structure is just now getting into a position where white squares can be exploited. Most of Fritz's advantage (which is not large) is due to superior pawn structure. Am I wrong?

The game is drawn. Between grandmasters the draw would already have been fixed; but Fritz plays on because he does not know how to offer draw.

Seirawan mentioned that Fritz is not allowed to make the first draw offer - Kramnik must do that.

torture chess for the audience

Oh well, Deep Fritz was better the whole game but not by enough to win.

I enjoy endgames and endgame studies. Fritz played well today, but I think Kramnik played better overall. As we know, it's difficult to defend with knight against bishop with pawns on both sides . Not only that, but how many of us in a serious game would would isolate all our pawns against a better placed bishop and King. Cudos to Kramnik, it was an instructive endgame. Show me a chess player who doesn't enjoy a well played endgame and I will show you a chess diletante.

Mig, by the way, Fritz played 1.e4.

I enjoy endgames and endgame studies. Fritz played well today, but I think Kramnik played better overall.

-- Posted by: chesstraveler at December 1, 2006 15:07

Kramnik played better overall, which is why he had to defend the whole game? Huh?

Excellent game today, pretty bad that it was played, two more as this one and people are free to watch Topa-Polgar ( which I intend to follow blindfold if there is audio broadcasting).

Back to Kramnik, he spotted another endgame on the openings screen and off he was to play it before the screen was turned away from him.
On the birght side... now we know that comps are equal to GMs in technical endings, they even press for a win but there just isn't enough material to play with in such postions, as Leko and Kramink have proved many times, too many times.
So.. draw today, draw yesterday and draw tomorrow.

Notyetagm,

That's my point, it's more difficult to defend than attack. Yes, Fritz had the advantage throughout most of the game today, but Kramnik held the position and drew in an instructive manner. I did indicate in my previous post that "Fritz played well today." I was mostly focusing on the endgame play of Kramnik. I hope you can discern the difference.

The comp understood that it had to trade the last rook...
Kramnik tried in vain to win that endgame, but the chances for that weren't really great. Deep Fritz kept things together...
Kramnik played very well (Seirawan, Jussupov, Pfleger admitted that, amongst others): Out of a seemingly very unpleasant position he managed to equalize and even to keep some chances to punish Fritz for the slightest error to come, but it never came.
I admire Kramnik for handling the knight..Good game both, thank you!

I agree that Kramnik played well, but notyetagm has it right: if he really played better he wouldn't have been on the defensive for the entire game.

The statement "It's more difficult to defend than attack" is misleading. It is more difficult to defend than attack when the opponent has a strong attack. On the other hand, in the situation like today where the advantage is minimal and no clear avenue of attack exists, it is far easier to defend.

>.....gxf4! I find that a very surprising move. Here was I thinking much of the purpose of ...g5 was to secure an outpost on f4 for the knight. Anyone got an explanation? >

otherwise Fritz plays f5 and the squares g6-e6-d5 are controlled so no Nf4

>I have a nasty feeling that if we understood the first thing about what was going on this would be a very instructive game...>

lol rdh, this sounds as despair, that's all hope left for this game ?

>I agree that Kramnik played well, but notyetagm has it right: if he really played better he wouldn't have been on the defensive for the entire game.>

Not really, he actually played better then Fritz.
He deliberately went for a slighly worse endgame right from the opening (Ng5). An advantage of seeing the opening book and a small price if this avoids Fritz's monster tactics.

He then slowly equalized but there wasn't anything concrete that Fritz could do (maybe Karpov but not Fritz, there were no tactics, a symmetrical pawn structure, and all postion play) so it was only a matter of time for this to happen.

Then he could have even won if Fritz's pawn-play had been very bad and it would have allowed Kramnik to create strong squares for the knight.

Well guys, we might very well be at an impasse on this one. So as Yogi would say: "When you come to a fork in the road, take it." ;)

can somebody explain me why Frizt played Bb5? a move without any logic. this is a match between human number 1 and computer number 5 or 10.

