One interesting endgame

Hi. I played a game against a friend of mine, and we reached this position, where we agreed a draw (if you press backwards, you can have the game if you'd like to).

Now after a series of analysis with some friends (even 2000 rated ones), we haven't been able to come to a final veredict of whether this is a draw or win for black.

Chessmaster vs Chessmaster 40 min game resulted in a win for black, but I'd like us to discuss that. What do you think? Post your thoughts on this endgame.

Good question.

I like b5 + and then Knight f4.  Black's overwhelming pawns should win.

Pretty complex.. I think black wants to play 1.b5+ to prevent any kingside expansion by white. I'm not sure about the rest. I've gone over some lines and it looks pretty drawish but both sides need to be very careful.

Some thoughts (from a beginner, so probably error-laden) on this interesting endgame:

What happens after 38.... Nf4 ?  39. Nxf4 gxf4 and white's forced to rush back and stop the passed f- pawn while black's king devours the queenside pawns.  Nf4 also threatens Nxh3, but that seems to be quickly stopped by Nf2.  Black's knight is more actively positioned, but I can't see a direct line for black leading to a win.  

38... f6 also looks promising, as 39. exf6 Nxf6 and 39. ... fxe5 40 Nxe5 eliminates white's remaining center pawn, frees the now passed e-pawn, and deprives the white knight of any outpost besides f5, which isn't incredibly helpful in stopping the pawn.  The knight can be driven away from its blockade by Kd6 and Nd7.  However, 39. Kd4 fxe5 40. Kxe5 may give white a way into black's position.  a5 followed by Nc5 (40...Nc7 41. a5 (... b5 42. Nc5) bxa5 42. bxa5 and then Nc5) puts even more pressure on the black e-pawn.  

Looking quickly, I see 38... a5 countered by 39. b5+ and Black's king is locked out, probably good enough for a draw for Black.  38...b5+ 39. axb5 axb5 + Kd4 moves white's king closer to the potentially vulnerable f6 square, while hampering black's game.  White's b pawn has a death grip on a5 and c5, and it will be difficult getting the black king into the position.

38... Ne3+ 39. Kd4 (any king move) Nxg4! 40 hxg4 h3 41. Nf2  ties down the knight, and forces the king away from the defense of the other pawns.  This could lead to very interesting things; this line may end in a loss for black, but I think this offers the best chance for an interesting game.

Out of curiosity, what did the computers play?

For some reason I couldn't change the result, but the final position (at move 73) is a winning position for black.

Pretty complex, could see it being a draw, but with the extra pawn I don't see why black didn't play on to see how it ends...

This diagramn shows that Nf4 is a losing strategy for black.

Wow, I hadn't considered the triangulation, or that b5+ led to black having no way to capitalize on white's need to stop the f-pawn. Nxg4 hxg4 h3 Nf2 h2 and then b5+ probably leads to more of the same thing. 

 I tend to agree... my choice of moves, exactly. Black has a chance of winning. bondiggity, Black has to play b5+ first, not Nf4 (playing Nf4 first is a loser). By playing b5+ first, and if white replies axb5, black gets the opposition (with Nf4 first, white gets the opposition, and black cannot cross white's pawns), which makes a lot of difference. If white plays 2.Kb3! in response to 1... b5+ then 2... Nf4 3.Nf2 (3.Nxf4 gxf4 and white king cannot cover both flanks ) Kd5 

 I agree, but during the game I saw some ghostly images of the white knight causing trouble. Fear, basically. But it was a friendly match, so we had no reason not to play on, really.

Had I met this position in a tournament game though, I'd definetely take the draw.

I am not sure if the triangulation is perfect for black.

Why not 41.b5+?

I believe it is a draw.