Deep Fritz(C) – GM Kramnik [B86]
12.05.2006
Click here to replay the game.
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 (Kramnik decided to play the super sharp Najdorf. This means that Kramnik at least is trying to go for the win to even the match.)
6.Bc4 (Fritz responded with Bobby Fischer’s favorite Bc4 variation.)
6…e6 7.0-0 Be7 8.Bb3 Qc7 9.Re1 Nc6 10.Re3!?!?!?!? (THIS is a computer move 🙂 I have not seen this move before. GM Rustemov said: “Kramnik should be laughing at the board.” I would love to see Kramnik’s face after Fritz made that move.)
10…O-O (A logical reply.)
11.Rg3 (This is Fritz being Fritz. Tactically speaking, there is nothing wrong with it. However, moves like these are simply NOT human.)
11…Kh8 (Another logical reply to get out of the pin.)
12.Nxc6 (Another surprising move to me. 12.Be3 or 12.Bg5 are both playable.)
12…bxc6 (Qxc6 is also playable but this is probably a little better.)
13.Qe2 (One possible idea is Bf4, followed by e5. Black cannot allow White to push e5.)
13…a5 14.Bg5 (White is playing agressively but Black is OK so far. Black is not in any serious danger.)
14…Ba6 15.Qf3 (This is one of those positions where Black has to be careful about the Kingside attack or the e5 threat. Other than that, Black is doing OK by counter attacking on the Queenside.)
15…Rb8 16.Re1 c5 (The threat is c4 to isolate the Bishop on b3. White now has to defend the pawn on b2 with Bc1 or Rb1.)
17.Bf4 (These are some posibilities: 17…c4 18. e5 dxe5 19. Bxe5 Qb7 20. Bxb8 Qxf3 21. Rxf3 Rxb8 22. Ba4 Rxb2 += or 17… Nd7 18. Qg4 Rg8 19. Bc1 Bf6 += or a simple 17…Qb7 as suggested by GM Kiriakov on ICC is also fine.)
17…Qb7 18.Bc1 (As a human, I evaluate this position as equal. However, the computer seems to like White a bit better.)
18…Ng8 (The idea of this move is to put the Bishop on f6 which is a better square. If people see this game for the first time without knowing the names of the players, they must think White is an amateur with Re3 Rg3 stuff. How computer brains are so different from humans.)
19.Nb1 (The position is equal. The idea of this move is to play c3 to clear the c2 square for the Bishop on b3.)
19…Bf6 (White has 2 main choices here: 20.c3 or 20.Na3. I like the more natural 20.c3)
20.c3 g6 (I do not like this move. This is a strategical mistake and not tactical. The rule is never weaken your pawns unless you have to. This is a very important rule as Black defending with the Sicilian. Black was doing fine until now.)
21.Na3 (Kramnik was doing fine until 20…g6. Now, I do like White better. It is still not lost but White has a more comfortable position with a potential attack on the Kingside.)
21…Qc6 22.Rh3 Bg7 23.Qg3 (Black’s position is quite unpleasant. I do not expect a blunder from Fritz in this position and I see no chances for Black to win unless there is a machine malfunction. 23…h5 24.Bc4 Nf6 25.Qd3 Bb7 26.Bb5 Qb6 27.Bf4 Rd8 28.Qc2 Kg8 29.Qa4 Qc7 30.Rf3 is one possible line which was given by Rybka computer.)
23…a4 (24.Bc2 is expected. I would not want to be Black looking at the 2 Bishops, Queen and Rook aiming for my Kingside.)
24.Bc2 Rb6 (White’s idea now is to play Qh4 and then something like e5 eventually to open up the diagonals for the Bishop. e5 first is also fine. I do not like Kramnik’s position at all.)
25.e5 dxe5 26.Rxe5 (26… Bxe5 27. Qxe5+ f6 28. Rxh7+ Kxh7 29. Qh5+ Kg7 30. Qxg6+ Kh831. Qh7#)
26…Nf6 27.Qh4 Qb7 (Kramnik has about 18 minutes left to defend a very difficult position.)
28.Re1 (The problem for Black is there is no counter attacking chances on the Queenside and White’s pieces are in excellent position to attack on the Kingside.)