I suppose you must be right, Ovidiu, about f5. But then, with the kingside sealed like that, can Black really have any problems? He can arrange h5xg4, then if the knight comes to f6 or even e5 he might get somewhere. Or he could just go f6 and sit; surely he could defend with play only on the queenside?

It is possible that it's just a draw either way, of course.

Bb5 seemed quite sensible to me - surely it's desirable to provoke ....c6, isn't it?

Ovidiu,

Nice comment (at 16:54) re Kramnik "outplaying" DF10 by accepting manageable weaknesses in order to obtain a simplified position.

Maybe the flipside of this strategy would be for DF10 to accept manageable weaknesses in order to obtain a hairy, highly tactical position.

Interesting comment, Greg. But I think (and, indeed, hope) that programs are still too "stupid" to do that. While they do understand "compensation" now, they are not capable of doing anything speculative.
I may be wrong, of course. (As I often am.)

rdh, it may be draw this way too, actually even win if Fritz plays weak after f5 and Kramnik comes with Nd6-e8-f6-d7-e5 but it is risky since now Fritz continues with his own ideas by placing Bf3 defending g4 and attacking Qs with a6

for instance 1..Nd6 2.f5 Ne8 3.Be4 Nf6 4.Bf4 Nd7 ( with the idea f6/Ne5 )5. a6 and it becomes unclear

for certain Kramnik saw this and and played safely the plan with Ne7 (defending now c6) and b6 and neutralizing Q-side. Typical Kramnik.


>
Maybe the flipside of this strategy would be for DF10 to accept manageable weaknesses in order to obtain a hairy, highly tactical position. >

It already did something like this in the 3rd game when it sacrified a pawn for activiy.

rdh why would you provoke c6? bd3 is better.

>>Not really, he actually played better then Fritz.
He deliberately went for a slighly worse endgame right from the opening (Ng5). An advantage of seeing the opening book and a small price if this avoids Fritz's monster tactics.>He then slowly equalized but there wasn't anything concrete that Fritz could do (maybe Karpov but not Fritz, there were no tactics, a symmetrical pawn structure, and all postion play) so it was only a matter of time for this to happen.

Then he could have even won if Fritz's pawn-play had been very bad and it would have allowed Kramnik to create strong squares for the knight.<<

Your ability to interpret chess games based on automatically knowing who is and who is not going to make a mistake when is way beyond me. Today the outcome was a flat draw with Kramnik hardly ever having initiative. He got the kind of simplified position he often got through this against Kasparov and Leko. Perhaps, since this is the kind of position a player has an advantage on over computer, Fritz could have overplayed his hand and lost, but Kramnik was unable to get him to do so. Overall outcome? 35 more moves of non-existent positional play than it would have been against a human.

Yuriy,

I want to interpret what you are saying correctly and I am not trying to be cute here, but would you please define "non-existent positional play" for me?

I'd appreciate it if someone would explain why Kramnik played ...Ba6 (inviting b4) instead of playing ...Bxc4 straight away. In the resulting positions, White's pawns were more threatening on a4-b4 than they would have been with the pawn back on b2.

Ovidiu--

I'd think DF10 would be programmed to do a great deal more of the sort of thing you pointed out in Game 3. Is Kramnik "taming" DF10, or is DF10 not "trying" for hairy, unbalanced positions?

The chessbase commentary for game 4 adds practically no analysis value whatsoever. There's no indication of where white might have gone wrong - not even a suggestion, and this was an interesting game. White is much better and then suddenly Black equalizes?!

Mail it in,
I think that Chessbase is trying to sell Fritzes - and so they want everyone to believe that Fritz brilliantly forced an advantageous game from Kramnik. The reality is that Kramnik directed the game to an inferior ending so that he could remove pieces. Even if Kramnik was worse, he was completely in control - an interesting thought. =)

@greg

If Fritz were let alone he would try but he is bound the follow the opening book at the beginning
and, as Kramnik himself wisely noticed after Bahrain, the fate is decided in the opening.

