|
< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> |
| Jul-22-11 | | messachess: With 17..b5, black would love, of course to open the <c> file for his rook on a8 to bear down. Not going to happen. 18.c5 is 'right back atcha!'--setting the stage for the exchange sac. 20. Qxb4. After 25.Bxc6, whites bishops have come rather extravagantly into their own and black's troubles loom.--but not too bad it would seem (black down a pawn) except that after 25..Re2, white dominates the queen side with black's rook stuck on <a7>. Of course, black didn't have to blunder and lose a piece with 41..kf6 (time trouble?) Still, the pawn on <b5> has to fall to <R-d5> (which is why black wanted the <K> on <f6> to protect the <N>. |
|
| Jul-22-11 | | twinlark: <al wazir>
<23...Nef2+ 24. Rxf2 Nxf2+ 25. Bxf2 Bd4 26. Bxd4  > The DSB still protects g1. |
|
| Jul-22-11 | | Garech: Superb game from Kramnik.
-Garech |
|
| Jul-22-11 | | Hesam7: As a long term fan I am very worried, his opening preparation isn't what it used to be. Here with White pieces Kramnik completely loses the opening battle against an opening (KID) he traditionally did very well. He was lucky Ponomariov did not find 19. ... Nxe4! after which Black secures an advantage in every line (I will post some analysis later) |
|
Jul-22-11
 | | Honza Cervenka: <Hesam7> I see that after 19...Nxe4 20.Qxe4 (what else?) 20...Bxb3 21.axb3 Rxe4 22.Bxe4 Bxa1 23.Rxa1 black won the Queen for three minor pieces but I am not sure that he has an advantage there. In fact, I would rather play as white from there. |
|
| Jul-22-11 | | DrMAL: <twinlark> Thanx...I think LOL. At the level in this game, many if not most mistakes are judgment calls. Players often take risks to go for a win instead of a draw, and their play shows up as a "mistake" when the GM purposely chose it and it was not best after all. In this game, key choices were made by black on moves 27, 28 and 36. There were others, these moves seemed to involve mostly judgment so they fit together as a whole to me. I let this position compute overnight and here is the result. I think 36...h4 may still produce a draw so I will take a hard look, it's a very interesting game anyway:
click for larger view Houdini_15a_x64: 30/83 5:33:34 99,965,545,305
-0.62 36. ... h4 37.Nf6+ Kg7 38.Ne8+ Kg8 39.Nc6
-0.77 36. ... hxg4+ 37.hxg4 Kg7 38.g5 Rh2 39.Nc6
-0.92 36. ... Ne7 37.Nc3 h4 38.Ne4 Kg7 39.Rd2
-0.92 36. ... Kg7 37.gxh5 gxh5 38.Rg1+ Kh6 39.Rg2
-0.98 36. ... Rh2 37.gxh5 Nd6 38.hxg6 fxg6 39.Nf6+
Here, Pono's judgment was to ignore 36.g4 a strong move. I am admittedly far best in middlegame play and least so in endgame play, so I very much welcome any help, cheers. |
|
| Jul-22-11 | | DrMAL: <Hesam7: He was lucky Ponomariov did not find 19. ... Nxe4!> I think Pono saw this and decided otherwise, we are talking GMs here. In any event, yes, 19...Nxe4 evaluates best giving slight advantage to black whereas 19...Bf5 evaluates as second giving slight advantage to white. The purpose of 19...Bf5 was to stir up more complexity in the game rather than reduce it via 19...Nxe4. As such it serves as an excellent fourth example to my initial post on judgment. Pono tends to like making fancy moves where they are not required, he has had a lot of success with that versus less (extremely) strong opponents. It's a strategy that only goes so far. |
|
| Jul-22-11 | | DrMAL: Ooops I got sidetracked onto 19...Bf5 instead of 19...Bc4 that was played and my text got messed out (a line got lost). Meant to write after "via...Nxe4." above that: 19...Bc4 probably third best objectively, attempts to create even more complexity by allowing white the positional sac he made. This was surely also seen by Pono but he chose to do it anyway. It's a judgment call, I kinda beat this to death but it seems in prior posts I got misunderstood for making statements such as "mistake." The mistake usually referred to judgment. Computer analysis is tricky this way. It can compute via tactical combinations what is "wrong" and what is "right" depending on how good its evaluator is. Presently, there is no output about how positional strategy affects this even though it is an integral part of the software. In order to use computers correctly one needs to make sure it has enough time to compute. But in addition, it helps to understand chess a bit more before making statements, particularly if one's aim is to trash another from the wording in their posts. I do not aim this statement at anyone in particular, the very few I've encountered are in an ignore list. |
|
| Jul-22-11 | | Hesam7: <Honza Cervenka: <Hesam7> I see that after 19...Nxe4 20.Qxe4 (what else?) 20...Bxb3 21.axb3 Rxe4 22.Bxe4 Bxa1 23.Rxa1 black won the Queen for three minor pieces but I am not sure that he has an advantage there. In fact, I would rather play as white from there.> Damn! I only looked at 20. Bxe4 and 20. Qxb4. But I think my evaluation of 19. ... Nxe4! still holds. Although 3 very complicated responses have made my analysis prohibitively long :( 20. Qxe4 Bxb3 21. axb3 Rxe4 22. Bxe4 Bxa1 23. Rxa1 dxc5 24. Nxc5 (24. Bxc6 Rb8 is worse because Black gets to keep the c5-pawn) 24. ... Qc7! 
