This is Fritz's favourite line:

47... Qe7 48. Qd6 Nc8 49. Qb8 Qd8

click for larger view

Now white has a number of tempting options, all with evals in the of more than +4 - eg 50. ef or 50. Bd7 (my preference, blocking off the d file) or 50. e7.

Unzicker vs A Dueckstein, 1967 (47.?)

White to play and win.

Material: B for N. The Black Kg8 has 3 legal moves and is surrounded by weak light squares. The White passed Pe6 steals f7 from Kg8 and is a candidate for e8=Q. In support of e8=Q, Qd3 controls the adjacent open d-file, while Bb5 controls e8. Both sides have possible P breaks, at c4 and e5. If tactically feasible, 47.e5 opens up the Black Kg8, permitting mating threats to accelerate the promotion of Pe6. The White Kf3 is vulnerable to pointless checks from Qc7, nothing else.

Candidates (47.): e5

47.e5 (threatening 48.Qd6)

Black cannot permit 48.Qd6: Pe6-e7-e8 becomes unstoppable, whether or not Qd6 is captured.

(1) 47…Qxe5 48.Qd8+ Kh7 49.e7

[49.Qxb6 probably permits perpetual check]

<[Toga values 49.Qxb6 at only 0.75 P. This candidate provides an interesting example of when <not> to calculate.]>

Black must check, because after e8=Q, White can prevent perpetual check with Qe8-e2.

49…Qxf5+ (making Bb5-d3+ deadly) [Qc3+ 50.Bd3 is curtains]

The White Kf3 heads for c2, where the checks cease, and e8=Q finishes the game.

<[Toga confirms the game variation 50.Kg2 as best. My plan of interposing on the Q-side is also successful, but White must be careful to avoid e3 because of …Nd5+.]>

(2) 47…fxe5 48.f6 (threatening 49.e7)

(2.1) 48…Qb7+ 49.Qe4 Qxe4+ [Nd5 Bc4]

50.Kxe4 gxf6 51.e7 then 52.e8=Q

This variation is critical. Because Black cannot exchange Qs, he must submit to a K-side attack.

(2.2) 48…gxf6 49.Qg6+ Kf8 [Kh8 is similar]

[49…Qg7 50.Qxg7+ Kxg7 51.e7]

51.e7 (threatening 52.e8=Q+)

White can interpose against 51…Qb7+ with 52.Bc6 and flee against 51…e4+ with Kg2, so Black is finished.

Thats a good defense. I think even 47..Qe7 right away is hard to crack.

The best play I can find is

47...Qb7+ 48.Kf2 Qe7 49.Qd6 Nc8 50.Qd7 Qf8 51.exf6 gxf6 52.Qc7

47... Qb7+ 48. Kf2 Qe7 49. Qd6 Nc8 50. Qb8 Qd8 (forced, if you want to save the knight) 51. Bd7

click for larger view

<goodevans> <Maybe I'm missing something obvious (I quite often do!) but I'm not seeing a clear-cut win after 47 ... Qe7. 48 Qd6 Nc8 49 Qd7 Qf8 forces black into a position of almost total passivity, but I don't see the coup de grace.>

<5hrsolver> <The best play I can find is 47...Qb7+ 48.Kf2 Qe7 49.Qd6 Nc8 50.Qd7 Qf8 51.exf6 gxf6 52.Qc7 To win one of the queenside pawns. There is still a lot of play left and the win is questionable.>

The key for white in this position is not to play Qd7, but Qb8 instead. This pins the knight and queen and prevents black from playing …Qf8.

Here is the position after 47 e5 47 Qb7+ 48 Kf2 Qe7 49 Qd6 Nc8 50 Qb8.

click for larger view

If black tries 50…Qf8, then white wins with 51 e7!

click for larger view

Now black cannot play 51…Nxe7 because 52 Bc4+ wins the queen. Similarly, 51…Qxe7 loses to 52 Qxc8+.

So, 50…Qd8 is forced, which loses a piece for black for black after 51 Bd7.

click for larger view

Didn't see your 51.e7!