I support Mig's position that in this match Kramnik gets to pick the openings, i.e. he can always steer the game toward postions where comps' strenghts are disabled.
True however that such technical positions are exactly what Kramnik likes to play anyway but he doesn't always get them against other GMs and nowhere as easy as he has got them up to now in this match.

This 4th game was paractically "over" at move 13, the rest was pretty much routine.
Uninteresting game-wise but of course instructive if what you want is to learn/understand chess in general.

To your question : Fritz is already "tamed" from the start by the condition imposed by Kramnik to be let to see the opening book during the game. With his encyclopedic and in depth knowledge/preparation of the openings he can always get what he wants. Except to some extent in the game 3 we are not given a chance to see how Fritz' really plays like.

Chessbase has staged a bit of fraud with this match but quite likely otherwise Kramnik would have refused to play --( why just get humiliated as Adamas ? he isn't a masochist)-- and all this big advertisment show (which is useful to all involved) wouldn't have been held.

Thus we get what we get. It is still interesting after you get over the frustration that it could have been different, more, and "for real".

Such is life, the issue is not "what is good and what is evil". Even in a world where everything is evil the lesser one is, by necessity, a good thing.

Ovidiu,

I'm trying to understand how Kramnik's knowledge of DF10's opening preferences allows him to dictate a peaceable, stable opening.

If, after 1 e4 e5, Kramnik knows that DF10 prefers 2. Nf3, then Kramnik can play the Russian defense, as he could against any human.

But if Kramnik knows that DF10's preference is 2. f4, and Kramnik dislikes the King's Gambit, then perhaps Kramnik plays 1...c6 instead.

DF10's only counterstrategy would be to make all its opening preferences as sharp as possible.

I think I'm getting it. Topalov could surprise Kramnik with sharp, but somewhat inferior openings that Kramnik would have to "solve" over the board. But if Kramnik has DF10's opening book in advance(?), and can see that DF10 has a preference for a sharp but inferior line, he'd be able to prepare for it and stomp it.

@greg

What Kramnik sees is only what moves Fritz expects from him according to its book, not what Fritz will reply to them.
However this matters little because Kramnik knows, of course, very well the theory of his openings and where each branch of the tree leads to.
So he picks up having in mind the final position.

In effect he can choose out of Fritz' opening book any of its final postions which he finds more favourable (Q-less, simplified, rigid pawn structure ), as he did today.

The only problem possible along this robotic path to an endgame would be a novelty from chessbase inserted in these known as leading to simplifications lines.

It hasn't happened at all yet, actually today was just a perfect example of this trick working smoothly.

A questions:

(a) Is there any indication on how frequently Kramnik goes to the bathroom during this match?...

(b) Is there any relation between the way Chessbase commentators were "objective" during the Kramnik-Topalov match and the way the analysis of these four games has been presented so far in Chessbase?

I have seen how the machine is praised in such an uninteresting draw that Kramnik did not have problems to equalize; sentences like: "tough draw" ... "Things are not looking good for Kramnik"...

(c) The fact that the Kramnik-Deep Fritz match was arranged before Kramnik-Topalov match, influenced the way Chessbase managed the information in that match? (thinking that with Kramnik's victory, DF 10 would face "the world champion" and not the loser of the match).

I don't know, but judging for the games and the lack of sharper play, I guess many people would refuse to buy that newer version of DF...



Sorry for the typo (I was going to put "just a question", at the end were "three questions", so unconsciously ended "a questions"...

Ovidiu,

What am I missing?

"The only problem possible along this robotic path to an endgame would be a novelty from chessbase in these known as leading to simplification lines."

Isn't this precisely the situation Kramnik faced against Topalov? Fear of a hidden novelty?

