|
< Earlier Kibitzing · PAGE 26 OF 26 ·
Later Kibitzing> |
| Jul-14-13 | | DWINS: <al wazir, newzild, aqb2002>, I believe that 69.Kf4 does win. After Black plays 69...Ke2, Stockfish 3 evaluates 70.g5 with an advantage of (2.62), 70.Kg3 (0.68) and everything else (0.00). I originally gave 70.g5 hxg5+ 71.Kxg5, when the Nalimov endgame tablebase calls it a draw after 71...Kf2. However, instead of recapturing with 71.Kxg5, White wins by playing 71.Kg4. Therefore, Blacks best move is 70...Kf2, but White still wins after 71.g3 h5 72.g6 which Stockfish 3 evaluates as (88.35). |
|
| Jul-14-13 | | mistreaver: Sunday. White to play. Insane. 69?
To my patzer mind this looks just like a winning endgame for white.
But he has to be careful. The natural continuation:
69 Kf4 Ke2
70 Kf5 Kf2
71 Kg6 Kxg2
72 Kxg7 Kg3
73 Kxh6 a2!
leads only to a draw.
So
69 g5!!
I think this is correct move, the point is of course undoubling of pawns but also either doubling black's g pawns, either placing a pawn on g6 and
giving the move to black while the kings are in opposition.
Black can take or refuse:
A)
69 ... hxg5
70 Kg4 Ke3
71 Kxg5 Kf2
72 g4 Kg3
73 Kh5 g6+
74 Kg5 Kf3
75 Kh4
and i think white should gradually win this, he can gain as many tempy as he wants, and black will have to give some grounds.
B)
69 ... h5 is even worse i would say
70 g6 Ke2
71 Kg3
and it is matter of technique.
I am not 100% certain that this is the right winning move order, but i don't see how else can white make any progress whatsoeer.
Time to check.
-----------------
Hmmm, I slightly considered the move 70 g3! in order to keep the oposition and give black the move, but i thought that it makes no difference whatsoever. It requires more then 10-15 minutes anyway to penetrate into secrets of this very simple yet complex endgame. |
|
Jul-14-13
 | | pawn to QB4: <To my patzer mind this looks just like a winning endgame for white.> hardly a patzer mind given the subsequent analysis. But I was there at the time and everybody around me was wondering what Carlsen was doing, playing on. I wonder how far up the food chain the view persisted...possibly all the way to Kramnik. |
|
| Jul-14-13 | | James D Flynn: White must create a passed pawn on the K-side. It will not matter whether his K is in front of or behind that pawn because he can always lose a move with his B. Candidates 69.Kf4, g5.
69.Kf4 g6 70.Ke5 h5 71.g5 Ke3(if h4(fixing the white pawn on g2) 72.Kf6 Ke3 73.Kxg6 Kf2 74.Bd5 a2 75.Bxa2 Kxg2 76.Kf7 h3 77.g6 h2 78 g7 h1=Q 79.Bd5+ winning the Black Q with a simple K and Q endgame win).
How could Black possibly draw this endgame? His advantages are the Pawn on a3 and the fact that White’s K-side pawns are doubled. If he could bring his K to h8 how could White make progress? Only by placing his K on f6 and forcing Kh7 when he can play g6+. Since the B can always lose a move White can always force g6+. Black’s position looks hopeless Let’s see what defensive ideas Carlsen comes up with. |
|
| Jul-14-13 | | James D Flynn: I am at a loss to understand why Kramnik chose to allow his K to be trapped behind his pawns. The only annotation indicates 69.g5 wins. I didn't analyse that because in principle I think White should avoid exchanging pawns, but I can believe it wins and I remain sure 69.Kf4 wins. |
|
| Jul-14-13 | | Marmot PFL: I found 69 g5 by a process of elimination but made a mistake a few moves later and the game was drawn. The win is very difficult so I would not Kramnik for missing it. |
|
Jul-14-13
 | | chrisowen: I go figure bus effect back it now ledge in hack eg, cuffed sidesteps 69.Kg3? allow black king cane,
ground to make up the distances 69...Ke3 angage,
whips around light still enacting g5 being farm,
banded big gains nifty pawn doffed mycap see free,
aintoff 70.Kh4 Kf2 and black draws sugar missed,
doublette 69.g5 or 70.g5 when energy comes to life,
whites camp I mean ogle cadences 69...Kc3 70.Kf4,
and angle light ring bate ar find go 70...Kd4 etc,
pinball a 71.g4 weight is over ignoble success ah,
good maneovre e6 coralled in g5 bats in my belfry for black a oomph g-pawn can't be stopped I guess, following 69.g5 h5 now f4 l0 wins or betcha again,
see fed cobbled 70.Kf4 or g3 calms in a tour,
fateffuls in good believe it is ok etc doing go be true delight race in light flash path to victory, 70.Kf4 Ke2 71.Kf5 72.Kf2 oh bind d5 one changes a,
71...Ke3 72.Kg6 in stiff and true h5 clogging get,
swallowed in band off aid away it draw in kingg3,
local crowd pieces the puzzle cracks creakings,
69.g5 or 69.Kf4 goofed see g3 ah hastle in e3,
koinus h4 and chance has got away it just dandy be,
finish a low tiffin cakewalks 69.Kg3 Ke3 70.Kh4,
hogtied off... |
|
Jul-14-13
 | | Jimfromprovidence: I found two ways off of the same branch
69 g5 hxg5 70 g3 Kd4 71 Kg4 Ke3 72 Kxg5 Kf2 73 Kf4.