Now black cannot play 51…Nxe7 because 52 Bc4+ wins the queen.>

47. ♙e5!

If 47. ... ♙fxe5 48. ♙f6! ♙gxf6 49. ♕g6+ ♕g7 50. ♕xg7+ ♔xg7 51. ♙e7 and now Black cannot stop the pawn promotion.

Alternatively, 47. ... ♕xe5 48. ♕d8+ ♔h7 49. ♙e7. Now Black can put up some resistance with 49. ... ♕xf5+ but White can easily move to safety.

For example:
50. ♔e2 ♕c2+ 51. ♔f1 ♕b1+ 52. ♔f2 ♕c2+ 53. ♔g1 ♕c1+ 54. ♗f1 ♕e3+ 55. ♔h2

Please elaborate.
>

Sunday 6 December 2009

<47.?>

Target: 15:00;000
Taken: 14:17;791

Material: White piece exchange up, ♗ vs ♘

Candidates: Qd7, Nd7, e7, Bc4, Qc4, Qd5......? argh...!... <[e5!?]>

-ML-
A difficult position for both colors: White has a clear passed pawn on e6 but dealing with a doubled pawn, Black is struggling with a weaker Knight but has his position seemingly under control due to placement.

White's shot here has to have something to do with the e6-pawn, but my move has, at first sight, not a lot to do with it. After all kinds of typical piece and Queen sacs, a pawn sac seems more fruitful.

<47.e5!> White moves the weak doubled pawn (what's its worth anyway). In fact, this just looks like a dead pawn for White. Quick analysis shows, however, that Black might not at all be so happy with this move. Keeping it on the board will lead to the winning Qd6, which is the actual idea of the move, supporting the e6 pawn doing so. Without a capture, this is pretty much the winning move, if the Knight moves to b8, White will follow with 48.Qd7 winning. No, the two capture variations are the most critical, where fxe5 might be a quick end for Black.

/A\
<47....fxe5 48.f6!> of course, the idea is the breakthrough. e5† is useless, and keeping it on the board leads to e7 winning. Capturing, however, is just as bad:

<48....gxf6 49.Qg6†> interposing loses quickly, 49....Qxg7 50.Qxg7† Kxg7 51.e7 and unstoppable. Neither King move does not really matter, both Kf8 and Kh8 meet:

<49....Kh8 50.Qxf6† Kh7 51.e7 > so, Black has to capture with the Queen on move 47, making this a Sunday puzzle.

/B\
<47....Qxe5> obviously, now the Black Queen gave up the defense of d8. A critical square, compromising her own King's safety. However, Black does threaten some kind of perpetual with Qxf5†. Not important though, if White makes a new Queen unopposed.

<48.Qd8† Kh7 49.e7!> and the hunt begins.

<49....Qxf5† 50.Kg2 Qc2† (50....Qe4† 51.Kf2 Qf5† 52.Kg1 transposes) 51.Kg1! Qc1† 52.Bf1> now Black has to continue checking, as threatening g3 won't work. After 52....Qe1? 53.e8=Q Qxg3† 54.Bg2 no more checks are available.

<52....Qe3† 53.Kg2 Qe4† 54.Kh2> with the Bishop now on f1, Black is losing his checks soon:

<54....Qc2† 55.Kg1> like, now. Without c6 available and the Bishop on f1, Black can no longer check White on g1 from c2. White will promote unopposed, and I'm not seeing effective ways to create threats:

Woohooooo! Easy week though. I checked my escapes with Rybka, although 50....Qe4† 51.Kh2 is simple, Kf2 indeed transposes and wins in the same way... 51....Qh3?! 52.e8=Q is playable. I kinda missed Qb7† but this seems irrelevant, as the same threat of Qd6 remains. If 47....Qb7† 48.Kf2 Qe7 49.Qd6, Black's position is still not holdable, whatever played. I don't really need much analysis for it.