click for larger viewFrom the diagram I think Black is better. Examples:
<A> 25. Nxb4 Qxf4 26. Bxc6?! Rd8! 27. Nbd3 (the b4-knight was threatened and 27. Nbxa6?? loses to 27. ... Qf6!) 27. ... Qd2 28. h3 Rxd3! (28. ... Ne3 is my engine's recommendation but after 29. Bxe3 Qxe3 30. Re1 I don't see how Black can make any progress) 29. Nxd3 Qxd3 30. hxg4 hxg4 and this position deserves a diagram: 
click for larger viewAnd this is practically lost for White. Me vs Engine: 31. Ba4 f5 32. Bc6 g3 33. Ba7 Qxb3 34. Bf3 a5 35. Rd1 Qb5 36. Be3 a4 37. Bf4 a3 38. Bxg3 a2 39. Bf2 Qb2 40. Bd4 Qxd4 41. Rxd4 a1=Q+ 42. Rd1 Qe5 43. Rf1 g5! and here I stopped because the evaluations went below -5.00. <B> 25. g3 Qd6! 26. Bd4 (only way of defending against ... Qd2) 26. ... a5 27. Bf3 a4! 28. Kg1 (28. Nxa4? Qe6 and White is lost because the a4- knight is pinned; 28. bxa4 b3! 29. Nxb3 Qe6 and Black will win material because the White minor pieces are uncoordinated) 28. ... a3! 29. Ne4 Qd8 30. Bxg4 hxg4 31. Nf6+ Kg7! (31. ... Kf8? 32. Nh7+ Ke8 33. Nf6+ Ke7 34. Re1+ is much better for White) 32. Nxg4+ Kf8 33. Nf6 a2 <C> 25. h3 Nf6 26. Bf3 Qxf4 27. Nd3 Qf5. Here I don't have a forced line but here is a sample line: 28. Rd1 Ne4 29. Nd4 Ng3+ 30. Kh2 Qf6 31. Kxg3 Qd6+ 32. Kh4 Rc8! and another diagram worthy position arises: 
click for larger viewIt is hard to evaluate this. But a good sign for Black is that despite his huge material deficit engines do like Black's position: Stockfish 2.0.1 gives -1.21 @ depth 29. |
|
| Jul-22-11 | | twinlark: <DrMal>
I spent some time on looking at 36...h4 and it's a close thing, but I think White can still extract a win from it. In practice, it would probably be a win for White as these sort of defensive actions are extremely difficult for the defender, with the attacker generally having an easer job of it with a wider array of options. <Honza Cervanka>
Seems that that variation plays out to a forced draw, namely it's a dead end which is maybe why Pono didn't play it. After 19...Nxe4 20.Qxe4 Bxb3 21.axb3 Rxe4 22.Bxe4 Bxa1 23.Rxa1: 
click for larger view23...Qc7 seems close to forced. A likely line would be 24. Nxb4 d5 25. Bf3 Qxf4 26. Nxc6 Qf6 27. Nd4: 
click for larger view and now 27...Nxh2 28. Bxh2 Qxd4 29. Rd1 Qf2 followed by ...Re8-Re1+ and after the rook exchange a draw by perpetual. There are other sub-variations of course, but this whole variation seems like a dead end. |
|
| Jul-22-11 | | twinlark: <Hesam7> In your variation B, after <25. g3 Qd6 26. Bd4 a5 27. Bf3 a4 28. Kg1 a3 29. Ne4 Qd8>, how about <30. Rd1> instead of <30. Bxg4>? 
