After 36...Rg1 (as you suggested)

37.Rf3! Ra1 (if 37...Rxf3+ 38. Kxf3 Ra1 39. Rxb3! if 39...Rxa4? 40. Nxc5+)38. Rxf5 gxf5 39. Nc5+ Ke7 40. Rb7+ Ke8 41. Rc6! Nd8 42. Rxb3

You can check by yoursel, there are other sub-variations, but in any lines White finishes having always 1-2 pawns more, or 1-2 passed pawns, because after 30. Bxb3 and 31. e5+ White has more space, a very strong Rc1 and a passed pawn that make the difference.

After 30 e5+, black has limited options, all no good. The king can't move to the c file because his c pawn is pinned, losing the bishop. It's either 30...Kd7 or 30...Ke7.

click for larger view

Also, the a1-h8 diagonal opens up, allowing the bishop to put two pieces en prise with one move, 31 Bb3.

click for larger view

Black's error:
(+1.38):28...Nb7.
Alternative:

Analysis by Rybka 4 x64: depth 18:

1. (0.42): 28...h6 29.Rg3 Ke5 30.Bh1 Rxf2 31.Rxg6 Rh2 32.Rxh6 Rh3+

2. (0.44): 28...Rf8 29.Bd1 Rxf2

Black increases deficit:
(+2.41):29...Rxf2.
Alternatives:

depth 17:
1. (1.07): 29...Bxd1 30.Rxd1+ Kc5 31.f3 Kb6 32.Nc2 R4f6 33.h5 gxh5 34.Nd4 Rh6 35.Rdh1 e5 36.Nc2 Nc5 37.Rxh5 Rhf6 38.Rh6 Nd3 39.Nb4 Nxb4 40.axb4 Rg6 41.R1h5 Re7 42.Rxg6+

2. (1.60): 29...Nc5 30.f3 e5 31.Be2 Ne6 32.Nd5 c3 33.Nxc3 Rb7 34.h5 g5 35.h6 Rf6 36.Rh5 Nd4 37.Rxg5 Rxh6 38.Ba6 Rc7 39.Bf1 Rf6 40.Bg2 Nc2+

3. (1.89): 29...Na5 30.Bxb3[] Nxb3 31.Rxc4 Rxf2 32.Rc6+ Ke7 33.Nd3[] Re2+ 34.Kxe2[] Nd4+ 35.Ke3 Nxc6 36.Rf3 Rg7 37.e5 g5 38.hxg5 Rxg5 39.Kf4 Rf5+ 40.Ke4 Rxf3 41.Kxf3 Nd4+ 42.Kf4 Nb3 43.a4 h6 44.Nf2 Kd8

Black in more trouble:
(+4.22):37...Rc3+.
Alternative:.
depth 19.
1. (2.35): 37...Rxh1 38.Bxh1 c3 39.Kd3 c2 40.Kd2[] Kd6 41.Ng4 Kc5 42.Ne3 h6 43.Nxc2 g5 44.hxg5 hxg5 45.Bf3

after your 37.Rf3 there's the pretty much forced line 37...Rxf3 38.Kxf3 Nd4+ 39.Kf2 Rd1 40.Nc5+ Kc6 41.Ra5 and, in my opinion, its only black who can play for a win.

but i dont think there's much sense in bringing up now one line after the other. your statement "easy and sure win" is just not quite right, i think

Here's a breakout with Fritz 10:

<30. e5+ !!> This little deflection (pseudo) pawn sacrifice takes the Black King out of his comfort zone, so any legal move results in a forced loss.

Bad and actually losing for White is 30. Bf3? Rxb2 31. e5+ Kxe5 32. Bxb7 Rxb7 when Black's pawn majority wins easily.

White gets an endgame advantage after 30. Bxb3 cxb3 31. e5+ Kd7 32. Nd3 Rf1 33. Rc3 Na5 34. Rh2 Rd1 35. h5 g5 36. Rd2 Rxd2 37. Kxd2 Rf5 38. Nc5+ Ke7 39. Nxb3 . However, a win, even if there, is problematic and far from simple or assured.

<30... Kd7> This is the best move in a lost position.

If 30... Kxe5, White gets a winning Knight fork after 31. Bxb3 cxb3 32. Nd3+ .