A) 47... fxe5 48.f6 (threatening 49.e7 and 50.e8=Q+)

A.1) 48... gxf6 49.Qg6+ Kf8 (49... Qg7 50.Qxg7+ Kxg7 51.e7 + -; 49... Kh8 50.Qxf6+ and 51.e7 + -) 50.Qe8+ Kg7 51.e7 and Black can't stop the pawn (51... Nd7 52.Qxd7; 51... Qc8 52.Qxc8 Nxc8 53.e8=Q).

A.2) 48... Qb7+ 49.Qe4 Qxe4+ 50.Kxe4 gxf6 51.e7 + -.

B) 47... Qxe5 48.Qd8+ Kh7 49.e7 (49.Qxb6 Qxf5+ and the white king probably can't escape the perpetual) Qxf5+ (49... Qc3+ 50.Bd3 c4 51.e8=Q Qxd3+ 52.Qxd3 cxd3 53.Qd8 + -) 50.Ke2

B.1) 50... Qg4+ 51.Kd2 Qg5+ 52.Kc2 Qf5+ 53.Bd3 + -.

B.2) 50... Qe4+ 51.Kd2 Qg2+ 52.Be2 Qxg3 53.Qd3+ + -.

B.3) 50... Qe5+ 51.Kf2 Qf5+ (51... Qb2+ 52.Be2 + -) 52.Kg1 Qb1+ 53.Bf1 + -.

B.4) 50... Qc3+ 51.Qd3 Qe5 52.e8=Q + -.

click for larger view

Black has one last trick: 51...c4! but 52 Qxc8! Qxc8 53 Bxc8 cxb3 54.Ba6 ends it.

<5hrsolver: [snip] The best play I can find is 47...Qb7+ 48.Kf2 Qe7 49.Qd6 Nc8 50.Qd7 Qf8 51.exf6 gxf6 52.Qc7 to win one of the queenside pawns. There is still a lot of play left >

I agree completely with your last comment. Playing against Crafty it was the best I could find. After 52...Ne7 threatening the f5 pawn

click for larger view

A lot of “E” level chess is based on hope. In the opening, the E player will often say to himself, “Gosh, self, I hope he doesn’t see I’m surreptitiously attacking his f-pawn.” Or take the middle game. That’s where we say, “I hope he doesn’t notice he’s about to get forked.” Yes, we could and often do substitute the heteronymic dirty word or even throw in a slant rhyme if we feel literary (“Ha! You’re about to get ------ by a fork!”) When we do so, we feel like we have something in common with our higher rated brethren. This makes us happy and gives us something else to think about when we have no idea what move to play.

Like, say, this puzzle.

See, in a position like this, with relatively few “big” pieces on the board to move and no traps to hope for, the “E” player does a few things that some of the higher rated players would kindly call “counterproductive.” Others, more blunt, less kind, less willing to give us the chess version of a hug, would call it “BSing.” What kind of BSing? I’m glad you asked.

How we look at the position depends a great deal on who we’re playing. Against a higher rated player, we might look at the board in a daze before shifting a piece of wood to relative safety, hoping the other guy gets overconfident and blunders. I mean, heck, we blunder so he might, too. But for this puzzle, I can’t do that because I’ll never solve the damn thing. So instead of facing Dueckstein, who has enjoyed watching Fischer, Tal and Keres tip their kings, I’m going to pretend I’m facing an “E” player on FICS. This is fun because it means I look first for a <check!>

When in doubt, we “E” players love to <check!>; we love it almost as much as we love exchanging pieces—it gives us something to do and it buys us time to make our next plan (another <check!>). Now, I could move Be8 and make a <check!> on the next move. It doesn’t do much because the king could just move closer toward the passed pawn but I get to go home to the wife and kid and say, “Guess what gang? I had Dueckstein in <check!>” My son would take half a breath of excitement and the wife would raise the eyebrows and wonder if something good wasn’t coming from playing this game. They would never be the wiser. But as 47.♗e8 doesn’t really do anything, we really should move on.

