|
< Earlier Kibitzing · PAGE 16 OF 18 ·
Later Kibitzing> |
| Jan-24-11 | | Kazzak: Go to the Chessbase-channel on YouTube:
http://www.youtube.com/user/ChessBa... |
 |
Jan-24-11
 | | alepo: After 30..Bxg7, the most straight and brilliant line would have been, 31.Nf5 Rc7 32.Nxh6+ Kh7 33.Nf7+Kg8
34.Rh8++. |
|
Jan-24-11
 | | Domdaniel: Live games become part of the database before any moves are played, but data still has to go into the relevant fields. The players, ratings, site, and date are known -- but not, obviously, the opening. So 'Uncommon Opening' is used as a temporary place-holder. Somebody will change it later. Now nobody will ever have to ask that question again. |
|
| Jan-24-11 | | geladar: Chessgames.com seems to have some difficulties to identifying specific openings..this is definitely not an "Uncommon Opening".. |
|
| Jan-24-11 | | BobCrisp: Why is this <Uncommon Opening>? |
|
| Jan-24-11 | | frogbert: kramer, i use engines when i watch these events. but i have started to turn the "audience sound off". booing and cheering at all the wrong moments spoils the experience for me. and clearly, a number of "fans" get a totally wrong impression of what's going on by their abuse of engines. it's mostly due to "user/operator errors" of course. would you let lay-men interpret lie-detector tests? some tools are "dangerous" in the wrong hands. |
|
| Jan-24-11 | | frogbert: btw, the Nxe6 line that eyal explained isn't even a clear win, by far. but it forces a continuation where white doesn't get another chance of deciding the game. in reality carlsen's moves were better than the immediate engine recommendations on several occasions. even objectively. |
|
Jan-24-11
 | | Hesam7: <Eyal: Nakamura might have considered simplification with 17...Nec4 18.Bxc4 (18.Qc1 Nxb2!; 18.Qd4 Nxb2! 19.Kxb2 Qxc3! 20.Qxc3 Na4+) 18...Nxc4 19.Qd4 Na3+ 20.bxa3 Qxc3 21.Qxc3 Rxc3 22.Bxd6 Bxd6 23.Rxd6 Rg3.> I don't know after 17. ... Nec4 18. Bxc4 Nxc4 19. Qd3 Na3+ 20. Kc1: 
click for larger viewAnd now Stockfish 2.0.1 gives the following @ depth 30: 20. ... Qc4 21. Nd4 Nb5 22. Qxc4 Rxc4 23. Nxb5 axb5 24. b3 Rc5 25. Bxd6 Bxd6 26. Nxe6 fxe6 27. Rxd6 with an evaluation of +0.44. |
|
Jan-24-11
 | | Kinghunt: <frogbert: btw, the Nxe6 line that eyal explained isn't even a clear win, by far. but it forces a continuation where white doesn't get another chance of deciding the game. in reality carlsen's moves were better than the immediate engine recommendations on several occasions. even objectively.> I quite agree. For those not wanting to dig back through the kibitzing, here's Eyal's line: <Eyal: Again, the tactics of this 27.Nxe6 line is crazy: 27...Rxe6 28.Rxh6! gxh6 29.Qh3 Rce8 30.Rxf8+! Kxf8 31.Qxh6+ Ke7:And you have to see that White is winning here with 32.b4!! - trapping the black queen either by an immediate capture or a knight fork on d5 - unless it goes to d8, and then 33.Qg7+ mates...> The problem with this line, even ignoring how inhuman it would be to see 32. b4, is that even then it's not clearly winning, and in fact, probably draws. One continuation would be 32...Qc7 33. Nd5+ Kd7 34. Nxc7 Rxg6! and suddenly white has nothing more than one extra pawn. After 35. Qxg6 Nxg6 36. Nxe8 Kxe8, we end up at the following position, which, as GM King says, is probably drawn: 
