| Apr-23-20 | | Cheapo by the Dozen: My thinking process was roughly:
There's no way to force the queening of the c pawn. In particular, Black's king can arrive faster than White's can. It would be bad if Black were able to safely get his king to c7, blockade and win the pawn. Let's try to avert that by forcing his knight to the actual blockade square instead. f4/f5 is the natural way to do that. Temporarily losing a kingside pawn is fine.
Immediately pushing to a4 is pointless.
The game line emerged at that point. |
|
Apr-23-20
 | | agb2002: White has a bishop and a pawn for a knight.
The knight is overburdened with the defense of c7 and the g-pawn. Therefore, 35.f4 exf4 (35... f6 36.Bd5 Kf7 37.c7 wins) 36.exf4 gxf4 37.Bf5 Nc7 38.Kf2 and White is winning because the black king is tied to the h-pawn, the knight to the c-pawn but the white king is free since the bishop controls the b-pawn. |
|
| Apr-23-20 | | Walter Glattke: 35.Bf5 Nc7 36.f4 exf4 37.exf4 gxf4 same position 38.Kf2 wins. 38.-Nd5 39.Kf3 b5 40.Kf3 Nc7 41.Kxf4 Kg7 42.Ke5 b4 43.Kd6 Ne6 44.e.g.c7 Nxc7 45.Kxc7 b3 46.h4 Kf6 47.Kd6 b2 48.Bb1 Kg7 49.Ke7 f6!? 50.h5 other possibilities B) 44.Bxe6 fxe6 45.c7 b3 46.c8Q b2 47.Qc3+ C) If 35.f4 f6 36.Bd5 Kf7 also 37.f5 wins, but less good: 37.-Ke7 38.fxe6 Kd8 39.Kf2 b5 40.Ke2 b4 41.Bb3 "Zugzwang". f5 42.gxf5 etc. |
|
| Apr-23-20 | | mike1: Followed Walter Glattke's lines above, but found out (too late) that black can play 37.f6 instead of taking f4 and it is not as smooth as it looks |
|
| Apr-23-20 | | Brenin: Unlike Oparin, <Cheapo by the Dozen> and <agb2002>, I went for 35 h4 first, forcing 35 ... gxh (35 ... f6 36 Bd5; otherwise 36 h5 wins) and followed by 36 f4 exf 37 exf, with a similar strategy of tying the N to blocking the c pawn, advancing the K to pick up the Black h pawn, and then forcing the K-side pawns forward. I think it also works this way. |
|
| Apr-23-20 | | Gryz: The crucial point of this puzzle is:
Do I first play Bf5 and then f2-f4 ? Or do I first play f2-f4, then Bf5 ? I chose to play Bf5 first. That's wrong. The reason is this: if you first play 35 Bf5, Nc7, and 36 f2-f4 next, then Black does not have to trade both pawns. He can play 37 ..., f6, supporting the pawn on g5. White advantage is +2.5 then. If you play 35 f2-f4 first, then Black can't play 36 ... f6. Because White plays Bd5 ! Pinning Ne6. Black can only play Kf7 to save his Knight. But when he does, his Knight is still pinned. And white can then play c6-c7 and promote the c-pawn. So in this line, playing f7-f6 means black either loses his Knight, and/or the c-pawn promotes. I think this is the crux of the puzzle. I didn't see it. I don't get a cookie today. :( |
|
| Apr-23-20 | | WorstPlayerEver: <Brenin> Although slower, it works as well: 35. h4 gxh4 36. Bd5 Nc7 37. Bb3 Kf8 38. Kh2 Ke7 39. Kh3 b5 40. Kxh4 Ne8 41. Kh5 Nd6 42. g5 
click for larger view42... Ne8 43. f4 exf4 44. exf4 Ng7+ 45. Kg4 Ne8 46. Kf5 f6 47. gxf6+ Nxf6 48. c7 Ne8 49. c8=N I can't get a diagram here, because cg automatically promotes to a Queen. Then it's a draw after 49. c8=Q Nd6+: 
click for larger view |
|
Apr-23-20