If 30... Kc7 (or 30...Kc5 31. Bxb3 ) 31. Bxb3 exploits the pin on the c-file to win a piece.

If 30... Ke7, White uses the double attack and deflection (removing the guard or defender)tactic to win decisive material with 31. Bxb3 cxb3 32. Rc7+! Kf8 33. Rxb7! .

<31. Bf3!> Thanks to the King blocking the Rook's protection of the seventh rank, this move combines it's own obstruction of the f-file and a double attack on the Rook and the unprotected Knight to win decisive material.

<31...Rxb2>

If 31...Kf8, then 32. Kxf2 decides.

<32. Bxb7 Rf5 33. Bc6+ Kc7 34. Be4 Rxe5 35. Nd3 Rc2 36. Nxe5 Rxc1 37. Rh1 Rc3+ 38. Kd4 Rg3 39. Rf1 c3 40. Rf7+ Kc8 41. Kc5 c2 42. Kd6 Bd5 43. Bxd5 exd5 44. Nc6> 1-0 This final move surrounds and entombs the helpless Black monarch.

Even allowing that 37. Rf3 is not so good,there are other ways by which White can win, e.g.: 37. Rxb3 Nxe5 38. Rb7+ Kc6 39. Rxh7 and again White remains a pawn more, and 2 connected + passed pawns. But there is a practical point that you and other players are missing.

you're right, i dont want to analyze this position with you in detail. Not because i'm afraid, i'm still convinced that black has good drawing chances in your line, its more your stubborn style of argumentation.

first you say, white wins the b-pawn (clear and easy win), then, ok i dont take the pawn i'll promote my a-pawn at once (winning), then, ok i cant promote the a-pawn i play very strong Rf3 (winning), then, ok Rf3 is not that strong i... (winning), ad infinitum (i fear).

i think already our little discussion here shows that its not so sure and easy at all, and that was my whole point. and it seems to me that Volokitin, after all a almost 2700 GM, shared this view.

Analysis of a game means that people propose one or more lines, and then try to check and discuss together, to see what they find, and it is always possible to correct and improve the moves, nothing is final in chess.

I recognized myself -after your proposals - that the line played by Volokitin is winning, of course, but it is more risky - in a real game - than the one I suggested, although "my" line gives less advantage, surely less than the one I was thinking at first glance, therefore I changed my first opinion, I can admit that it is not so "easy" to win in that position.

As far as the topic of 2700 rating is concerned, sorry for you, but I can find you HUNDREDS mistakes played in real games not just by "almost 2700 rated chessplayers", but even by world chess champions!

You can just check under my nickname, and you will be surprised to discover some improvements (Ni-Hua (2710) - Hoyos (29.h4!); Tal (world champion) - Gasic (17.Nf4!!) Quinteros - Fischer (world champion) (23...Rxc4!!)) that nobody proposed before.

And I am analyzing WITHOUT any computer assistance, because every wood-pusher can ask Rybka, Fritz what to do, but without thinking over by himself.

Of course I can make mistakes as everybody else, but I'm writing here right to compare my proposals with the others.

Therefore, you are free to analyse whoever you want with, maybe I can survive if you don't want to analyze with me.

one last word from me too.

my impression was, that from the very beginning your evaluation of the position, for whatever reasons, was fixed (a doubtful one in my view). and the whole further discussion, analysis from your side had just one aim: to proove your point of view.

in that sense you were not open for the "truth" or even for changing your mind just a bit, but only you want to "win" the discussion. in these kind of "discussions" i'm not interested, they are just a frustrating waste of time.

now your last post sounds quite different, so i may have been wrong. if so, sorry for that.

The possibility of creating a double attack against the rook on f2 and the black knight with Bf3 suggests 30.e5+:

A) 30... Kxe5 31.Bxb3 wins decisive material (31... cxb3 32.Nd3+ and 33.Nxf2).

B) 30... Kc4 31.Nd3+ and 32.Nxf2 wins a rook.

C) 30... Kc7 31.Bxb3 wins a bishop.

...Kxe5 31. Bxb3 cxb3 32. Nd3+

...Kd(e)7 31.Bf3 double attack on Rf2 and Nb7