< Earlier Kibitzing · PAGE 3 OF 3 ·
|Sep-09-08|| ||johnlspouge: Tuesday (Easy): R Kuijf vs I Rogers, 1989 (31.?)
White to play and win.
Material: N+P for B. The Black Kg8 has 2 legal moves. White has control of the d-file with the battery Rd1 and Qd3. Black has a substantial counterattack along the h-file with the battery Rh5 and Qh3. The White Ne4 can destroy Bf6, the guardian of the entry point at d8.
Candidates (31.): Nxf6+, Qd8+
Because interposition is impossible [31…Nf8 32.Qxf8#], Black can accept the sacrifice or flee.
(1) 31…Bxd8 32.Rxd8+ Kf7 33.Nfg5+ Rxg5 34.Nxg5+ Kf6 35.Nxh6 Bxh6
White has won the exchange and has R+P for N.
(2) 31…Kf7 33.Nd6+ Ke6 34.Qxc8+ Kf7 [or Ke7] 35.Qe8#
|Sep-09-08|| ||whiteshark: I saw it until 36th when either 36.Bxa7 or 36.Rd6+/37.Rxc6 will create a 3:1 pawn majority on the queenside. With ♖+♙ vs ♘ up it's an easy win.|
<johnlspouge> in your line (1) it's '35.Nxh3 Bxh3'. :D
|Sep-09-08|| ||newzild: It's interesting that the silicon monsters are suggesting 31.Nxf6+ or 31.Qe6 as (apparently) more accurate wins.|
As a carbon-based monster, the Qd8+ win was the first to spring to mind, probably because there are only one or two branches on the analysis tree. Also, I think the human mind sees the back rank + fork pattern faster.
Thus, the line I saw was this:
(or 32... Nf8 33.Rxf8++)
With an easy endgame win.
I thought this was a good Tuesday puzzle - took me about a minute or two to solve.
|Sep-09-08|| ||JG27Pyth: <MAJ<In my engine-based evaluation of puzzle difficulty, an interesting behavior occurs.|
31. Nxf6 is found at Mon/Tue level settings, but with valuation way too low to be acceptable as a solution.
31. Qd8+ is found at Wed/Thu level.
31. Qd6 is found at Fri/Sat/Sun level.
With really lots of analysis, finally 31. Nxf6 rises again to the top.>
That's really rather amazing! I'll bet computer chess programmers would find this a position for further study.
But in terms of solving today...
Qd8+ is so clean (it's the Tuesday solution IMO) -- Nxf6 is just "computer chess" IMO... who wants to risk that sort of calculation when you can pick up a rook and positionally your opponent is utterly busted. It's a straight up win... numerical evaluations be damned. Nf6 is better in silcon land, but against flesh and blood I think Qd8+ actually convinces the resignation more promptly, as here.
I found Qd8+ and chose it -- I examined Nxf6+ and thought it looked interesting but the continuation was beyond my ability to calculate. Ergo, I am human.
-- Sincerely, a pool of viscous glowing liquid hooked up to wires and the internet in the basement of a lab at MIT..
ERrOr e44or penguindasfkldfas..
Please someone disconnect me... I am suffering...
|Sep-09-08|| ||johnlspouge: <<whiteshark> wrote: <johnlspouge> in your line (1) it's '35.Nxh3 Bxh3'. :D>|
I know you are only telling me because you know it irritates me...
Thanks, <whiteshark> :>}
|Sep-09-08|| ||DarthStapler: Got it|
|Sep-09-08|| ||YouRang: I found that 31.Nxf6 gxf6 32.Qd8+ Kg7 33.Qc7+ Kh6 34.Qf7 with Be3+ to follow led to a nice attack, although I wasn't convinced that it was "the" answer.|
I see that there are several winning moves for white, and I found one of them. Not exactly a great puzzle in that regard, but it's really about finding good moves.
|Sep-09-08|| ||playground player: <YouRang, et al> I don't have a chess computer, and have never used one, but I found Nxf6+, too. I doubt it's because I have a natural ability to think like a chess computer. Nxf6+ is just a good, compelling move.|
|Sep-09-08|| ||eblunt: <jimx: I'm going to go against popular opinion and say that I think this is a really good and original Tuesday puzzle.>|
I agree. Just because there are 3 winning moves, doesn't make it a poor puzzle. There's a clear winning sequence (or 3) to be found, consider it solved if you find any of them. After all, that's what would win you the game if you were in this position OTB as white.