But I could have put Dueckstein in <check!> Okay, now move on. Another thing: we “E” players spend a great deal of time in the land of “woulda, coulda, shoulda.” As in “I woulda had my Bishop on c4 if I knew I coulda had a discovered check.” Or, “I coulda been a great chess player by now but I had a life to live and so what if my fellow CG posters are living happy lives and playing chess well. The point is, it coulda been me!” Or, “I shoulda taken up checkers.” The “E” player frets about this and then explores the idea of getting away with a trap, like putting the bishop safely on c4 so we could use the dreaded <discovered check!> Look, I’ll even think it: “I could move it there and hope he doesn’t notice 47. ♗c4” But, of course, it’s quickly followed by 47… ♔f8.

Sometimes—and this is rare—we "E" players actually apply what we have learned. Now, I don’t want you to faint dead away but I did say to myself here, “Self, you read somewhere (okay, it was Silman) that if you control the square in front of a passed pawn, you’re in good shape. Right now, black controls that square thanks to the queen but if I move 47.♗d6 he doesn’t control that square. I am blocking his control. Which is great because then that could lead to an exchange of pieces--he could take the Bishop with the knight, I could take it with my queen and I’ve got a passed pawn and a queen controlling the key square in front of it. Might this work?”

It might if the damn King wasn’t as valuable as a bishop in the endgame. (47.♗d6 ♔f8 and then what?) So, how do I keep the king out of the fray? Maybe there’s nothing going on with a piece move or a passed pawn move. Maybe there’s a quiet move somewhere. Like…

47.e5

I don’t think black wants to take this pawn. If he does, it might open up a diagonal for the queen straight to the black queen. And black doesn’t want to take it with the queen because then the white queen advances to the back rank. But if black <doesn’t> take the pawn, the white queen advances anyway. Ergo…

47.e5 c4
48.♕d6 ♕c8
49.♕xb6 cxb3
50.♕c6

In predominately closed positions such as this, it's especially important to examine the possible pawn breaks, c4 for black and e5 for white. In this position, an experienced GM would probably examine 47.e5 first without thinking twice. Unfortunately, I wasted time analyzing 47.Bd7, 47.Bd8, and 47.Bc4 before I finally saw the light:

47.e5!!

The idea is to exploit the passive position of the BQ, which must guard both e7 and d8. This pawn opens another door to access the black king.

A) 47... fxe5 48.f6! gxf6 49.Bd7! Nxd7 50.Qg6+ Kh8 (50... Kf8 51.Qf7#) 51.Qe8+ Kg7/h7 52.Qxd7+ wins.

A.1) 48... Qb7+ 49.Qe4 Qxe4+ 50.Kxe4 and the e-pawn promotes.

A.2) 48... e4+ 49.Qxe4 and the e-pawn promotes.

A.3) 48... c4 49.e7! cxd3 50.e8=Q+ Kh7 51.Bxd3+ g6 52.Qe7+ wins

A.4) 49... Qb7+ 50.Kf2 Qh1 51.e7! Qh2+ 52.Ke1 Qh1+ 53.Kd2 Qg2+ 54.Qe2 and black runs out of checks, so the e-pawn promotes.

B) 47... c4 48.Qd6! and nothing can stop the e-pawn.

C) 47... Qb7+ 48.Kf2 fxe5 (Qh1 49.e7 and black runs out of checks quickly) 49.f6 and black can do no better than transpose to A.4.

D) 47... Qxe5 48.Qd8+ Kh7 49.e7 Qxf5+ (Qc3+ 50.Qd3) 50.Ke2 Qe4+ 51.Kh2 Qc2+ 52.Kg1 Qc1+ 53.Kf2 Qc2+ 54.Be2 Qf5+ 55.Kg1 Qb1+ 56.Kh2 and white breaks the perpetual.

Plan: <nimzowitsch says when in control of the open file, the major piece need to break through to the 7th or 8th rank> In this case, the e6 pawn need the help of the queen to promote.

candidate: e5!

1: Wolfgang Unzicker - Andreas Dueckstein, Krems 1967

click for larger view

Analysis by Rybka 3 : <18-ply>

<1. (2.17): 35...Rd8> 36.Rxd8+ Qxd8 37.e5 Kg8 38.Qe3 Nc8 39.Bb7 Ne7 40.exf6 gxf6 41.Qxc5

2. (2.27): 35...Kh7