How would it have helped Kramnik if, before each opening move in Elista, Topalov had said, "I think there's a 70% chance you'll play "move a", a 20% chance you'll play "move b" and a 10% chance you'll play "move c"?

Chesstraveler,

Thank you for asking:

What I meant by "non-existent" is unlikely to lead to an accomplishment, by "positional" I meant roughly the opposite of tactical and by "play", that pieces to continued to move. In other words, in a positional setting, Fritz continued trying to win, even though draw was by far the most likely outcome of the moves he made.

I have to agree with Yuriy about the comments over-hyping Kramnik play...

Kramnik just got in an slightly inferior technical endgame. He and Leko would do this with Black, any day, against any GM, and draw this. Nothing new. During the whole game he got zero chances to win, and could have lost if he slipped just a little bit (which, like for Leko, nearly never happens).

NOW, if he was playing Kasparov/Topalov/Anand in a 6 games match and was down 2-1, and just tradded all pieces, people would not praise him for "directing the game to an inferior ending so that he could remove pieces. Even if Kramnik was worse, he was completely in control".

They would call it sheer stupidity or at least cowardice, and people would praised Kasparov/Topalov/Anand for torturing Kramnik for a few hours, without having any risk.

It's stupid, because, as of today, Kramnik has nearly zero practical chances to win the match = he would have to win the 2 last games. Given the match conditions, this marks the point where humanity has surrendered to the machine.

>Isn't this precisely the situation Kramnik faced against Topalov? Fear of a hidden novelty?>

not precisely.

As far as potential novelties prepared by CBase go it is precisely the same. Actually Bxd5 played by Fritz in game 3 was a novelty (according to CB-website) and maybe that is why that game saw Fritz playing something.

But novelties aside (and they could not be that many when contrasted with the theory of openings in general) Kramnik did not know what were the drawish lines that Toplaov included in his repertoire (for lack of any better moves to be sure but nonetheless) so as to go for them.

for instance, giving the an exaggerate example for the sake of simplicity, if you play against Fritz
1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 and now you stop and look a the screen to see its opening book which gives for you (he expects) A.4 Ba4 B4.Bxc6..well that is good news, it means that Fritz has the well known drawish Bxc6 in his book.

Let's say that he would have not have it and thus the screen would have shown only 4.Ba4 (100%), well then you do NOT play Bxc6 ! Fritz would get out of his book and begin to play by itself and instead of the main line
4..dxc6 5.d4 exd4 6.Qxd4 Qxd4 7.Nxd4 draw he may have come out as well with 4..bxc6 which is not as bad as it seems and leads to a complex game.

Thus you now go on with 5. Ba4 and wait for the next moment and again look which of the moves that Fritz has in his book for you know that they lead to exchanges etc. After few such choices to steer the game it becomes rather impossible not to get in the end what you want.

Kramnik knows the theory and the kind of end postion which that they preach for the most moves. They are the present opening theory and implicitly Fritz's selection for it ("Fritz's book") is a subset of this theory.
But he still needs to be certain that Fritz will go for one to them rather than jump out of the theory book, start calculating, and playing by his own some weird move and blow up everything in a complex middle games.

That is why he came with this weird rule to be allowed to see the screen, he can infer rather easily what are the drawish(or difficult for comps to handle) lines present in Fritz' book and make sure he gets one.

>NOW, if he was playing Kasparov/Topalov/Anand

He isn't playing Kasparov/Topalov/Ananad he is playing something stronger than what mauled Adams with 5.5-0.5

>It's stupid, because, as of today...


No man, he does the best, plays the best possible strategy, to cut down the loses.
If he were to play otherwise not only that he wouldn't win the match but he would lose it crushingly as Adams did.

"He isn't playing Kasparov/Topalov/Ananad he is playing something stronger than what mauled Adams with 5.5-0.5"

From 1996 to 1999, Kasparov record against Adams was: +6 1= -0, that is 6.5-0.5
Of course in 1999 only, Adams reached 2720, but the point is Adams is not Kasparov.