click for larger viewFor example, if:
1. <30...a2 31. 31. Bxg4 hxg4 32. Nf6+ Kg7 (32...Kf8 33. Bc5+) 33. Nxg4+> and White can force the perpetual, or if 2. <30...Qe7 31. Re1> and if <31...Qb7 32. Nd6> or <31...Qc7 32. Bxg4 hxg4 33. Nf6+ Kf8 34. Nh7+> is a perpetual. |
|
Jul-22-11
 | | al wazir: <twinlark: 24. Rxf2 Nxf2+ 25. Bxf2 Bd4 26. Bxd4. The DSB still protects g1> You're right. Too bad, it was a lovely combination. Thanks. |
|
| Jul-23-11 | | Hesam7: <twinlark: <Hesam7> In your variation B, after <25. g3 Qd6 26. Bd4 a5 27. Bf3 a4 28. Kg1 a3 29. Ne4 Qd8>, how about <30. Rd1> instead of <30. Bxg4>?> Black's best seems to be 30. ... c5:
click for larger view<a> 31. Nxc5 Rc8 32. Nxb4 Qd6 33. Bf2 Qb8 34. Bxg4 (34. Nbd3 & 34. Ncd3 both are met by 34. ... Nxf2 35. Kxf2 and then 35. ... Qd6 & 35. ... Qb6+ respectively) 34. ... hxg4 35. Nbd3 (35. Ncd3? Rc3) 35. ... Rd8 (35. ... Re8 would allow 36. Ra1 which now loses to 36. ... Rxd3 37. Nxd3 Qxb3) and here I have hard time coming up with any moves for White. <b> 31. Bxc5 Qa5 32. Bxg4 hxg4 33. Nf6+ Kg7 34. Bd4 a2 35. Nxg4+ Kf8 
click for larger view36. Nge3 Qh5! 37. Nxb4 Qe2 38. Nbc2 Rd8 39. Re1 Qd3 40. Rd1 Qxb3 41. Kf2 Qb1 and White's pieces are practically frozen; 36. Kf2 seems like a better way of keeping teh Black queen out of e2 but Black has 36. ... g5! 37. Ne5 (blocking the 5th rank) 37. ... Rc8 38. Na1 f6 39. Nd7+ Kf7 40. Nxf6 gxf4 (so White can't defend the knight with a pawn) 41. gxf4 Qd8 with ... Qh8 to follow. |
|
| Jul-23-11 | | DrMAL: After 36...h4 here is what I think is best for both sides (checked for mistakes with Houdini): 37.Nf6+ Kg7 38.Ne8+ Kg8 39.Nc6 Rc2 40.Ne5 Be7
From here white has two main options, the first is more tactical and maybe more likely to be chosen: A) 41.g5 Kf8 42.Nf6 Rxa2 43.Rd7 Ra7 44.Kg4 Rxd7 45.Nfd7+ Kg7 This option has many subvariations (and transpositions), I played through as many as I could at various important branches. Here are two principal lines where black's play keeps trying to simplify while white maximizes pressure: A1) 46.Kxh4 Ba3 47.Kg4 Ne7 48.Nf6 Bb2 49.Ne8+ Kf8 50.Nc7 b4 51.Nd7+ Kg8 52.Kf3 Nc6 53.Ke4 Bc3 54.Nd5 Bd2
click for larger viewWhile white has good king position and an extra h-pawn, black's bishops seems to hold the two sides separated by three ranks. A2) 46.Nc6 f5+ 47.Kxh4 Ba3 48.Kg3 Nd6 49.Nde5 Ne4+ 50.Kg3 Nc5 51.b4 Nb3 52.Kf2 Bb2
click for larger viewPlaying out further, all that's left is a knight for each side and the three kingside pawns versus two (51.b4 was not forced, but other alternatives give very similar positions to the earlier line). The second main option is tactically simpler, here is one short line: B) 41.Rd7 f6 42.Nxf6+ Bxf6 43.Nxg6 Rc6
click for larger viewClassic three pawns for bishop where the pawns are divided 2:1 and 3:1 with nothing grimacing for black. There is still a lot of play left but the position is basically equal (slight advantage white). If Pono chose 36...h4 to take care of 36.g4 it seems to me he could have emerged with half a point. |
|
| Jul-23-11 | | Nietzowitsch: Game annotated by Jan Gustafsson : http://www.jan-gustafsson.de/partie... |
|
| Jul-23-11 | | DrMAL: <Nietzowitsch> Thanx for the link. Apparently, GM Gustafsson agrees completely with me.
