If <45...Re8 46 Rd8 Kc7 47 Rxe8> then 48...Bxe8 and now how do you propose that white will win?

click for larger view

Instead of proposing a single move, kindly forumulate the game plan for white to win, and propose a mainline with at least 4 full moves (like I did in my second post here J Trapl vs I Hausner, 1980).

My CPUs are busy with other stuff right now, but off-hand it seems that the black should respond to Rd6 with Bc6, abandon the f-pawn, exchange the R for N+P. In other words, aim for the position shown in my first post referenced above, but with one rook fewer on each side. Now black can keep shuffling his K between a6-b6 or his B along c6-h1, and how is white going to make progress?

J Trapl vs I Hausner, 1980 (35.?)

White to play and win.

Material: N for B. The Black Ke5 has 3 legal moves. Despite the lack of pieces on the board, Ke5 is vulnerable to mating threats. Black has a passed Pg5, the superior minor piece (B for N), and possible exposure of Nd6 and Pf6 to work with; White has the advanced Pf6 and the exposed Black K position.

Candidates (35.): Rxg5+

35.Rxg5+

The examination of forcing moves pays off here: despite initial appearances, 35.Rxg5+ is feasible. Moreover, the capture removes the cover from Pg5 provides against the lawnmower mate below. The Black Ke5 has his 3 legal moves:

(1) 35…Kxd4 36.Kd2 (threatening 37.c3#)

36…Bd5 37.b4 (renewing the threat 38.c3#)

Black cannot avoid mate:

37…Rc8 38.c3+ Rxc3 39.bxc3#

(2) 35…Kxf6 36.Rf5+

and White picks up B for P after 37.Rxf3, because if necessary, Re3+ releases any pressure from the pin …Rd8.

(3) 35…Ke6 36.Rg3 (threatening 37.Rxf3)

Black cannot counterattack Nd6:

(3.1) 36…Rad8 [or Rfd8] 37.Rxf3 Rxd6 38.Re3+ Kd7 39.Re7+ Kc6 40.Rxd6 Kxd6 41.Rxb7

White has 2 extra Ps, one already passed, and an aggressive R, which should win.

(3.2) 36…Bc6 [else, the White Rs gain tempo by attacking the B from near the 1-st rank]

37.Re3+ Kxf6 [Kd7 Nxb7+ wins]

38.Rd2 (threatening lawnmower mate 39.Rf2+ Kg5 40.Rg3+ Kh4 41.Nf5+ Kh4 42.Rh2#)

38…Kg5 39.Rf2 (renewing the threat of lawnmower mate)

I saw some of the power of 36.Nc4, but thought the lawnmower mate looked more promising. Unfortunately, here I missed 39...f5

40.Rg3+ Kf6

White can capture the Pf5, but probably cannot win the game. Toga II 1.3.1 evaluates my final position

click for larger view

at +0.86 (for what machine evaluation of an endgame is worth).

It seems like there is no clear win for white after 45..Re8. I would suggest for white 45.Rf7 right away. Black cannot take the rook because of the threatened back rank mate. The plan is for white to take the f pawn and then double rooks on the f file. After which a rook goes to f8.

A sample line would be 45.Rf7 Re8 46.Rxf5 Rg8 47.Rf7 Black is still in a bind. Black can still defend the planned rook invasion on f8. But since black can't do much anyway why not walk the white king up the board where it can participate in the attack. I have a line where the white king reaches the g7 square and then I play Rf8. From there the king goes back to the queen side and wins the trapped rook after sacrificing a rook for a bishop on e8 and displacing the black king. It seems to win although it goes round in circles.

How do you get the board picture for illustration? Maybe I can clarify.

See here: FEN Help Page

FEN string looks like this:

      8/rpk5/1NbR4/p1P2p2/8/1P6/1P6 /2K5 w - -

and is converted to a chess board once you post the message. I added a space after 1P6 above to prevent the conversion; otherwise it would show up as:

click for larger view

If??...IF??.. I (and several others) clearly said <no clear win> in the first place.