Plus Fritz drew its previous matches with Kasparov and Kramnik (2002, and 2003).

Kramnik would be very afraid to play Hydra. Not to mention the latest version of Rybka! lol

the idea that Fritz is stronger than Hydra is just totaly absurd. Rybka is also some 100 elo stronger that Fritz who for example threw away a good advantage in game 4 by playing the stupid move Bb5.

Zarghev,

yes and it would been way more interesting if Kramnik would have chosen to get out from the shelter of simplifyied position and instead go for wins.

but Kramnik takes into accounts his own limits and weakneses. He is no Kasparov in "fighting" postions. He is booksih, knowledgeable, technical and, rationally, tries to get into situations where he can play out his strenghts.

A very rational person rather than an 'artist' as he fancies himself.

The comments of zarghev and, perhaps ovidiu, assume that Kramnik's best chance for a win is to "get out from under the shelter of a simplified position."

But Kramnik may reasonably believe that his best shot at winning is to achieve a Game Four-type bishop v. knight setup, hoping that DF10 won't see far enough ahead and will put its pawns on the wrong-colored squares.

Anyone suggesting that this approach is "cowardly" or "stupid" ought to share with us his superior strategy.

> Fritz who for example threw away a good advantage in game 4 by playing the stupid move Bb5>

that was a quite good move forcing Kramnik to move pawns (c6) on the side of the board where Fritz was attacking (b5 intended)-- thing which increases the speed of attack and gives it targets.

Actually it is impressive because it means that the idea "manouver first to force you opponent to create weaknesses and only afterwards attack" has been somehow successfuly translated in numbers and encoded in Fritz.

Kramnik would have rather not moved that pawn.
After c6 Fritz's possible pawns assault on Q-side
with b5 had to monitored continously until the very end ( for instance if instead of 1... gxf4
1..Nd6 2.f5 Ne8 3.Be4 Nf6 4.a6[3..Kd6 4.c5+ Ke7 5.b5]).

Ovidu what do you mean artist as he fancies himself? In interview after interview Kramnik has presented himself for years as a logical rational above all professional player who does what he needs to win. His strategy here nearly worked he could have won the first game and drawn the rest - that was his best chance to take the million euros. I think everyone knows the top computer chess program are stronger than any GM in the world and we are fast reaching the point where its not really worth having a fritz 11, 12 etc. It remains to be seen whether there is a commercial incentive for computer programs to be developed that can beat any GM without an opening book ie just on programmed chess principles and strategy as well as huge calculating power. I hope they will try that.

Ovidiu,

Are you suggesting that you can't be both an artist and a "very rational person."

>Ovidu what do you mean artist as he fancies himself? In interview after interview Kramnik has presented himself for years as a logical rational above all professional player who does what he needs to win.>

could be Andy, I have not read all his interviews but as I remember those that I read he kept trying to associate his playing with art, painting, creativity and so on. It was true for Tal and Alekhine but for Kramnik's Petroff and Berlin ?
Maybe was just another "public relation" move but it seemed to me from how he wrote about it that he really wanted( believed) such thing to be true.
People have all sorts of delusions about themselves.

for instance from his website :

21.12.2005 "For us chess players the language of artist is something natural," says classical chess world champion Vladimir Kramnik in this indepth interview with German artist Ugo Dossi. Their indepth dialog probes the creative processes that take place in each field, and the intersection points between chess and art."

http://www.kramnik.com/eng/interviews/getinterview.aspx?id=75

Ovidiu,

For you, evidently, art must be exciting.

I prefer the excitement of ACDC to the precision of Bach. But I'd acknowledge Bach as the greater artist.

Love your Bach metaphor, Greg! In fact, some of his ultimate works (The Art of Fugue has Art in its title!) are not very accessible to most mortals.