Give yourself an extra pat if you got all 3.
|Sep-09-08|| ||Kasputin: <Frankly: I must say, I prefer all continuations in the Nxf6 line for White. Black either gets mated, or loses lots of material, or ends up looking very bad, whereas the game continuation, yes with a pretty temporary Queen sac, gives White the exchange in a depleted game that might well end up being drawn. In a game, faced with the choice of possibly getting the variation wrong somewhere and at best ending up at Bxh3, if all works out well, on the one hand, or lots of good stuff after Nxf6, on the other hand, I would pick Nxf6.>|
I don't understand this comment. Did you manage to navigate Nxf6+ and all its complexities by visualizing the end position without moving pieces or using a computer?
If you can manage that, then how could you think that the "exchange variation" (lets call it that) might be a drawn endgame? The queens are off the board (among other things this means that white's king is very safe). Also black's queenside pawns look very weak and creating a passed pawn or two for white on the queen-side looks straightforward. What targets (even potential targets) does black have?
On top of that, with best play, black ends up being behind the exchange plus a pawn. It is all forced. Any deviation and black gets checkmated. And to me at least (and others apparently), it is easier to see clearly the forced nature of the Qd8+ play in comparison to Nxf6+
I gave up on Nxf6+. Perhaps if I had played it over the board, I might have seen the correct line, but I would not have known it was a clear win initially. I can see that Qd8+ is a win. (BTW if black declines the white queen sac on d8, there is a way to force mate without taking the c8 bishop - at least I think so).
Maybe a computer would rate other initial moves higher than Qd8+. But if all the lines done to the correct end point turn out to be plus 7,8,9 etc... then they all win clearly. And playing something that ends up being plus 7 versus plus 9 makes no real practical difference. If you could see that either or both of them will win (and you see this with some measure of clarity), then congradulate yourself. Same goes for Qd6.
|Sep-09-08|| ||kevin86: I missed this because I was looking for a mate after Qd8+. This seems to be too complicated for a Tuesday...|
|Sep-09-08|| ||ClassZPlaya: For me, without a doubt this was one of the most difficult and interesting Tuesday puzzles ever offered. I started with 31. Nxf6 and failed to find a win. Also, I couldn't find a win for White after 31. Qd8+ Bxd8 32. Rxd8+ Kf7 33. Neg5+ Kf6 34. Nxh3 Bxh3. After giving up and reading all the kibitzes posted, of course correct is 33. Nfg5+! forcing 33. ... Rxg5. First Tuesday puzzle I failed to solve in a LOOONG time!|
|Sep-09-08|| ||jpolchinski: After 31 Nxf6+ gxf6 32 Qe8+ Kg7 33 Qc7 Kh6 34 Qf7, what about ... Qf5 threatening the knight? W can still pick up the exchange via Be3+, but doesn't come out any better than the 31 Qe8+ line. Presumably the computers are finding something better, but I'll take the forcing 31 Qe8+ any time.|
|Sep-09-08|| ||MostlyAverageJoe: <eblunt: ... Just because there are 3 winning moves, doesn't make it a poor puzzle.>|
Alas, there are many more than 3 winning moves.
31. a3 is also a wining move.
Taking a pass and letting black move again is also a winning "move" - even though black is not in zugzwang.
Just about anything that does not lose a piece in a senseless manner is a winning move.
Btw, CG's official position (Daily Puzzle F.A.Q.) states:
<The goal is to find the best move>
Not the second best, not the third best. THE best.
Tuesdays usually do not present any doubts as to which move is the best.