click for larger viewThe second way changes the move order slightly and is much more involved. 69 g5 hxg5 70 Kg4 Ke2 71 g3 Kf2.
click for larger viewIn this second case white must move his bishop along the diagonal until he can play Bd5, denying black's king access to the g2 square. This forces black's king to move two squares from the pawn allowing white to both play Kxg5 and protect his lone pawn. |
|
| Jul-14-13 | | SuperPatzer77: < Marmot PFL: I found 69 g5 by a process of elimination but made a mistake a few moves later and the game was drawn. The win is very difficult so I would not Kramnik for missing it. > Very true, <Marmot PFL>. I'd like to use the Nalimov tablebases for this difficult endgame. Not easy to find a win. We all overlook the 69. g5!! the Stockfish computer overlooks as well. SuperPatzer77 |
|
| Jul-14-13 | | SimonWebbsTiger: I am not sure what criteria cg.com uses when picking puzzles. This endgame is just too well known (well, if you follow tourney life and read the usual sources like chessbase and NIC). There have been a number of instances of puzzle positions which are "solved" in a millisecond if you have the erudition. Incidentally, when GMs blunder like this, one must remember they are alone at the board, playing a tough nervous battle for hours on end. The mistakes will arise so one should never be harsh on them IMO. I guess a brownie point for me for knowing this endgame, having followed the game and postmortem at the time; and a brownie point for those who have never seen it before and found what Kramnik missed. |
|
| Jul-14-13 | | devere: 69. Kf4 Ke2 70. Kf5 Kf2 71. Bd5 Kg3! is a draw. White is in zugzwang. |
|
Jul-14-13
 | | Once: <SimonWebbsTiger> I guess it depends on what you think CG.com are trying to achieve with these puzzles. Are they looking for a pure competitive test of our solving abilities, or perhaps simply an opportunity for us to talk? I think they sometimes choose interesting positions simply because they are interesting, not because they are a test of calculation. |
|
| Jul-14-13 | | whiteshark: |
|
Jul-14-13
| | Jimfromprovidence: <Eyal> did the analysis back in December, 2010, but I have not seen any kibitzer response today that explains that after 69 g5 hxg5 Kg4 Ke3 71 Kxg5? Kf2 72 g4 the position is a table base draw.
