< Earlier Kibitzing · PAGE 3 OF 3 ·
|Dec-22-07|| ||UdayanOwen: Actually, I was planning to analyze a little more deeply than that....
so give me an hour or more, and I will put something up.|
|Dec-22-07|| ||kevin86: The first few moves of this one is easy-it is just the details of the follow up moves that make the puzzle hard.|
The theme this week is very instructive-sometimes is can be quite profitable to sacrifice a minor piece or two to open the adverse king position.
|Dec-22-07|| ||TrueBlue: I thought it was very easy. I am sure something like this works:|
12. Bxh7+ Kxh7 13. Qh5+ Kg8
14. Re3 g6 15. Qh6 Bh4 16. Rh3 g5 17. Nf3 gxf4 18. Nxh4 Re8 19. Nf5 Qf6 20.
Qxf6 exf5 21. Qh8#
|Dec-22-07|| ||johnlspouge: <UdayanOwen: I was planning to analyze a little more deeply than that...>
Analysis is not necessary to play 12.Bxh7+. On positional grounds alone, it "obviously" is the right move . I just wrote the post to practice my analysis routine and to internalize* it.|
*This is to let you know that you are not the only one familiar with psychobabble :)
|Dec-22-07|| ||zb2cr: Okay, what Vladimir Vukovic calls the "classic Bishop sacrifice" is on order at h7. I was able to see the first 6 moves, for White and Black, but after I reached the position after Black's 18th, didn't see any way of carrying on. At that range, even with all forced moves, my board vision gets very cloudy. |
I'm going to give myself 1/2 credit.
Oh, and for those wondering why Black resigned after move 26. White is threatening 27. Rxg7+. Black can't allow this. If he does nothing to protect g7, then 27. Rxg7+ forces 27. ... Rg7 (27. ... Kf8??; 28.Rxg8# or 27. ... Ke8; 28. Rxg8+, Bf8; 29. Rxf8#). Then White plays 28. Qxg7+, Ke8; 29. Qg8+, Bf8; 30. Qxf8#.
Returning to Black's 26th move, his only way to protect g7 is to play 26. ... Bf8. But this opens him up to 27. Qg6+, Ke7 (forced); 28. Bg5#.
|Dec-22-07|| ||Jimfromprovidence: I think black can escape with 15...Qc7?!
|Dec-22-07|| ||patzer2: Notice how 21. Rg6! would have decisively exploited the pin on g7 (my post above) to prepare to deflect the King and win the unprotected Rook.|
|Dec-22-07|| ||UdayanOwen: 12.Bxh7+ Kxh7 13.Qh5+ Kg8 14.Re3 (14.Nf3 f6 leads nowhere).|
White's principal threat is Re3-h3 and mating on the h-file. Black will have to move the f-pawn on this move or the next to avoid this. This means he can play either 14...f6 or f5 now, or play one of many other moves first. Essentially, whichever move order happens will transpose to the other, so let's we'll give him two options now:
Against both of these f-pawn moves, the only other plausible candidates 15...Rg3 or Bh6 don't seem to hold up under analysis.
What I did next, rather than calculating what will happen after any number of possible black moves, was to first analyse what white is threatening, and only THEN work out what black moves will deal with that....
I don't have time to explain properly, but I came to the conclusion that there are pretty much no moves for black in either line that prevent the forcing, material recovering variations that are threatened by white (and several moves actually make things worse in my concrete analyses, including any piece to d7 (this square is needed later for the king) 15...Qe8 (blocking the king's escape route to d7) and 15...Re8 (16.Qh7+ Kf7 [16...Kf8 17.Rg3 ] 17.Bh6 Bf8 18.Rg3 )
So my chosen move is 15...Rb8, since when the smoke clears in the best forcing variations I can find for white, black is still in the game and this move, activating the rook and hitting the b-pawn, is the most useful.
14...f5 15.Rh3 Rb8 16.Qh7+ 16...Kf7 17.Bh6. Playing against g7 in this way is the only plan that I can find that leads to winning chances after 12.Bxh7. White will win g7 (and if my analysis is correct, an exchange), because attempts to hold it lose:
17...Rg8 18.Rg3 Bf8 (18...Qf8 19.Qg6#) 19.Qg6 Ke7 20.Bg5
Basically then, black's plan is to escape to d7 with a minimum of damage and hopefully with enough material to win or draw.
I've looked at lots of possibilities, but I'm not going to write an exhaustive analysis of the subtle variations... white seems to get the g7 pawn and win the rook for bishop one way or another no matter what black does, leaving white with two pawns and a rook for the two bishops, with the passed h-pawn as a positional reward.