 | | scormus: I went first for 35 f4 provoking the crucial pawn exchanges, but then wondered about Bf5 first (... Nc7 36 f4 etc). I thought they both were winning as one would transpose into the other. Checking with the engine I see they both win, but 35 f4 is to be preferred. After 35 Bf5 Nc7 36 f4 exf4 37 exf4 B has ..... f6 (38 fxg5 fxg5 39 Kg2) It's still winning for W (+3.0) but B maintains g5 so the there's a bit more work than in the text line where B cannot hold g5 since 37 .... f6 would invite 38 Bd5 1-0 It's quite instructive to realise that apparently transposing sequences don't necessarily transpose because the order can affect whether there is a mitigating move for the defender. |
|
| Apr-23-20 | | TheaN: I'm not sure but I think a lot of moves actually win. 35.f4 is in fact the best, but the thinking process may be more important for this puzzle: 1) The knight's stuck to Pc6 until the Black king can round it up. 2) Because c is further in than b and the bishop's aiming at b1 Black has no proper counter opportunities. 3) This means White will have to create an additional weakness on the king side. 35.f4 is, in hindsight, the easiest to achieve this as it creates a passed h-pawn. Typically, I went with the opposite route, essentially activating the White king instead. <35.h4?! gxh4 36.f4 +-> creating a majority. 36....exf4 37.exf4 just helps White but <36....Nc7> does allow <37.fxe5>; Black will have to go to the center to create activity. The king side fails after 37....Kg7 38.g5 +-:
click for larger viewInstead <37....Kf8> is more tricky as stealthily Black's creating two strengths also. White did not yet have time to pick up h4, so after an eventual Ke7 with Ne6, Bxe6 <loses> because Black has two passers and is in the square of the c-pawn. If we just allow the Black king to move forward, it probably draws. I was intending to create the same blockade with 38.Bf5 and 39.g5 but that's too slow after 39....b5!, forcing the White king into the b-pawn's square, else 40....Ne6 draws at least. Instead, <38.Bf5 Ke7 39.Kg2! b5 40.Kh3 b4>:
click for larger view
White cannot allow the b-pawn any further, or Black will have the same threats as White. So <41.Bc2! Ke6 42.Bb3+ Kxe5 43.Bxf7>:
click for larger view
Didn't go as far as this, I'll be honest. Here, White's on a slippery slope but wins after <43....Ne6 44.Kxh4 Kf6> required, as after 44....b3 45.c7 Nxc7 is forced: 45....b2? 46.c8Q b1Q?? 47.Qxe6#. But after 46.Bxb3 Black has no time to pick to e3: 46....Ke4 47.Bf7 +-. <45.Bg8 b3 46.g5+ Ke7 47.c7 Nxc7> same deal here without mate: 47....b2 48.c8Q b1Q 49.Qxe6+ +- but otherwise lost. <48.Bxb3 +->:
click for larger view
Won, but I concur 35.f4 might be easier :> |
|
| Apr-23-20 | | Walter Glattke: mike1/scormus are right so far, it is harder to play 37-.f6 38.fxg6 fg6 39.Kf2 Kg7 40.Kf3 Ke5 41.Kg3 Nd5 42.h4 and several variations to follow or to be before, but I accept this 3.0, looks pretty fitting to me. |
|
| Apr-23-20 | | Vermit: We need more endgames in our puzzles |
|
| Apr-23-20 | | mel gibson: The chessboard is too small -
I have to go CTRL ++++ to get
it to a reasonable size.
The site is also far too slow -
it took about 30 seconds to load this page.
At least the large black bars are gone. |
|
| Apr-23-20 | | Brenin: Thanks, <WorstPlayerEver>, I'm glad to see that while I didn't find the best solution, I wasn't entirely wrong. The under-promotion with check in your line is very neat. Thanks, also to <scormus>, <TheaN> and others for instructive analysis (much more helpful than a stream of SF output). I agree with <Vermit> that more endgame puzzles would be good. |
|
|
|
|
Kibitzing