Certainly, the puzzle has merits. I just don't think it is appropriate for Tuesday.
|Sep-09-08|| ||Once: I have seen this theme called a "reloader" - a piece makes a threat, the opponent takes that piece, another piece reloads to make the same threat (which cannot be avoided).|
In today's puzzle, we actually have a double reloader. First, we have a reloader (queen and rook) on the back rank to force the black king to f7. Then we have a knight reloader on g5 to win the black queen.
Maybe that's the theme of the week?
|Sep-09-08|| ||Dr. J: <MAJ> (or anyone): Sorry, I'm still not seeing it: after <31 Nxf6+ gxf6 32 Qe8+ Kg7 33 Qc7 Kh6 34 Qf7 Qf5 35 Be3+ Rg5*>, what's the devastating line for White that gives +19? . (*Incidentally 35 ... Nf4 36 Rd6 finishes.)|
|Sep-09-08|| ||whiteshark: <<johnlspouge > wrote: |
I know you are only telling me because you know it irritates me...>
. . . Y E S ! ! !
____( @@ )___
|Sep-09-08|| ||Once: <Dr J.>
click for larger view
Don't know about a +19, but after a couple of minutes Fritz rates this as +13.48. Here is its favourite line:
36. h4 Qxf3 37. hxg5+ Kh5 38. Qxh7+ Kg4 9. Rd6 Bf5 10. Kh2
click for larger view
And right now Fritz wants to give away the black queen with either Qg2+ or Qxe3. There may well be better lines, but this seems to be painful enough. Qh3# is threatened (amongst other unpleasantries).
|Sep-09-08|| ||zenpharaohs: VoodooMoves: "Does anyone here besides me solve these puzzles with brain power instead of computing power?"|
I think everyone solves the puzzle without using the computer first, and then goes to the computer to see what is really going on.
The game line is no guide to the "solution", so it really is useful to have a good chess engine to help. Even then, you can't just let the engine compute a move and then think that is the solution; there are some pitfalls of using chess engines that you have to look out for.
|Sep-09-08|| ||mworld: wow, toughest tuesday ever for me :( I thought that after Qd8+ the game traded down to an end game where the outcome wasn't clear (for all I knew there still could have been a draw looming)...so I discounted it and tried to get the Nxf6+ line working...but my brain started to fry thinking all the lines through|
|Sep-09-08|| ||johnlspouge: <<whiteshark> wrote: [snip] . . Y E S ! ! ! [snip]>|
Es ist klar, dass eine Antwort hier nur eine kurzfristige Lösung für ein langfristiges Problem ist :)
Where should I send the bomb?
...i.e., I like the motto on your profile :)
|Sep-09-08|| ||gawain: Yessss! I am more pleased than usual at seeing this one, though I don't know why. Maybe it's just how those knights work together to force the fork of the Black K and Q I looked at Nxf6+ first but quickly turned to the back rank check Qd8+ just to see what would happen. Things fall into place prettily then.|
|Sep-09-08|| ||SpoiltVictorianChild: I spent a while trying to make Nxf6 work, then found the line johnlspouge suggests, but kept trying to find a better knock-out blow, thinking that wasn't decisive enough for a Tuesday puzzle :P|
|Sep-09-08|| ||Dr. J: <Once> Thanks. At your diagram after <31 Nxf6+ gxf6 32 Qe8+ Kg7 33 Qc7 Kh6 34 Qf7 Qf5 35 Be3+ Rg5 36. h4 Qxf3 37. hxg5+ Kh5 38. Qxh7+ Kg4 39. Rd6 Bf5 40. Kh2> Fritz's "idea"is that if 40 ... Qe4 (or Qe2) then 41 Qh3+ Kf3 42 Qg2+ Kg4 43 f3+ Qxf3 44 Qh3# Still, if this is all best play for both sides, and it probably is, then I, for one, would have no hesitation choosing the slower but much simpler game continuation.|
|Oct-24-08|| ||patzer2: White forces Black's resignation with a winning Knight Fork combination, beginning with the Queen sham sacrifice
< Earlier Kibitzing · PAGE 3 OF 3 ·