< Earlier Kibitzing · PAGE 2 OF 2 ·
|Feb-08-18|| ||latebishop: I wondered about 44...Re8 as after 45.d7 Qxe7 46.dxe8+ Qxe8 47.Qxd4 White is only a pawn up, although I guess that's enough at Grandmaster level?|
|Feb-08-18|| ||nalinw: I cheated - Stockfish says
44. ... Re8
keeping the pressure on.
|Feb-08-18|| ||agb2002: White is one pawn ahead.
Black threatens Q(R)xd6.
The first idea that comes to mind is 41.Rxd4:
A) 41... exd4 42.e5
A.1) 42... Qg7 43.e6
A.1.a) 43... Rf6 44.e7+ followed by 45.exd8=Q wins.
A.1.b) 43... Rff8 44.f6 Qg6 45.e7+ followed by exd8=Q and Rf5 with the threat Rxg5 looks winning.
A.2) 42... Qh8 43.e6 as in A.1.
A.3) 42... Rxd6 43.exd6 followed by Rd1 looks very god for White.
B) 41... Kh7 42.Rd2 + - [2P].
C) 41... Rxd6 42.Qxd6 exd4 43.Qd5 followed by e5, e6 looks winning.
|Feb-08-18|| ||diagonalley: hmmmmm... after a minute's gaze, the prospect of three connected passers for a rook showed up... theoretically that's a no-brainer... (so time to test the theory!)|
|Feb-08-18|| ||saturn2: I saw 41 Rxd4
and black gets squeezed. White gains back the rook under positive circumstances.
|Feb-08-18|| ||AlicesKnight: I saw the 41.Rxd4 break, but would like to know more about 42.... Rxd6 in reply to 42.e5. Perhaps 43.exd6 Kg7; 44.d7 is good enough (e.g. ... Rf8; 45.Re1 etc.)|
|Feb-08-18|| ||yadasampati: It took me less than 10 seconds to see Rxd4. The resulting wall of pawns after exd4 and e5 is obviously unstoppable!|
|Feb-08-18|| ||Once: That's a beautiful combination. Not too difficult to see although it takes a little calculation to be sure that black can't wriggle out. |
I liked the position after 43. e6
click for larger view
Yes, White has a pawn roller for a rook, but he also has an immediate tactic. If Black tries to move his attacked Rf7 White has the devastating threat of 44. e7+ which is both a discovered check and an attack on the other rook.
And if 43...Re6 44. f6
We can even take the black king out of the equation and the combination still works. This is the position with the black king nudged to h8 and with black to move:
click for larger view
White hasn't got a discovered check, but he will force home one of his passed pawns. There's nothing that black can do.
|Feb-08-18|| ||morfishine: After <41.Rxd4> White has a pawn roller|
|Feb-08-18|| ||thegoodanarchist: Easiest Thursday ever!|
|Feb-08-18|| ||malt: Have <41.R:d4> ed4 |
(41...Kg7 42.Rd3 Rfd7 43.Rfd1 )
42.e5 Qg7 (42...Qh6) 43.e6
|Feb-08-18|| ||mel gibson: Stockfish 8 agrees:
41. Rxd4 (41. Rxd4 (♖d1xd4 ♔g8-g7 ♖d4-d3 ♖f7-d7 ♖f1-d1
♖d8-e8 ♖d3-c3 ♕f6-f7 ♕d5-d2 ♔g7-h6 ♖c3-c6 ♖e8-f8 ♖d1-f1 b6-b5 ♕d2-c3 ♕f7-f6
♖f1-d1 ♔h6-h7 ♕c3-c5 ♕f6-f7 ♖d1-d5 ♖f8-d8 ♕c5xb5 ♔h7-g7 ♕b5-c5 ♔g7-g8
♕c5-c4 ♔g8-f8 ♕c4-c1 ♕f7-f6 ♕c1-c5 ♔f8-g8 ♖d5-d2 ♔g8-f8 ♕c5-d5) +5.64/32
score for White +5.64 depth 32
|Feb-08-18|| ||gofer: Well the rook sac looks sound...
<41 Rxd4 ...>
Ignoring it just means that black is now two pawns down, but it is probably black's best option. Accepting the sacrifice seems a bad idea.
<41 ... exd4>
<42 e5! ...>
42 ... Qe7?
43 dxe7 Rxd4
44 e8=Q+ +-
<42 ... Qg7> (Qh8/Qh6 seem to be worse)
<43 e6 ...>
43 ... Rf6?
44 e7+ +-
43 ... Rdf8?
44 e7! Re8
45 Re1 +-
43 ... Rff8?