He wrote 26...Nh6? (26...a5!) perhaps a bit too harsh, as the penalty was not so severe. But in writing "Kramink now takes over" his point is the same. It was indeed here that the Kramnik took the initiative. Instead of criticizing 28...Nf5 he merely gave an exclam to 29.Bb5 instead (black had simpler better moves to avoid this, as I pointed out). Same thing. He marked 36...Kg7 as dubious but reinforced it with "tough, but against Kramnik probably hopeless." Jan only offered 36...hxg4 as the alternative, maybe he felt that was sufficient to survive. I have not carefully looked at the consequences of this move, as after I analyzed the alternative 36...h4 it was good enough. He also gave 19...Bc4 a question mark, writing "strong instead was 19...Nxe4!" I pointed out later this was bad and offered Nxe4, but did not want to pick on Pono, this choice is more a matter of style and Pono did not mind the positional sac, basically asking Kramnik to prove it (as he did from black's later decisions). Much better verification than simply making sure via Houdini I did not make mistakes, cheers! |
|
| Jul-24-11 | | twinlark: <Hesam7>
Thanks. I'm convinced. I found some subvariations stemming from your analysis that stiffen White's resistance (eg: after 30...c5 31. Nxc5 Rc8 32. h3>), but it's definitely uphill for White all the way. Makes you wonder about both GMs though, as it was not only Kramnik's preparation that was lacking. |
|
| Jul-24-11 | | twinlark: <DrMal>: Try 36...h4 37. Rc1. |
|
| Jul-24-11 | | DrMAL: <twinlark> I suggest you try it and convince yourself or, better yet, post some analysis on it, cheers. |
|
| Jul-24-11 | | twinlark: I have convinced myself of it and I will post some analysis of it. |
|
Jul-24-11
 | | Richard Taylor: Kramnik certainly knows how to move the pieces! |
|
| Jul-24-11 | | twinlark: <DrMal>
<36...h4 37.Nf6+ Kg7 38.Ne8+ Kg8 39.Nc6 Rc2 40.Ne5 Be7> White has <39. Nd5>: 
click for larger viewThis should win, eg: <39...Rh2 40. Ndf6+ Kh8 41. g5 Ne7 42. Nd6 Kg7 43. Kg4 Rxa2 44. Kxh4 > 
click for larger view |
|
| Jul-25-11 | | Robyn Hode: DrMal, what UCI are you using for Houdini? |
|
| Aug-14-11 | | Hesam7: I have already covered <1> 20. Qxe4 so here are some other tries for White after 19. ... Nxe4!
click for larger view<2> 20. Nxb4?? Bxb3! 21. axb3?! Nef2+ 22. Bxf2 Rxe1 when after both 23. Bxe1 Bxa1 or 23. Raxe1 Nxf2 24. Rxf2 Qxb4 it is over. <3> 20. Qxb4 Qxb4 21. Nxb4 Nxc5 22. Rad1 Nxa4 23. bxa4 d5 24. Nxc6 Rac8 25. Nd4 Bd7 26. Bxd5 Bxa4 27. Rc1 Rcd8 28. Bb3 Bxb3 29. Nxb3 Re2! leads to an inferior endgame to Black rooks activity. <4> 20. Bxe4 Bd5!
<4.1> 21. Bxd5 Rxe1 22. Raxe1 cxd5 23. cxd6 Bc3 where Black has an overwhelming advantage. Note that Black can play 20. ... Bf5?! with a similar idea: 21. Bxf5 Rxe1 22. Raxe1 gxf5 23. cxd6 Bc3. But this second version is much better for White, just compare the two diagrams: 
click for larger view
click for larger viewIn addition to Black's much better pawn structure on the Kingside the black pawn on d-file makes defending/advancing the White d6-pawn much more difficult. <4.2> 21. Qb4! (combining the ideas of <4.1> & <3>) 21. ... Qxb4 22. Nxb4 Bxe4 23. Rae1 a5! (the knight on b4 is trapped!) 24. Nb6 (24. cxd6 axb4 25. d7 Re7 won't help, for example: 26. Bb6 Bxg2+! 27. Kxg2 Rxd7 and Black is a healthy pawn up) 24. ... axb4 25. Nxa8 d5! 26. Nb6 Bc3 
click for larger viewBlack is a full exchange down but his domineering minor pieces give him a sizable advantage. |
|
| Nov-21-11 | | pablo333: 11... a6 can't be right, black lags behind in development after this. |
|
 |
|
< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> |
Kibitzing
sac: 21.Nxb4