28…h5 29.Rf3 (Black’s position is getting worse. It would be a miracle for Kramnik to hold this position as Black with such little time.)
29…Nh7 (This is basically over. Black will lose the a pawn and the threat on the Kingside is still there. As I mentioned above, this is an extremely hard position for ANY human to defend, especially with little time left.)
30.Qxa4 Qc6 (Now Fritz can move the Queen back to h4 and maintain the attack. The difference now is White is up a pawn. As human, I prefer to trade Queen and win the endgame.)
31.Qxc6 Rxc6 32.Ba4 Rb6 33.b3 Kg8 (White is a lot better. I believe Kramnik will have a hard time drawing this game.)
34.c4 Rd8 (Black has about 10 minutes for the last 6 moves.)
35.Nb5 Bb7 36.Rfe3 Bh6 (I do not understand this move. 36…Nf6 is more preferrable to me.)
37.Re5 Bxc1 38.Rxc1 Rc6 (This is why 36…Bh6 does not make sense. Black is making White’s position easier.)
39.Nc3 Rc7 40.Bb5 Nf8 +- (Black has a number of problems. One is the c5 pawn. The other is White a pawn.)
41.Na4 (Attacking the c5 weak pawn. The problem for Black is if Kramnik plays Rdc8 to defend the c5 pawn, he would be giving up the d file.)
41…Rdc8 42.Rd1 (White accomplished another task which is to take over the d file. Black has no counter play.)
42…Kg7 (Now, one of White’s plan is a3, b4 and create the Queenside pawn storm. The problem for Black is there is no plan. Black can only wait.)
43.Rd6 f6 44.Re2 e5 (I prefer something more quiet like Kf7. 44…e5 will only help White open up the position with f4 later.)
45.Red2 g5 46.Nb6 Rb8 (The easiest way for White to win is to start trading pieces. Black also cannot defend against the big pawn advantage of the Queenside for White.)
47.a4 White wins – Kramnik resigns. Fritz won 4-2! 
Way to go, Kramnik – nice choice, and I imagine GM Polgar agrees?
Nothing venture, nothing venture! I hope he can get the full point.
Was the machine out of book for move 9, it looked like it took some time?
In the seventh round of Air Marshal Subroto Mukerjee Memorial All India Open FIDE Rating Chess Tournament, playing with black pieces against IM Shetty Rahul, Umakant Sharma found with a blue tooth device. Umakant Sharma having 2484 rating is a top seeded player for this tournament. It will be interesting to see what action arbiters take against him, it would be the first case in India where a player found carrying electronic means of communication device. Mr. Vishal Sareen said that the investigation is going on.
http://www.cpai.in
umakant sharma zoomed from 1900 to 2484 in almost 1 year!!!!good that he is caught…
Susan’s laughing at the rook movements, I heard if the game’s result will be something to laugh about in the end…
Oops, I meant “I wonder”, not “I heard” as I wrote in the previous post…
Extremely excited but also somewhat apprehensive, anything could happen in a sicilian. However Kramnik-Leko in 2004 does set a precedent.
I will be cheering Kramnik all the way this game.
Has anybody considered the possibility that Kramnik intentionally lost the earlier game for an undisclosed, under the table payoff from Chessbase to boost Fritz sales just prior to Christmas?
where can you watch this game live?
If Sharma is found guilty of using the bluetooth device during play(it has games and chess program) and this is the first offense, penalty should be 3 year suspension and loss of rating (start over from scratch).
Thats a bit of a conspiracy theory 🙂 I dont think Kramnick needs money that much 😛
I think that this new(?!) idea will be very popular after this game…
Yes, anonymous, ChessBase will make tons more money because Fritz can find an elementary mate-in-one.
e5 is a real threat now!
The way Fritz plays, is where he is best! TACTICALLY AND ATTACKING! He showed already the capability of Strategic mode like Yasser and Kramnik, and now he show the capability of Fisher, Kasparov.
If I were Kramnik, I wouldn’t laugh facing this game because Kramnik knows who his opponent is. I wouldn’t think DF10 plays COFFEE SHOP CHESS because I know how he calculates.