click for larger viewSo Carlsen's line is definitely better than the engine move. A rare case of a computer miscalculating a tactical sequence. |
|
Jan-24-11
| | Hesam7: Just watched Maguns's commentary. Given how complex the game was I was a bit disappointed that he did not go into much detail. He thought Nakamura's decisive mistake was 19. ... Qc5?. The two of them discussed 19. ... g6 after the game and that might have been a better choice. But Black's best move which is not mentioned by Carlsen is 19. ... Nec4:
click for larger viewNow Stockfish 2.0.1 gives the following @ depth 32: 20. Bxc4 Nxc4 21. Qd3 Qb6 22. Nb3 g6 23. h6 Qb4 24. Rdf1 Ne5 25. Qg3 Nc4 26. Qd3 with an evaluation of 0.00. |
|
Jan-24-11
 | | Eyal: <Hesam7: 17. ... Nec4 18. Bxc4 Nxc4 19. Qd3 Na3+ 20. Kc1 [etc.]> Yeah, for some reason I missed Qd3 during the game and looked only at Qc1 & Qd4 - I agree that it's clearly better for White, though maybe not as bad for Black as in the actual game. In order to significantly improve the setup for Black you probably have to go earlier - for example, 12...a5 (which was used successfully a couple of times by Rashkovsky - http://www.chessgames.com/perl/ches...), either provoking some weaknesses if White plays a4, or pushing the pawn to a4, kicking away the knight on b3 before it can go to d4 and generating some quicker counterplay on the Q-side. |
|
Jan-24-11
| | Eyal: Btw, the video of Carlsen's game analysis is on (http://www.chessvibes.com/reports/c...) and he also mentions 12...a5 as an alternative for Black. With regard to the opening, he refers to 8.Be3 as a "tricky sideline" (instead of the more usual 0-0) - hoping for 8...0-0?! (instead of Be6), and now the delay of K-side castling - normal in the 6.Be2 line - allows White to shift gears and start an "English attack" type of game, with Q-side castling and a pawn storm on the K-side, which one usually sees only after 6.Be3, by 9.g4! [as was already explained here, without the bishop on e6 Black can't counter that by 9...d5, and once Black has committed himself to K-side castling, 9...h6? to stop g5 would be bad, since it only helps White's attack after 10.h4!] A couple of other interesting points - The prophylactic 18.Ka1 is explained by the line 18.Nd4 Nbc4 19.Qe1 Qb6, after which 20.Bc1 would fail to 20…Na3+ 21.Ka1 Qxd4! [engines show 21...Nc6!! as even stronger - in fact, winning by force] followed by 22…Nxc2+. Carlsen mentions the need to defend against ...Nxb2 ideas by Black as very important in his considerations - this relates to his playing 26.Rdf1, so that the rook wouldn't be under attack in such a case, as well as 24.Qg3 a couple of moves earlier - keeping an eye on Nc3. It's interesting to compare that with another computer recommendation - <24.Qf1> (which, unlike the spectacular 27.Nxe6/32.b4 line, <is> objectively very strong...); one of the reasons such a move is harder to find/play is that it removes support from c3, even though an engine would show you in an instant that 24.Qf1 Nxb2 doesn't work for Black because of 25.Nxe6 Qxc3 (or 25...Rxe6 26.Bxd6!) 26.Be5!! Qxe5 27.Qf7+ Kh8 28. Rxh6+! and mate. |
|
Jan-24-11
| | Hesam7: White could have ended the game much faster with 24. Qf1!