44 f6 +-
<43 ... Qf6>
<44 e7 ...> +-
|Feb-08-18|| ||zb2cr: 41. Rxd4!, exd4; 42. e5, Qg7; 43. e6 and White will recover his Rook with a dominant-looking endgame.|
|Feb-08-18|| ||Walter Glattke: 43.-Kf8? 44.f6 Rxf6 45.e7+ Ke8 46.Rxf6
Qxf6 47.Qg8+ Kd7 48.Qxd8+ or 46.-Rd7 47.Rf8+ Qf8 48.exf8Q+ Kxf8
|Feb-08-18|| ||ventricule: For me, the hardest line to analyze was
<41. Rxd4 Rxd6 42. Qxd6 exd4 >
where it temporarily looks like white did not win anything. Exchanging queens with <43. Qxf6> and pushing the d and e pawns is tempting, but is not as clear cut as the main line.
Stronger is <43. Qd5> as <agb2002> rightly suggests, but seeing this requires a bit of work.
|Feb-08-18|| ||morfishine: To pull this off against the powerful Yusupov is truly remarkable|
|Feb-08-18|| ||Eduardo Leon: The pawns bulldozed their way through. Not difficult to see, but still beautiful.|
|Feb-08-18|| ||RKnight: To all you eagle-eyed rook-sac problem solvers: I failed to even consider 41. Rxd4. What's your secret? I solved the Tues and Wed puzzles and others too, but occasionally total blindness takes over.|
|Feb-08-18|| ||mjmorri: <RKnight> The pawn tsunami resulting after 41.Rxd4,exd4 makes it at least worth looking at.|
|Feb-08-18|| ||patzer2: I missed finding the win for today's Friday puzzle (41. ?) because I couldn't resist the positional move 41. Rc1, putting my Rook on an open file.|
Unfortunately for me, this puzzle is not one that calls for quiet positional play. My 41. Rc1 fizzles out to a level position after 41. Rc1 Qxd6 42. Qxd6 Rxd6 43. Rfd1 b5 = (0.00 @ 34 ply, Stockfish 8).
Instead, White wins with the aggressive tactical move 41. Rxd4!! exd4 42. e5 +- (+6.36 @ 31 ply, Stockfish 8) when White's combination wins back the sacrificed Rook with decisive advantage.
P.S.: According to the computer, Black's clear losing move was 27...Bf6 allowing 28. f4 +- (+1.99 @ 29 ply, Stockfish 8).
Instead, the second player can put up more resistance with 27...Ng6 28. c5 dxc5 29. Nxc5 Qc6 30. Qd5 Qxd5 31. Rxd5 Rb8 ± to +- (+1.57 @ 35 depth, Stockfish 8).
Earlier in the game, Black can apparently improve on 14...Rxe6 15. Bxf3 Rxf3 16. Ne4! ± (+1.02 @ 30 ply, Stockfish 8) with 14...Qxe6 15. Bxf3 Rxf3 16. Ne4 Qc8 17. Ba3 ⩲ (+0.64 @ 40 ply, Stockfish 8).
Though it's difficult to prove Black's Knight is "dim on the rim" after the seldom played 5...Nh6!? 6. Nge2 ⩲ (+0.35 @ 30 ply, Stockfish 8), I prefer the popular move 5...d6 = (0.00 @ 31 ply, Stockfish 8) as in Black's win in Ivanchuk vs M Mchedlishvili, 2017.
|Feb-08-18|| ||Geronimo: I could see as far as 43...Qf6 on the text moves, but only count a half-point for this elegant puzzle, as I didnít even consider the alternative line 41. Rxd4 Rxd6 42. Qxd6 exd4. Also, 44. e7 is a beautiful move, and better than recapturing the rook on f7, which I probably would have done breaking the pawn roller and losing a big part of the advantage.|
|Feb-08-18|| ||CopyBlanca: What if Black does not accept the sacrifice of the rook and just tries to keep the position blocked? Penrose gave an example of such a blocked position where if the computer declines the sacrifice it is a draw but a computer can not see that opening the position is much worse than a gain of a rook and will always take the rook. Penrose's point is that a computer does not know what is doing and that a ten year could see it is best to leave the position blocked.|
|Feb-08-18|| ||CopyBlanca: I should add that this was not Jonathon Penrose , but Roger Penrose the expert on artificial intelligence. In his example all 16 pawns opposed each other and the position was 100 per cent blocked. This position is close to his example.|
|Feb-09-18|| ||landshark: Anybody else like 43 f6? From my quick look it seems to put Black in a world of hurt....|
< Earlier Kibitzing · PAGE 2 OF 2 ·