THIS WILL BE VERY EXCITING TO SEE WHAT FRITZ CAN DO IN HIS BEST MODE!
I have played this line (upto 12.Nxc6) at least a couple of times in online blitz games (and I’m rated about 1800). Both times my opponents responded with Qxc6. I ended up with 1.5-0.5 I am yet to find why Re3 followed by Rg3 loses straight off the bat
eli
Wow!!! The Najdorf !!! This is incredible for Kramnik – he has only played this opening as Black less then ten times and only won once! Amazing!
15..Rb8 — which Rook?
This is what Mikhail Tal would have done games ago and there would have been a lot less 1/2 1/2 games if he were playing. Deep Fritz would have been pulverized by now. Lets hope Kramnik can attack now without his usual positional Drawnik Style getting in the way.
“White now has to defend the pawn on b2 with Bc1 or Rb1.” But Fritz sees another opportunities!
Rybka recommends Rb1
DF10 programmers should let Fritz play this mode since game 1. But I understand because nothing to lose for DF now.
For Kramnik, because he is fatigue, one move fail means die.
17.Bf4! very very sharp for fatigue Kramnik. He will blunder somehow.
Robert,
Your comments make me think that you have invested your money in Deep Fritz stock. Why the disdain for Kramnik and the ceaseless praise for Fritz? Just curious.
Susan wrote:
10.Re3!?!?!?!? (THIS is a computer move 🙂 I have not seen this move before. GM Rustemov said: “Kramnik should be laughing at the board.”
I am following the game with a Deep Fritz 8 on a 2.8 Ghz Pentium 4 machine. So, it is obviously vastly inferior to the setup DF10 is running, BUT……….
I am observing several moves made by DF10 now, which are 3rd, 4th choice on the list what I am seeing on my screen. Such as the move Susan referred to, as well as move 17 by Fritz. It was on the screen, but as an inferior move compared to other moves. Bc1, Rb1, Rh3 would have left DF8 in the positive, the only move Bf4 what turned it into 0. Of course it is possible and likely that DF10 on a much faster machine does better calculations, but still, I find it a bit weird.
Gabor
Good or not, 11.Rg3 is probably the most “human” move in the posistion. You confuse me, Susan.
~kt
Come on K: 17) …c4!
Hóember said…
15..Rb8 — which Rook?
Az Rab8 bastya 🙂
Gabor
Köszi, sejtettem, de nem lehettem biztos.
>>Of course it is possible and likely that DF10 on a much faster machine does better calculations, but still, I find it a bit weird.
It is to be expected that now Fritz is rigged to lose so the final result becomes 3-3. Such result will be a win-win situation.
3-3 is l null-null I think, not win-win as nobody gains
Hóember said…
Köszi, sejtettem, de nem lehettem biztos.
Hozd be:
http://www.rag.de/microsite_chess/index.html
Kattints: English version
Kattints: Live games
Kattints: here (a Click here to trace….mondatban)
Egy masik ablak fog megjelenni.
Ott kattints Server 6 (vagy akarmelyik azt hiszem)-re
Es akkor eloben kovetheted a jatszmat.
Gabor
After 18.Bc1 Rybka evaluated the position exactly equal (0.00). However, it recommended 18…Bc6 instead of 18…Ng8. On the last, Rybka recommends 19.Qg4 (+0.19).
Keyser Söze!
i am afraid that white with bishop on c2 may gain strong attack…
is 20.c3 playable here?
So did Kramnik *find* Re3 before the match, when he studied the program at home? Or did he *see* it during the game, while looking at Fritz’s opening book?
19.Nb1 and then 20.c3 is lame. N returns to its start with no perspective for development (c3 blocked). Rybka says 0.00 here.
20…c4 somewhat better than 20…g6
I see it! Mate in 173!
21.Na3 (+0.16 Rybka)
Fritz plays the attacking chess of TOPALOV and defending chess of KRAMNIK.
Anybody know where I can buy Fritz stock?
21…Bc6 instead of 21…Qc6. Now +0.33 (Rybka) – White significantly better off.
