| Aug-17-13 | | niemzo: If the pawn endgame is lost, and it seems pretty bad, this game shows a weakness of computer chess. Houdini in the official site was giving it 0.3. I think it's the horizon effect at work here or something similar. |
|
| Aug-17-13 | | Marmot PFL: Grischuk may have been in his usual time trouble or he would realize that the K + P ending is lost. Maybe he should trade rooks on move 31 and try to hold with queens on. |
|
| Aug-17-13 | | Marmot PFL: Usually a bad idea to trade pieces when opponent has a protected passed pawn. Grischuk vs Kramnik, 2013 |
|
| Aug-17-13 | | Nerwal: Grischuk and problematic pawn endgames, continued : Potkin vs Grischuk, 2011 (this time he won though). Houdini can give +0.3 all it wants. This type of position is extremely unpleasant to play for black. I suspect out of 100 games white would score 80 against equal opposition, maybe more. |
|
| Aug-17-13 | | Marmot PFL: Probably 100 out of 100...black just runs out of moves and loses material. |
|
| Aug-17-13 | | Marmot PFL: Although 38...c4 might be a chance to hold. 38... c4 39 Kd2 Kd6 40 Kc3 Kc5 is much harder than the game, if white can win it will take more work. |
|
| Aug-17-13 | | Marmot PFL: Still hopeless for black, as white can lose moves by triangulating with his king to get opposition. |
|
| Aug-17-13 | | twinlark: <21...Nb4> may have been an invitation to a draw: <22. Rxb4 Qxb4 23. Nh5+> and Black won't get away from knight and queen checks unless he takes the knight, and then he can't get away from the queen checks on the weakened black squares. |
|
| Aug-17-13 | | csmath: Classical Bc4 Gruenfeld, topical variation.
Reference game here is WC game:
Topalov vs Anand, 2010
LQL plays a new move
12. Rfd1!?
while Grischuk plays as Anand
12. ...Rc8
Grishchuk has achieved equality after the opening but he started playing for draw by exchanging pieces. 27. ... Bxe2?! [...Rc4 is more active and black has good countergame but white has a passer thus Grishchuk just wants to draw] 28. ...Qd6!?
[blocking passer with queen is positionally dubious though Grishchuk has parted with knight which was a perfect piece for blocking. White is now better] 29. ...Rc5?
30. ...Rdc8?!
[closing the scope of queen with rook and then removing queen protection is also against solid positional principles, it is always a possible pin target] These are serious though subtle positional errors.
31. Kf8?
[Yet another positional error putting the king into a double pin.] 32. Qa3!
This game is now over. You can certainly see a number of positional errors of a top GM. After the exchanges the pawn ending is lost. |
|
| Aug-17-13 | | twinlark: Comps often need a bit of time to figure out these kinds of endgames. There's no doubt the K+6P ending is dead lost for Black, as he will run into zugzwang. But Black was lost even with the queens on the board. Position after <34. Rxc5>: 
click for larger viewNow if Black recaptures with the pawn instead of the queen, <34...bxc5>, then the White queen infiltrates and supports the advance of the passed pawn: <35. Qa6> followed at some stage by Qc6 and White will win, although not as easily. Black had to keep at least one rook with <33...bxc5> instead of <33...Rxc5>, but then has an uphill battle to restrain the infiltration of White's queen. Seems like Black's c-file strategy was flawed. |
|
| Aug-17-13 | | csmath: Grishchuk has achieved equality after the opening meaning LQL has not improved Topalov's plan. However in his desire to draw Grishchuk has exchanged minor pieces, in particular it would have been advisable to keep knight [I know I would have never exchanged knight in that position with passer on d5]. Then he simply violated solid positional principles and LQL pinushed that immediately. Top 10 GM violating positional principles:
1. Do not close the scope of queen with a piece that can be pinned. 2. Do not use queen to block pawn, use minor piece instead, knight preferable. 3. Do not remove protection of queen that can be pinned. and after all the worst error:
4. Try to remember elementary pawns endings - the ending with pawns after 35 moves should be known to every master, it is elementary won. [Forget about engine evals, I can see this win on a mile, you do not need engine for that.] |
|
| Aug-17-13 | | csmath: This is second time this year already that Grishchuk gets into elementary lost pawn ending [check that funny game with Kramnik in London] I think Grishchuk has skipped some classes in 1st grade of chess school while smoking pot. :-) LQL apparently did not. |
|
| Aug-17-13 | | twinlark: <2. Do not use queen to block pawn, use minor piece instead, knight preferable.> That seems to be the problem that essentially landed him in difficulties. It was the exchange of the second rook that lost him the game. I mentioned earlier Black's c-file strategy. It seems he would have been better off seizing the c-file immediately with <29...Qc5>, instead of the problematic <29...Rc5> which merely allowed White to oppose the heavy pieces and ultimately exchange them all to a won K+P ending. Black's position after <29...Qc5>: 
click for larger viewand it's as solid as a rock. The queen no longer spends its valuable self blockading the pawn, the job now being available to the rook, and White can't oppose on the c-file, eg: if <30. Rdc2 Qc3! 31. Qxc3 Rxc3 32. Rc2 Rdc8 33. Rxc3 Rxc3> and Black is copacetic: 
click for larger view |
|
| Aug-17-13 | | csmath: On the matter of principles with heavy pieces you want to keep queen on board when you have a passer and you want rook ending when your opponent has a passer. So yeah, absolutely the second diagram looks ideal for black. |
|
Aug-18-13
 | | HeMateMe: Worthy of textbook inclusion. How to gain the opposition. |
|
| Oct-19-13 | | fgh: <Grischuk and problematic pawn endgames, continued : Potkin vs Grischuk, 2011 (this time he won though).> <This is second time this year already that Grishchuk gets into elementary lost pawn ending [check that funny game with Kramnik in London]> And another one for this "collection":
A Riazantsev vs Grischuk, 2009 |
|
Kibitzing