click for larger viewThe Queen is heading for the f7-square after Nxe6. The bishop does not have a retreat square that still covers f7 so Black has to play: 24. ... Rc7 25. Nxe6 Rxe6 26. Nd5 Rxg6 (Black can't save the c7-rook: 26. ... Rd7 27. Qh3!) 27. Nxc7 Qxc7 28. Qd3: 
click for larger view |
|
Jan-24-11
 | | Bobby Fiske: For this game, Magnus won the 500 euros for the best game of the day in Round 8. |
|
Jan-24-11
| | Kinghunt: < Hesam7: White could have ended the game much faster with 24. Qf1! The Queen is heading for the f7-square after Nxe6. The bishop does not have a retreat square that still covers f7 so Black has to play: 24. ... Rc7 25. Nxe6 Rxe6 26. Nd5 Rxg6 (Black can't save the c7-rook: 26. ... Rd7 27. Qh3!) 27. Nxc7 Qxc7 28. Qd3> But like so many other lines that have been posted, it's incredibly difficult to calculate to a clear conclusion. Suppose 24...Qb4. Then we have 25. Bc1 Ne5 26. Nxe6 Rxe6 27. Qf5 Rce8 28. Nd5, reaching the following position: 
click for larger view
Black to move
White has an intimidating looking attack, but stop and analyze the position for a second. The sole reason white's attack looks worrying is because of the awkward placement of black's queen. If black's queen were on d7 right now, black would be fine. So then the question becomes, can you calculate a line to show that white's attack is actually going somewhere before black gets his queen to a better position to defend? I don't know about to a GM, but this looks like anything but a slam dunk conclusion to me. The silicon beast spits out 28...Qa4 Rg1 29. Rdg1 Qd7 30. Rh3 Rb8 31. Rc3 as winning because white will shortly win an exchange by force, but really, going into a sequence like this where you have to go so deep to make sure the attack doesn't just fizzle out into nothing is just inhuman. |
|
Jan-24-11
 | | Rolfo: <For this game, Magnus won the 500 euros for the best game of the day in Round 8.> Well deserved |
|
| Jan-24-11 | | frogbert: <<24.Qf1> (which, unlike the spectacular 27.Nxe6/32.b4 line, <is> objectively very strong...); one of the reasons such a move is harder to find/play is that it removes support from c3, even though an engine would show you in an instant that 24.Qf1 Nxb2 doesn't work for Black because of 25.Nxe6 Qxc3 (or 25...Rxe6 26.Bxd6!) 26.Be5!! Qxe5 27.Qf7+ Kh8 28. Rxh6+! and mate.> eyal, i made more or less exactly this comment on icc during the game, when people were screaming about 24. Qf1 - claiming that it was rather unlikely that a human would find Qf1 for exactly the reasons you mention (Qf3 looks more normal, as it keeps an eye on c3 while still threatening tactics against f7). this was when carlsen still was contemplating his 24th move, of course. but people were going: "oh, no - how can he miss Qf1 ..." when carlsen played the logical <and> strong 24. Qg3 instead. *sigh* |
|
| Jan-24-11 | | frogbert: <28...Qa4 Rg1 29. Rdg1 Qd7 30. Rh3 Rb8 31. Rc3 as winning because white will shortly win an exchange by force, but really, going into a sequence like this where you have to go so deep to make sure the attack doesn't just fizzle out into nothing is just inhuman.> i totally agree with that assessment, kinghunt.
i've been looking at 24. Qf1! for a while now, and as you say, the <tactical> reason why it wins, is that white wins a piece for a pawn, alternatively an exchange, by force, <some time after move 35> - 10-12 full moves ahead in a position where there are dozens of equivalent lines to choose from. while an engine only can justify such a line by tactical means (it knows nothing else), a human will never consider that line and choose it because he sees the win of material. it might be chosen for "positional" reasons - as a decent way of improving the positions of one's pieces and continuing the attack - but finding a <tactical justification> for 24. Qf1 is pretty hard. in particular because you need to be sure that black doesn't have anything nasty against b2 and such - which includes finding the Nxe6, Be5 and Rxh6 solution to the direct Nxb2. in my opinion, carlsen plays more to the point with 24. Qg3 and 26. Rdf1, increasing the pressure on black's king's position to the maximum. while engines might find ways to defend, a strong player like nakamura simply cracked under that pressure and the game was over much faster than after the very deep (and tactically complicated) move 24. Qf1. |
|
| Jan-24-11 | | frogbert: consider this position, after white's <36th move> for instance:
click for larger viewblack to move. what's his best move?