As with Bach, one good measure of greatness is what people will think 300 years later...

"I think everyone knows the top computer chess program are stronger than any GM in the world and we are fast reaching the point where its not really worth having a fritz 11, 12 etc."

But remember that Fritz is also marketed as a chess trainer with many useful playing modes, an interface to databases, and so on; it's not just an engine. I've used Fritzes 8 and 9 and the improvement in the engine's strength from 8 to 9 hardly could matter to a weak player like me, but the improvement in Fritz's interface and its additional features made the upgrade worth it. I don't know about Fritz 10, though.

"It remains to be seen whether there is a commercial incentive for computer programs to be developed that can beat any GM without an opening book ie just on programmed chess principles and strategy as well as huge calculating power."

Perhaps not a *commercial* incentive but there are a ton of individual ventures out there, many of which are producing engines of great strength. I have no doubt that some competent hobbyist will tackle a project such as you describe.

I do not own a recent version of Fritz and what caught my eye in the chessbase reports was the high quality of Fritz's 3D graphics.
Has anyone used it ? What is your experience ?

Yuryi,


Fischer was once asked to give the reasons for his success at chess. His reply was: "Simple. First, Learn the chess openings, really learn the one's you play for white and black. Second, keep the pressure on them every second, THEY WILL CRACK. No exception, play every game to the end."

In my mind, that sounds how most chess computers are progammed to play and may be the reason for the "non-existent positional play" of the last game. Only thing is, Kramnik didn't crack.


how about this one
"Chess is a matter of delicate judgment, knowing when to punch and how to duck."—Bobby Fischer

Zarghev,

I was not responding to any sort of comments overhyping Kramnik's play--didn't see too many of those. I simply disagreed that he played better than Fritz.

Chesstraveler,

Valid analysis, but I am not sure Fritz put much pressure on Kramnik... He had a slight advantage in space, but all of Kramnik's position was well-defended. It wasn't really much of a "defender cracks under pressure" position which is why I called the play non-existent. The original statement which I disagreed with was "Kramnik played better than Fritz"--what you are saying doesn't seem to disagree with me, though I may be wrong.

I've only tried Fritz's 3D interface a couple of times. It's not bad, certainly much better than many, but I've never liked playing computer chess in a 3D view. Maybe it's just long habit. Is it really a better idea, though, because it more closely approximates playing with a real board? Not just for aesthetic reasons but because you'll perhaps better learn to "see" things on the real board by playing the computer in a mode that emulates it. It is, I think, easier to "see" the whole board in 2D mode.

A minor nit: whenever setting up a chess board I always point the knights to the side. I don't like to see them pointed forward, as Fritz does.

Zarghev,

I was not responding to any sort of comments overhyping Kramnik's play--didn't see too many of those. I simply disagreed that he played better than Fritz.

Chesstraveler,

Valid analysis, but I am not sure Fritz put much pressure on Kramnik... He had a slight advantage in space, but all of Kramnik's position was well-defended. It wasn't really much of a "defender cracks under pressure" position which is why I called the play non-existent. The original statement which I disagreed with was "Kramnik played better than Fritz"--what you are saying doesn't seem to disagree with me, though I may be wrong.

In the arrangement of the match between both parts, Is the final result supposed to be a tie 3-3? So far, what seems unlikely is a way in which Kramnik could win a game with such a conservative play, but I am not sure if the parties agreed a result of 3.5-2.5 in favour of Fritz.

On the other hand, the match can be seen as a victory of the machine in the sense that the human player tries to play "anti-computer" strategies and simplifications, assuming that in another situation the computer would win. So far, the conclusion is that Kramnik might be able to draw against the machine, but has no intentions to win ... the question of whether a human player could DEFEAT the machine in a match remains unanswered. If strong players, for example can prepare strong theoretical novelties and execute in the board to win games, or using deep strategic ideas or speculative tactics to win... is it possible to defeat a computer like DF in that way?