Shipov thinks 20…g6 is fine. He sees it as a sound positional move: http://online.crestbook.com/kr-fr-06.htm
Rybka prefers 22…Bg7 to 22…a4
“Anybody know where I can buy Fritz stock?”
http://finance.yahoo.com/q?s=BA
Study how this Fortune 500 company ( a member of the DOW Industrial Average Index)
http://finance.yahoo.com/q/cp?s=%5EDJI
engineered a major turn around following 9/11/2001 and their stock tripled since March, 2003.
There is a threat Bxe6, and Qxg6!
some where..
Too early for Bxe6. 2 pawns for B against Kramnik? Nah …
Rybka says 24.Qh4 (+0.36)
25.e5… it’s getting more and more interesting!
24…Rb6 ? Isn’t there a Bb7?
look how DF protect back rank mate with two bishops! Beautiful!
If only Topalov with additional solid defence, he will look like Fritz.
Draw.
should have played Bb7 not queen
fritz owes him a mate-in-one
does anyone know if Kramink is still laughing at the board??
Why would Kramnik take a system like the Najdorf and resort to his old positional tricks? This makes no sense and has been his problem with the Najdorf from the beginning. Even Kasparov has joked that Kramnik has been none too successful because of this positional approach. A positional approach to the Najdorf is suicide… without a change in Kramnik’s style hoping that the Najdorf will bring victory is wishful thinking.
Bad opening choice from Kramnik.
Match was very boring.
Even elite players cannot outcalculate the machine so only real chances are
1.Gaining clear edge AND simplified position straight from preparation.OR
2.Just playing b3 or g3 lines and trying to get a closed game. Stupid chess may actually be best.
Kramnik should resign soon.
This match was IMHO quite boring.
But Kramnik playerd good.
I hate computers.
Chess has no future.
It is interesting how long Fritz is thinking wether to capture Kramnik’s bishop or not.
I don’t get this…
b to h file? hoping to exchange the bishops? why?
This is the scenario.
The operator forget to move the pieces and default.
Then this will be the funniest match ever
Well, I think it’s a well established fact: Computer chess programs are superior to the best human players!
This would’ve been very interesting if a team of GM’s would’ve played Fritz.
Contrary to what some have said, I believe this match WILL actually go down in history. It’s not the first time a computer defeated a human in chess (Deep Blue over Kasparov), but it is the first case of a more or less off-the-shelf consumer program running on an off-the-shelf consumer machine beating a human. That’s huge.
It would have been better if Fritz did not use a special “deep” version, but it might be that Kramnik could not be convinced to play an off-the-shelf release.
Kramnik is doing an excellent job in defending this position .
If this game adjourned, Kramnik will call Mark Dvoretsky to help him in the endgame.
Adjourned?Why do you think so?
“This match was IMHO quite boring.
But Kramnik playerd good.
I hate computers.
Chess has no future.”
No need to be so melancholy Anon. Chess will be alive and well 100 years from now. The ability of computers to play well should not diminish your enjoyment of this great game.
If Kramnik could hold, there is a rule of adjournment after 6 hours of play.
Chess has a future. Computers are much better than people are but there will still be a human champion. Plus, people will continue to learn from computers, though, not being able to overcome them.
the best human player in, say, 100 years will be light years ahead of todays best player. Why?
Computers.
Or 56 moves. From the rules:
“Mr. Kramnik shall have the right to adjourn any game after 56 moves even if six hours of play have not been completed.”
After which point Kramnik will go home and have all the engines and members of his team analyze the position to death. Would that be enough for Kramnik to get a draw?
If nothing else helps advance your king (old chess rule)
“This match was IMHO quite boring.
But Kramnik playerd good.
I hate computers.
Chess has no future.”
c’mon…!
humans have not given up track and field just because a motorcycle runs the 100 meters faster than Asafa Powell does
As of S.Polgar queenside pawn storm plan comment::
Fritz doesn’t see that honey.
Chesscomputers are just cheating. Disable database, and they try to play 1/ Nc3 as first move lol.
as if 1.nc3 was losing…
“chess computers are just cheating…”
I doubt you could even draw with Fritz with anything disabled on the computer. Plus…computers are not intelligent enought to “cheat.” They are not self-aware.
Wow! The World Human Chess Champion is simplfy getting clobbered.