i doubt many humans would pick rybka's first choice here for black. this is after one of dozens of possible lines after 24. Qf1 - where black has successfully manouvered his queen back to d7, shuffled his rooks some, and white has taken on e6 and played a3, Kb1, Bd2-c3, Qf5-h5, Rdf1 and Rhg1-g2. the reason white has been able to do all that, is that black is in an advanced type of "zugzwang" - not really being able to do anything useful except waiting for white to get his pieces to optimal squares. it's worth noting the similarity of the placement of white's pieces here with how carlsen's heavy material ended up in the f, g and h files in the game, too. did you solve the exercise above? rybka "thinks" that Qe8!? is a decent try for black here, "hoping for" the fork Nc7 winning the exchange for white. go figure ... (going deeper it finds that it needs to give up a whole piece instead - oh well, you get my point i guess). |
|
Jan-24-11
 | | Ulhumbrus: After 19 Nd4 the move 19...g6 pays a price for getting Black's King's Bishop into play : it disturbs the King side pawns. This suggests keeping the King's Bishop on e7 and waiting for g6 before playing ...Bf6 to get the bishop into play. In that case what is Black's King's Rook doing on e8? This suggests playing 17...Nb6-c4 inatead of 17...Rfe8. Another interesting alternative comes earlier at move 13. In reply to 13 f4, instead of 13...exf4, 13...f5! may end White's attack before it can begin. |
|
Jan-24-11
| | Hesam7: <And there’s the first blow at point-blank range! Hand-to-hand fighting is beginning. The f7-pawn, it seems, is overloaded with responsibilities. It needs to keep an eye on the bishop – which is in trouble from the white knight – and to protect the g6-point. It seems the advance of the white pawn was an unpleasant revelation for Hikaru. He’s having a long think. The clocks show: 0:45 – 0:45.> -- GM Shipov during his live commentary <I felt somehow that after 19. ... Qc5 20. g6 in a sense the game is strategically decided.> -- GM Carlsen commenting on his win after the game. So Shipov from the comfort of his own home and with various engines at his disposal still got "out-evaluated" by Carlsen. |
|
Jan-24-11
| | Hesam7: <Eyal: <Hesam7: 17. ... Nec4 18. Bxc4 Nxc4 19. Qd3 Na3+ 20. Kc1 [etc.]> Yeah, for some reason I missed Qd3 during the game and looked only at Qc1 & Qd4 - I agree that it's clearly better for White, though maybe not as bad for Black as in the actual game. In order to significantly improve the setup for Black you probably have to go earlier - for example, 12...a5 (which was used successfully a couple of times by Rashkovsky - http://www.chessgames.com/perl/ches...), either provoking some weaknesses if White plays a4, or pushing the pawn to a4, kicking away the knight on b3 before it can go to d4 and generating some quicker counterplay on the Q-side.> Why not go another move earlier and play 11. ... b5 ? I have not checked to see if there is a tactical refutation but it is a lot more in the spirit of Najdorf. |
|
| Jan-24-11 | | dale2222222: pretty |
|
| Jan-24-11 | | The Chess Express: White's opening system looks pretty convincing. |
|
Jan-25-11
| | Hesam7: After 11. ... b5 I came up with this: 12. Nd5 Nc5 13. f4 exf4 14. Bxf4 Nc6:
click for larger viewAnd Black looks way more comfortable compared to the game. My engine came up with: 15. Nxe7 Nxe7 16. Nxc5 dxc5 17. Qxd8 Raxd8: 
click for larger viewDespite my engine's evaluation I would rather play Black. Any improvements for White? |
|
|
|
< Earlier Kibitzing · PAGE 16 OF 18 ·
Later Kibitzing> |
sac: 29.Bxh6