Here are some sample lines, after 17.Bh6:
17...Ke8 18.Bxg7 Kd7 (18...Rf7?? 19.Qg6, when 20.Rh8 will give white a winning material advantage no matter what black does). I can't see anything better than 19.Bf8... black's king is safe, and white hopes that his passed h-pawn and extra rook will be superior to the two bishops.... and they may well be, because the black pawn structure will not be conducive to the bishop's activity.
17...Rg8 18.Rg3, and no matter what black does on this move, he will have to give up his rook the next after 19.Bxg7 Rxg7 (since he can't allow a winning discovered check). Again, black will be able to escape to d7, and the same strategical analysis about the rook and h-pawn versus two bishops applies.
I can't find anything better for white....
This analysis is continued in my next post, where I analyze the 14...f6 line...
|Dec-22-07|| ||patzer2: <Jimfromprovidence> It's a complicated position, but 15...Qc7!? and 15...Rb8!? were better tries than the move played and might have held. My computer program shows them both equalizing (Fritz 8, 0.000 @ 15 ply), but I personally think the position is unbalanced and unclear.|
|Dec-22-07|| ||patzer2: <Udayowen> <So 14...f6 15.Qh7+ Kf7 16.Bh6> Is this after 14. Re3 in the game? If so what is wrong with 15...KxQ?|
|Dec-22-07|| ||johnlspouge: <UdayanOwen: Against both of these f-pawn moves, the only other plausible candidates 15...Rg3 or Bh6 don't seem to hold up under analysis.>|
For what it's worth, I looked at them both before settling on 15.Rh3, and I agree.
|Dec-22-07|| ||UdayanOwen: <Johnlspouge> I don't think 12.Bxh7 should automatically be played here on positional grounds. I think that there is positional justification for the sacrifice, and I don't think it would be illogical for it to be played purely on positional grounds....|
However, white has slower, less risky options like 12.Nf3 Nxd3 13.Qxd3, when without a material investment white still has a strong piece superiority and space to maneuver on the kingside, whilst the cramp for black is still likely to make defending difficult.
So if I was going to decide, I'd want to generate quite a bit of sample analysis of 12.Bxh6, to get a feel for just how easy/hard it will be for black to defend, before taking the risk.
|Dec-22-07|| ||Jimfromprovidence: <patzer2> <It's a complicated position, but 15...Qc7!? and 15...Rb8!? were better tries than the move played and might have held. My computer program shows them both equalizing (Fritz 8, 0.000 @ 15 ply), but I personally think the position is unbalanced and unclear.>|
The move 15... Rb8 looks like it leads to a draw by repetition if 16 Qg6.
I'll take a stand and say that 15...Qc7 does no worse than a draw for black.
|Dec-22-07|| ||UdayanOwen: (continued from my previous post)
After 12.Bxh7+ Kxh7 13.Qh5+ Kg8 14.Re3...
Now I will look at the move 14...f6 15.Rh3. In this variation it is important to show that 15...fxe5 fails, since it is a plausible move giving black options to defend g7 with Bf6... 16.Bxe5, when black is helpless to stop a winning sacrifice on g7 (16...Bf6 17.Qh7+ Kf7 18.Bxf6 Kxf6 [18...Qxf6 19.Rf3 ] 19.Rf3+ Ke5 [19...Kg5 20.Qxg7+ Kh5 21.Rh3 Qh4 22.g4#] 20.Ke5 Qxg7+, winning the f8 rook ).
So 14...f6 15.Rh3 Rb8 16.Qh7+ Kf7 17.Bh6. Again, trying to hold g7 loses, 18...Rg8 19.Rg3 Bf8 (19...Qf8 20.Qg6#) 20.Qg6+ Ke7 21.exf6 Kd7 (21...gxf6 22.Rxg8 Bxh6 23.Qh7+, ) 22.Qf7+ Be7 23.Rxg7, and now white will win back at least a piece on either g8 or d7, (23...Rxg7 24.fxg7 and white will queen).
Again, black will end up being forced to give up his g7 pawn and rook for the bishop, and hide his king on d7... I don't have time to write down the lines (It's 4:45am, and I've really got to start getting to bed earlier), but it's worth noting that a key difference in this f6 line is that black can try to win a pawn with 17...fxe5, but this seems too risky after 18.Bxg7 Kd8 19.Bxf8 Bxf8, when the black king seems vulnerable.
I suspect white finds something better in the game than what I have....