> Chesscomputers are just cheating. Disable database, and they try to play 1/ Nc3 as first move lol.
Humans are cheating. Prevent them from studying openings and ………
This is a titanic strugle for draw.
We must congratulate Kramnik if he saves the game.
Not a borring draw.Not borring at all.
no pawn storm… simply promote the a one 🙂
1-0
I think that advancing black king is on Kramnik’s mind.
This is the end not for CHESS, but for Human and Machine match.
4-2 for DF10.
Who guess correctly? lets check our comment here.
as if 1.nc3 was losing…
I doubt you could even draw with Fritz with anything disabled on the computer.
Some champions have spoken …
=))))
nb1-c3-b1-a3-b5-c3-a4-b6 resigns 🙂 bravo knight ! my match prediction was right he went for win and lost
It was a good match. Kramnik now has made $500,000! Not bad for losing!
The robots are coming!
what are the stats on kramnik’s bathroom breaks against fritz. I’d like to model a statistical distribution to establish some correlation results
Now I’m going to play Re3-Rg3!!!! Not so amateurish afterall. LOL.
I am reposting my comment that I left after game 4.
Anonymous said…
Lets be honest folks. We all know that Kramnik Is going to lose this match Don’t we? He already said he lost most of his games against his training version at faster time controls. His love of the draw and general fear of the machine almost gaurantees he’s going to either lose or draw the rest of these games.
When Kramnik blundered into checkmate the most amazing thing is he wasn’t even being lazy. He was paying attention and calculating. Also this isn’t an isolated event. I forget the game but he’s suffer from at least one other very basic hallucination.
Marcellus Wallace
Friday, December 01, 2006 8:18:23 PM
maybe if Kasparov comes out of retirement like Obi Wan Kenobi coming back to fight the Empire
I think people still underestimate the quality of the struggle that Garry put up against Deep Blue, back in ’97. Now that super fast consumer grade hardware is available to run programs, like Fritz & Junior, we see how strong these computer systems actually are. Positional assessments aren’t as important to the machines, because they have a much greater ability than humans to find good lines (drawing or winning) in positions that our concepts tell us shouldn’t be favorable.
I think that even Kasparov from his best days (ELO 2850) wouldn’t be able to defeat DF10.
To draw-maybe..
kramnik choosing the najdorf was a huge mistake. he never even got a feel for the initiative.
would it have been good if Kramnik played c4 instead of a4 and then blocking the white bishop with e5?
What a shame! Nice try Kramnik
I hope we can see another victory from humans against the machine. I dont mind if it comes from Topalov, 😛
What a shame that Kramnik made a very avoidable blunder in Game 2 that probably cost him the match (since he likely would have played a quieter opening in which he’s proven himself Fritz’s equal were the match equal after 5).
In this game, although Kramnik made some minor mistakes before 27…Qb7 that put him in a somewhat inferior position, I really don’t understand this move. He got nothing for the pawn, so whether he intended to sacrifice the c6 pawn or not, he basically just dropped it, game over. Kramnik is a great player, but he makes more basic tactical mistakes than a lot of past world champions did.
Is it the end of Human’s fight against the Machine? Suddenly I want Kasparov to be back.. or maybe Arno Nickel can fight Fritz next.
Has anyone ever played at a GM level against such a computer chess program (it’s a program, not a “machine” btw) with the opening book totally disabled?
I think that GK’s strategy of deviating from theory as soon as possible is the best when facing chess software. On the other hand, it looks kinda scary that a chess program could play “out of book” as early as move 10, and still be able to beat the World Champion in Chess.
BTW, does this win mean that DF10 is the World Champion in Chess? 😛
I couldn’t agree more with something an anonymous person said a few lines above: $500,000 for a loss, not bad.
Great game. Not bad at all. Kramnik just finished his job to test this program. And got a great pay too.
i don´t believe this is the end of human chess. i think this match marks the end of cp-chess developement. in ten years,very much likely, some version of deep fritz may run on any cellular phone, but no one will bother. do you compete in extracting roots with your calculator?