click for larger view |
|
| Jul-14-13 | | ceebo: 69.Kf4 is also a winning move. It is very tricky because the winning line involves giving black a passed h-pawn which is very difficult to capture without allowing the black king to get to f8 with a resulting fortress. Here is the line:
69.Kf4 Ke2 70.g5 Kf2 71.g3 h5 72.g6 Kg2 73.Bc4 Kh1 74.Bd5+ Kh2 75.Kf3 Kg1 76.Bc4 Kh2 77.g4 h4 78.Kf2 Kh3 79.Be6 Kh2 80.g5 Kh1 81.Bf7 Kh2 82.Bd5 Kh3 83.Kf3 Kh2 84.Kg4 h3 85.Kh4 and white wins 
click for larger viewThe computer says that after 69.Kf4 white can force a winning 5-man ending by move 97. |
|
| Jul-14-13 | | thegoldenband: I'm quite sure I would've played 69. g5 in this position -- and quite sure I would not have found 70. g3! |
|
| Jul-14-13 | | DWINS: <SuperPatzer77:
Not easy to find a win. We all overlook the 69. g5!! the Stockfish computer overlooks as well.> Stockfish 3 finds 69.g5!! within 3 seconds. It evaluates it as (88.94) at depth 39. |
|
| Jul-14-13 | | waustad: I must have remembered it from when they played since I thought of g5 immediately though I wasn't sure of the lines that made it good. It looked like one of those where creating a weakness was more valuable than the pawn, or alternatively getting a pawn that far foreward would be nice. It seemed like a Smyslov sort of thing to do. |
|
| Jul-14-13 | | devere: So either 69. g5 or 69. Kf4 followed by 70. g5 can win, as long as White plays g3 before recapturing the pawn on g5. For any Grandmaster who saw all this over the board it would have been a career highlight; similar to Shirov's famous ... Bh3!! against Topalov (Linares 1998). It's a very nice problem. |
|
| Jul-14-13 | | SuperPatzer77: < DWINS: <SuperPatzer77:
Not easy to find a win. We all overlook the 69. g5!! the Stockfish computer overlooks as well.> Stockfish 3 finds 69.g5!! within 3 seconds. It evaluates it as (88.94) at depth 39. > <DWINS> Yes, but the old version of Stockfish did overlook 69. g5!!! Now the latest version of Stockfish finds this winning move later on. This is one of my favorite endgames - It is too tough to solve. Poor Kramnik didn't win this game against Magnus Carlsen of Norway. SuperPatzer77 |
|
| Jul-14-13 | | Amarande: <James D Flynn: How could Black possibly draw this endgame?> That's how I managed to find g5 here, though oddly enough the drawing resource I sought to guard against doesn't really seem to be an actual threat. Specifically, rather than thinking of ... g5 (which seems to create difficulty because it blocks up the White King and pawns), I was actually thinking along Bishop-ending lines: namely, that I have an LSB and King's side pawns, so the main thing to avoid would be to somehow be stuck with an h-pawn in the end due to having the wrong Bishop. As such, I thought ... h5 at some point would be Black's chief ruse, so that if I captured and then advanced the pawns normally to obtain a passed pawn (candidate first) I would be trading off the g-pawns and end up with an h-pawn. Thus, g5 occurred to me, as after this Black can no longer foist an h-pawn upon me, and thus the ending seemed to be a clear win as long as Black could not enforce the win of the other g-pawn. (This is apparently indeed how the win goes, although I did not recognise that it would be quite as complicated as other posters have noted.) (And of course, ... h5 doesn't really help Black anyway: White will get stuck with the wrong RP, but White will easily capture the g7-pawn, and Black won't be able to get back to the corner in time as necessary to actually draw such endings ...) |
|
| Jul-14-13 | | TomOhio: Wow... I saw g5 immediately and never even considered anything else. What an odd choice White made. |
|
| Jul-15-13 | | RandomVisitor: The position after 68...Kd3 is <clearly> a mate in 30. How stupid of me not to have seen this.
click for larger viewRybka 4.1 x64: <TB6> <[+M30] d=36 69.g5 hxg5 70.Ba2> Kd4 71.Kg4 Ke3 72.g3 Kd4 73.Kxg5 Ke3 74.Kg4 Ke4 75.Be6 Ke3 76.Bg8 Ke4 77.Ba2 Ke3 78.Bb1 g6 79.Ba2 Ke4 80.Bg8 Ke3 81.Bf7 Kd2 82.Kf4 a2 83.Bxa2 |
|
Jul-03-16
| | dernier loup de T: I like to imagine that Karpov "in his prime" (1984!?) would have played 69.g5!.. (?) |
|
Apr-11-21
 | | MissScarlett: This game is still living in Kramnik's head....and he still gives the impression that the win was straightforward: <Vladimir Kramnik on his "absurd" failure to win with an extra piece against Magnus Carlsen in the 2010 London Chess Classic... and after the game he ended up getting food poisoning!> https://twitter.com/chess24com/stat... |
|
 |
|
|
< Earlier Kibitzing · PAGE 26 OF 26 ·
Later Kibitzing> |
Kibitzing