<45. Red7 <Re8>

46. Rf7 Ra6 47. Rxf5 Ka7 48. Rdf4 Rxb6 49. cxb6+ Kxb6 50. Rg5

(50. Kd2 Be4 51. Rg5 Re6)

50... Re2 51. Rff5 a4 52. bxa4 Bxa4 53. Rf6+ Bc6 > This is courtesy of <hms>, apparently the evaluation hovers around +1.00 but black has a fortress.

Indeed, I can quote myself as saying "Rd8 saves black for a while".

<Anatoly21> while it is possible that you know how to win that endgame, blatant assertion <White is winning here. Its a won endgame> is not going to convince anyone.

Thanks for the proposed line, although I also asked you to <forumulate the game plan for white to win>. In your line, < 48. Rd6 Bc6 49. Rf6 Be4 50. Kd2 Ra6 51. Ke3 Ra7 52. Kd4 Bc6 53. Rf6 Kd8 57. Rxf5 etc>, black has improvement with 51...Bc2 (continuing 52. Kd4 Bxb3 53. Kc3 Bd1 54. Kc4 Be2+).

I did some exploration of my idea for Black's resistance. Suppose that black would succeed exchanging R for P+N. White would take the f-pawn and advance the king. We might get a fortress like this:

click for larger view

None of my engines (Hiarcs12.1, Toga, DeepShredder11, Rybka3, Fritz10) is making any progress in the above position (just as they did not after 45...Re8).

Not surprising, since it is most likely a draw. Rationale: white pawn on b2 is pretty much useless, so let's drop it. And let's drop the black a5 pawn to get the position to 6 pieces:

click for larger view

Nalimov says it is a draw. On the other hand, this:

click for larger view

is a win in 89, with numerous places in the wining line where white must find the single move that wins.

46. Rd7-f7 Ra7-a6
47. Rf7xf5 Kb8-a7
48. Rd4-d1 Ra6xb6
49. c5xb6+ Ka7xb6
50. Rd1-g1 Re8-e2
51. Rg1-g5 a5-a4
52. b3xa4 Bc6xa4
53. b2-b3 Ba4-c6
54. Rf5-e5 Re2xe5
55. Rg5xe5 Kb6-a6
56. Kc1-d2 Ka6-b6
57. Kd2-d3 Kb6-a6
58. Kd3-d4 Ka6-b6
59. Re5-f5 Bc6-g2
60. Rf5-c5 Bg2-c6
61. b3-b4 Bc6-g2
62. Kd4-e3 Bg2-h1
63. Ke3-f4 Bh1-c6
64. Kf4-f5 Bc6-h1
65. Kf5-e5 Bh1-c6
66. Ke5-d6 Bc6-g2
67. Rc5-a5 Bg2-f3

And Black draws. Apparently I was wrong. I was pretty sure there was a win here, both from my own instincts and the fact that it was CG puzzle, but apparently not.

You claim <White is winning here. Its a won endgame> and back your 'analysis' up with <If you want to know how to win this position you should be asking Fritz or Dvoretsky's manual:>

I show people the same respect they show me, if anyone here was being a <jerk> it was you.

I think most people here would agree your initial posts were condescending to say the least. That said, I see you have admitted your mistake and since I am only interested in the chess anyway I am just happy we have reached a consensus.

The legacy of Chess Analysis is a history of our perceptions, our assumptions, even our biases as to style.

This was a pretty puzzle position. White's intuitive vision (NOT precise calculation) was spot on. The probabilities of maintaining the upper hand were excellent. Victory in Chess often requires imagination and risk-taking which may border upon speculation.

But above all is its underlying synergy - which rarely permits us to 'see' all aspects and nuances....

Here OTB the key move had enough psychological power to assure a certain momentum...and the point is during the game itself - it worked.

By way of comparison look at Anand's recent win against Kramnik in game 3 of the current WC. From a teaching point of view White gives up valuable tempii to gain a 2 pawn advantage while permitting an obvious weakening of the King's position. Black in turn permitted a potentially fatal weakening of the KIng...in short what a game!!

Kramnik cracked under the pressure. Humorless analysts with their computers may try to prove that the game could be drawn or that Anand might even have lost... but what would be the point? Rg4 was a beautiful move: one might say designed to play the player ...and it worked!!

Incorrect.

45. Red7 was ok because after 45...Re8 white can still take time out for 46. Rf7 and the f pawn drops.