Kramnik made a mistake of going into quiet positional lines earlier in the match. Then after realising he could not grind fritz down, he desparately switched to a Sicilian ultra sharp game in last game knowing that humanity will be hanging with him. Too little too late!
From playing comps alot myself, I’m starting to believe that you have to go toe to toe with them. Get kasparov out of retirement just like they did to that gun slinger Lee Marvin in the flick Cat Boulue!!!
http://www.fide.com/ratings/id.phtml?event=5017459
suspicious rating improvement
Move 10 of Fritz is a theory novelty. Maybe if the opening book of computers would be restricted to fewer moves we could see some more interesting games in the future.
Yasseer wants comps to be handicapped in future matches with humans, My Goodness ,Kramnik got the most favourable conditions ever, how much more handicap should the comps give? Every other GM did not get the same privileges Kramnik got. I really wonder how Topalov or Kasparov would perform with such privileges. With these advantages Kramnik had some superior positions but still could not win, whose fault is that? All this flimsy excuses for Kramnik is unnecessary. As a complete package Kramnik’s best years are behind him. As a complete package Kramnik simply is NOT the best HUMAN player in the world, PERIOD.
Limiting the opening books of computers to earlier moves should not be regarded as a handicap. Au contraire, this would allow them to make more innovations in the opening, like in this game, at 10.Re3
Kramink have playing wery well but last games he trye too wins and after 19.Nb1 (The position is equal. The idea of this move is to play c3 to clear the c2 square for the Bishop on b3.)
i will sage that wery human fritz play after (kramnik play ng8!)
Soo big surprice for Kramnik are not re3 but nb1! play like human!
Knightonranden.
Interesting commentary…GMs poo-poohed some Fritz moves – till they saw the light.
:-)))
http://chessbase.com/newsdetail.asp?newsid=3524
“1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bc4 e6 7.0-0 Be7 8.Bb3 Qc7 9.Re1
9…Nc6.
This position has occurred at least nine time in international tournament chess, with the moves played being 10.Bg5 (four times), 10.a3, 10.Nxc6, 10.Qd3, 10.Be3 and 10.f4. The move and manoeuvre Deep Fritz plays has certainly never been seen before.
10.Re3 0-0 11.Rg3 Kh8 12.Nxc6 bxc6 13.Qe2 a5 14.Bg5 Ba6 15.Qf3 Rab8 16.Re1 c5 17.Bf4 Qb7 18.Bc1.
The manoeuvre Bc1-g5-f4-c1 (with Ra1-e1 incerted) caused some amusement in the commentary boxes in Bonn, but as with Rf1-e1-e3-g3 the GMs slowly started to see meaning in the “madness”.
18…Ng8 19.Nb1. Amazing: a mirrored knight retreat on both sides. Once again the GMs doing live international commentary broadcasts were surprised and amused by the computer moves, but slowly started to find reasons for it.
19…Bf6 20.c3 g6. The commentators who had given Black a substantial advantage (while Deep Fritz persistantly saw White in the lead) now slowly started to have doubts about Kramnik’s position.
21.Na3. This was not the move the GMs anticipated (…Nd2 was the reason given for 19.Nb1). Deep Fritz is still playing moves that take a while for humans to appreciate.
21…Qc6 22.Rh3 Bg7 23.Qg3 a4 24.Bc2 Rb6.
Kramnik said later that immediately after he had played the move 24…Rb6 he had seen the computer’s reply:
25.e5! dxe5 26.Rxe5 Nf6 27.Qh4 Qb7 28.Re1 h5 29.Rf3 Nh7 30.Qxa4 Qc6
White is now a pawn up and firmly on the path to victory.
31.Qxc6 Rxc6 32.Ba4 Rb6 33.b3 Kg8 34.c4 Rd8 35.Nb5 Bb7 36.Rfe3 Bh6 37.Re5 Bxc1 38.Rxc1 Rc6 39.Nc3 Rc7 40.Bb5 Nf8 41.Na4 Rdc8 42.Rd1 Kg7 43.Rd6 f6 44.Re2 e5 45.Red2 g5 46.Nb6 Rb8 47.a4 and White resigned.
Fritz has played this game very impressively, as all the GMs and Vladimir Kramnik himself have admitted. 1-0.”