I think that there is excessive publicity and also excessive respect for some players to a program (computer player) whose strength is probably overrated. First, DF is not the strongest chess program in the world and second... if those matches are just for money, why not try to test the computer even if you lose? As far I have seen, is not a very nice use of the title of WC from Mr Kramnik (at least for the fans, but I doubt he cares about it with the money he is receiving)


all theory, Kramnik's 17.h4 novelty (17.Bf6 1/2 Geller-Spassky sf2 1965)

Ovidiu's ever-mutating theory as to how Kramnik can 'choose the openings' seems now to come to this; he has the advantage of being able to prevent the computer from finding itself out of its opening book.

It would seem to follow that the programmers would do well to unplug the opening book altogether.

Strange that the programmers don't seem to have cottoned on to this strategy.

rdh, it is exactly the same idea, why "ever mutating" ? ...rhetoric instead or arguments ?

There's no need for arguments; you're right. If Kramnik perceives it as an advantage that the line he's prepared turns out not to be in his opponent's preparation, he can indeed instead play some move that is in his opponent's preparation. Personally I doubt whether that's quite as useful as you imagine. And your comments that aiming for quiet technical positions is the only way to go because that way he might score a few draws doesn't seem to take into account how close the match has been so far, does it?

I'm more interested in today's game frankly, since for some reason my machine won't do the live commentary. Or is Fritz just about to repeat?

Sorry, would help if I typed what I meant. 'If Kramnik perceives it as a DISadvantage....'

it will repeat

>And your comments that aiming for quiet technical positions is the only way to go because that way he might score a few draws doesn't seem to take into account how close the match has been so far, does it?>

It has been close but then what this match proves is not that Kramnik can play in equal terms against Fritz but that he can still hold against him in simple postions, with difficulty but nonetheless.

If anyting it acknowledges that Fritz is stronger since the rules had to be bent so as to make sure that these will happen, and the match be possible (and get the benfits of advertisement).

My only issue that it should not be classfied as being in the same class as the former matches.

If you want CB could have organized something like "poor chess" instead of standard ( let alone FRC) chess. say "king, 5 pawns, a pair of rooks and 1 bishop in the starting positon" GM vs. Fritz match.

Today's Chessbase article calls the 5th game "fighting chess"; I'm not sure if I can agree with that. What I saw was a lot of early exchanges to reach an early middle game and then draw. And it wasn't all Kramnik's doing this time. I was a little surprised by the way Fritz played.

My apologies, should have read early endgame instead of middlegame.

>Today's Chessbase article calls the 5th game "fighting chess"; I'm not sure if I can agree with that.>

What else could one expect ?
Kramnik has become Chessbase's poster boy. This match is an incestuous family affair, and the products of such the relationships are usually early abortions.
So much about "match" and being on an adversarial postions.

Today's game saw Kramnik's choice for trading Qs and getting his endgame being a Gheller-Spassky 1965 game with 17.h4 ! novelty (improving over Gheller's 17.Bxf6 gxf6 and draw 10 moves later).

Afterwards Fritz logically opened Q-side ( at the cost of another piece trade ) while Kramnik neatly kept the balance with his K-side counterplay and draw was the logical result. A neat 13 moves game ( 17 being theory and another 3 repetition) and a lesson on how to play to keep the balance in the game of chess.

After their recent offenses, Mel Gibson and "Kramer" apologized to anyone who'd listen.

After forcing him to a) minutely examine five abortions and b) post fifty comments about them, Freidel and Kramnik ought to be on their knees begging Ovidiu's forgiveness.

Twitter Updates

    Follow me on Twitter

     

    Archives

    About this Entry

    This page contains a single entry by Mig published on December 1, 2006 4:56 AM.

    Lawson on Fritz & Tal was the previous entry in this blog.

    Chess, What's Chess? is the next entry in this blog.

    Find recent content on the main index or look in the archives to find all content.