Some initial positional analysis shows that the threat of e7 and queening is huge because there is only on piece defending e7 the bishop on f6 and one piece on the back rank; the Rook. Also the h-file is huge concern to Black because white can get the Queen or Rook on d1 (via d3) over to attack easily. Black does have some possible counter play with cxb2 but right now which has this covered with Bishop and Queen the only other possible sticking point is the Black Knight which is holding almost everything together and is threatening Nxf2 and possibly later c3 with forks against Rooks and Queens.

The key has to be the Black Bishop and that e7 square whilst threatening a direct attack on the king (becuase the Black pieces are just to far away to help!) The only move that threatens both is

17. Ne5.

17...B*e5? The black bishop cannot take becuase the immediate e7 is nasty and Nf6 doesn't help after e8=Q and the threat of back rank mate looms large

17...Qc7 is the only other option to protect against e7 but now White can directly attack the King

18. Ng6+ hxg6 the only move that keeps black alive but the h-file is now open

The problem is how to make use of it. White has a couple of choices Qg4 or Rd3. The problem with moving the Queen is that she may still be required for e7. Moving the Queen also fails to develop another White pieces which is required to prise away the Knight or Bishop which are now required to protect the h-file or king.

19.Rd3 threatens h3 mate just as well

Blacks options are limited because now the Black Bishop is serious overloaded with helping to defend e7 and the King via h6

19...Bg5 is bad because of 20.Qxe4...21. BxB...22.Rh3#

19. Ng5 is a little better protecting h3 but

20.Bxg5 Bxg5 and White can now play e7 becuase the Black Bishop cannot take the pawn becuase of Rh3#.

This position looks great for White - two passed pawns on the seventh rank lots of mate threats and the black peices are still of no help.

Summary: I'd play Ne5 confidently

In the opening, as black, I certainly would have closed up the game some with c4! on my tenth or eleventh move - this keeps the d-file closed and leaves white's center pawns on squares his pieces would like to occupy, thereby denying him a lot of immediate activity.

Meanwhile, with white I should open the game with dxc5 on my tenth or eleventh move to open the d-file and activate my pieces as much as possible. Black's 11...b4? is optimistic but unprincipled and ultimately wrong, a real gift to a brilliant player like Georgiev.

However, it seems to me, even Georgiev made at least one identifiable error. White's 14.e5? is a rotten move that should have spoiled his attack. After 14...Ne4! 15.exf7+ Rxf7! is not nearly as dangerous for black as the game - black gives back material to ease the danger to his king.

The immediate 14.exf7+! seems much stronger because 14...Rxf7?? runs into 15.Ng5! which is not possible in the previous line because black's pieces control g5. Black is *forced* into the line which occurs in the game (14...Kh8 15.e5). Move order is important!!

Kiril Georgiev vs Dlugy, 1983 (17?)

White to play and win.

Material: 2Ps for N, but the Black Q-side cannot develop: both Nb8-c6 and Nb8-d7 drop a piece. Thus, White is effectively up a R. The Black Kh8 is stalemated. Moreover, White has a potent K-side attack spearheaded by Pe6 and Pf7, backed by Bb3. The White Rd1 has an open file, and the White Bc1, Qe2, and Nf3 require activation, but have immediate access to the Black K-side. The Black counterattack is dynamic but unfocused. White has some tenderness in Pf2, which is protected by Qe2 and Kg1 and attacked directly by Ne4 and indirectly by Qb6 behind Pc5. Unless watched, c5-c4 could therefore embarrass White, although he has the resource Bc1-e3. Black controls the long diagonals, and Bb7 supports Ne4, which in turn supports the attack on Pb2 by Bf6 and Pc3. If White responds to c5-c4 with Bxc4, Qb3 also supports cxb2. The Black Bf6 is essential to preventing e6-e7, which suggests the following candidate, to seek a check against the stalemated Black Kh8.

Candidates (17.): Ne5

17.Ne5 (threatening 18.Ng6+ hxg6 19.Rd3 and mate along the h-file)

The candidate is typical of the "deep" Sunday puzzles: it activates pieces, namely, Nf3 and Qe2. Beyond the tactical threat to the stalemated Kh8, Black cannot let the positional threats in Nf3-e5 go unchallenged: Ne5 blocks Bf3 controlling the diagonal a1-h8 and supporting cxb2, Black's major Q-side threat.

17…Bxe5 18.e7 (threatening 19.exf8=Q# or 19.e8=Q)

Black must start to give the Greek gifts back, to regain control of his back rank.

(1) 18…Nc6 [Bd6 19.e8=Q] 19.exf8+ Rxf8 20.Qxe4

White has R+P for B, and the Black threats 20…cxb2 and 20…c4 are now nullified, because of

(1.1) 20…cxb2 21.Bxb2 Bxb2 22.Qe8

(1.2) 20…c4 21.Be3

and without forking Ra1 and Bc1, …cxb2 has little venom.

(2) 18…Nd7 19.Rxd7

After a R drops, White has R+P for B, and again, the Black threats 20…cxb2 and 20…c4 are nullified.

(2.1) 19…cxb2 20.Bxb2 Bxb2 21.Qxb2 c4

22.e8=Q Raxe8 [else, down material]

23.fxe8=Q Rxe8 [else, down material] 24.Qxg7#

(1.2) 19…c4 20.exf8+ Rxf8 21.Be3

and because the escape clause Rxb7 is available, White must emerge at least a P ahead with an attack against the Black Kh8.

Alas, RV's Rybka found a most likely drawing line quickly. So did Hiarcs, but I looked at the puzzle too late to make the analysis worth repeating.

Toga II 1.3.1 agrees with <RandomVisitor>'s best variation. If the computers are correct, today's puzzle was "flawed" - which takes nothing away from its interest, of course.

Thanks for the correction, <MAJ>. Humans and computers have different modes of analysis: in the past I have taken comfort from your posts indicating that we humans are still more effective than computers at analyzing some chess positions.

click for larger view

<21-ply +1.51 17.Rb1> Nc6 18.Qxe4 Nb4 19.Qf5 Bxf3 20.Qxf3 c2 21.Bxc2 Nxc2 22.Qe4 Nd4 23.e7 Bxe7 24.Qxe7 Qg6 25.Be3 Qf5 26.Qd6 Rfd8 27.f8Q+ Rxf8 28.Bxd4 cxd4 29.Qxd4 Rae8 30.f3 Qa5 31.a3 Re2 32.h4

14.exf7+ Kh8
15.e5 Ne4

[ply 20/72 time 3:51:08 value +1.20]

17.Rb1 Nc6 18.Qxe4 Nd4 19.Rxd4 Bxe4 20.Rxe4 Qc6 21.Bc2 Qd5 22.e7 Bxe7 23.Rxe7 Rxf7 24.Re3 Qxa2 25.Rxc3 Rd8 26.h3 Qa5 27.Be3 Rc7 28.Ra3 Qb5 29.Ne5 g6 30.Bd3 c4

Who'da thunk?

< <JG27Pyth> wrote: [snip] Expecting humans to play that kind of position with computer accuracy isn't fair IMO.>

It's marvelous to me the courage that a human needs to play such a position. It's something I could not have conceived a year ago. My, how we have grown :)