So let's check it out...
|Dec-22-07|| ||UdayanOwen: Thanks <patzer2> I @#$%ed up cos I missed a move, I have deleted that post, repaired it, and re-posted it.|
|Dec-22-07|| ||UdayanOwen: I think black might have gone wrong by trying to hold g7... I have analyzed some lines where black burns back some material... I'm unclear whether white has a clear advantage or what, what do people think?|
For example, the line:
12.Bxh7+ Kxh7 13.Qh5+ Kg8 14.Re3 f5 15.Rh3 Rb8 16.Qh7+ Kf7 17.Bh6 Ke8 18.Bxg7 Kd7 19.Bxf8 Qxf8.
Are the rook and two pawns, with a passed h-pawn, against the two bishops who are a bit restricted, enough to win? Possibly, I'm not sure. It's an unusual set of imbalances that I'm not sure how to assess accurately... It probably is an advantage to white, but black's not doing too badly here... Certainly not getting checkmated.
|Dec-22-07|| ||johnlspouge: <UdayanOwen: <Johnlspouge> I don't think 12.Bxh7 should automatically be played here on positional grounds.>|
With the plethora of posts giving Bxh7+ without justification fully expected, the length of my analysis embarrassed me, particularly when it was not conclusive. Its final position tied Black in knots, but had no death blow. I like the idea of "sampling" the analysis of a complicated line, however, and it makes good sense to me. Thanks for your further insights on when to calculate, and when not.
|Dec-22-07|| ||CaptGeorge: Let's see...
Positions: about even
Black Attacks: Nc5xBd3
12. Bxh2 ... Looks good, but speculative for me. I don't see a forced mate or material gain here. ..humm...
12. Re3 ... Protect d3B & prepare kingside attack
I don't see a clear line, but this looks like my best move. Time for peek!
ok... white went with the Bxh2...appears he speculated too.
|Dec-22-07|| ||Eggman: Incidentally, I see that many are referring to today's puzzle as an example of a GREEK GIFT or what Vukovic called the CLASSIC BISHOP SACRIFICE. Not so. The combination would only qualify as such if there were a White Knight on f3.|
<Patzer2> S#@%, I dunno, maybe I didn't consider 19...Qe8 or maybe or thought (horror of horrors) that 20.Ng4 would then be mate. A blunder on my part. I can't seem to make 17.Nf3 work, at least not for a victory. Frustrating.
|Dec-22-07|| ||patzer2: <eggman> That's OK. I thought the 14...f6 line completely equalized, but after 14...f6 15.Rh3 fxe5 16. Qh7+ Kf6 17. Bxe5! it looks like White gets a clear advantage. |
Perhaps <Jimfromprovidence>'s 15...Qc7! or 15...Rb8 are Black's best chance(s) at equality following the Bishop sacrifice. At least that's what the computer's trying to tell me.
|Dec-22-07|| ||alphee: Got it until 16.....♔f7 but totaly missed 17.♖g8.
The first move was to be choosen between 12.♗xh7, 12.♗h6 or 12.♕h5 that came as the first candidates and only the first one had a logical following.
This reminded be of Vukovic's Art of Attack where chapter 6 is all about the bishop sacrifice on h7 with good examples. "Timeo danaos et dona ferentes" ...
Let's see what sunday will bring.
|Dec-22-07|| ||erniecohen: Like everyone else, I'd play Bxh7 OTB, but this is a terrible puzzle:|
- The "solution" is the most obvious move;
- The "solution" doesn't appear to be ultimately winning against best defence, or even obviously better than other, quieter continuations;
- There doesn't appear to be a clear, winning solution.
|Dec-22-07|| ||Steve Case: I was move for move with white until that pawn chased the Knight on the other side of the board. After that, I just sat back and watched (-:|
|Dec-22-07|| ||Terry McCracken: < erniecohen: Like everyone else, I'd play Bxh7 OTB, but this is a terrible puzzle:
- The "solution" is the most obvious move;
- The "solution" doesn't appear to be ultimately winning against best defence, or even obviously better than other, quieter continuations; - There doesn't appear to be a clear, winning solution.>
It's a beautiful combination. White will win.
Check with deep assisted computer analysis.
White always knew he was winning and at worse would only draw.
It takes awhile and many moves before a computer sees it's winning, it's very deep.
|Dec-23-07|| ||erniecohen: Hi Terry, I'm afraid I just don't see the win if black doesn't try to defend g7 (e.g., see UdayanOwen's post). I'm not saying that white might not get an advantage, but I don't see anything close to a clear win, and that's kind of expected when you post a position as a problem.|
< Earlier Kibitzing · PAGE 3 OF 3 ·