How would black win this game?

Here is a game I played as black and felt at some point I had a positional advantage and couldn't push it home to a win. So either my positional evaluation is way off or I need to learn how to convert better. Any suggestions welcome.

You're better than me, as your Online rating would indicate, so consider everything I say to be fairly salty .

I think the key point is at 15. ... Bxe5.  My initial thought on seeing that move was that it willingly gave up the Bishop pair.  Continued play based off of a timely c4 (possibly even as a pawn sac) could open the center more favorably for you.  But, you did initially seem to win a pawn, so I initially considered that move fairly neutral - giving up one type of advantage for another.

However, upon seeing 17. Nf1!, my evaluation of that exchange changed drastically.  Now the center is open, you no longer have the Bishop pair, and White's Bishop is freed.  White now has options to play against the weak c-pawns.  A simple Ba3 after the 19th move would win one of them as you don't have enough dark square defenders.  Your play against the weak light-squares was appropriate, but as we see, White can get enough defenders present with the Queen and Rooks.

While I'm not sure what a better 15th move you could have made, but Bxe5 was clearly the turning point.  Perhaps a plan could be based around Qb3 and centralizing the Rooks followed by opening the center.  I'd say you're distinctly worse by 19. Qxe3, though perhaps salvagable had you gotten a draw during the repetitions.  I didn't count the repetitions, but I'm guessing that your opponent didn't actually give you an opportunity to claim such a draw, anyway.

I appreciate your time and thoughts. This gives me something to work on.

I'm still not sure on 15. ... Bxe5 because if I allow white to play Ndc4 his knights start to look menacing. I won't be able to maintain the knight on e3. But the way it turned out wasn't great either. Unless there really is something to that light square attack.

"A simple Ba3 after the 19th move would win one of them [c-pawns] as you don't have enough dark square defenders."

This is a good point, but it's not simple. 20. Ba3?? Qg2#.

21. Ba3 Re6 22. Bxc5 Rf6 and the queen is overloaded defending g2 and c5.

I definitely agree that 15. ... Bxe5 and the long term consequences made my c-pawns a target, though.

I could  have claimed a draw after 33. Qe2 instead of playing 33. ... h4.

Based on the variation I gave with 35. ... Rxg3+, There may still be some teeth left to black's attack. This was the feeling I had during the game that kept me from taking the repetition draw.

Unless I am missing something, it seems to me that the alternate line you gave starting with 35...Rxg3+ is simply winning.  And after 37...Bf5 it looks won.  How can he stop Rh3 without giving up significant material?  and I can't see any improvement for moves 36 and 37 that you listed for him.

I don't think Bxe5 was a mistake, actually. If you had followed it up with 16. ... Bd5, intending moves like Re8 and Ng4, black looks pretty good. That pawn can't be defended well, so it looks like a goner. Just don't take it immediately because of the pin...

I am not the best but maybe  15.Qc1 f6    16.Nf3 Bd5    17.Nh4 Bxf4  18.xf4  Qxf4 would give you a good advantage?

Good lord, loomis, I've spent the better part of the past evening and all this morning looking at your game (to come up with three short lines, two of which are inaccurate! LOL. I'm not exactly Dvoretsky, am I?)  I finally went to the computer (firebird running in the fritz gui) to see how I'd done... well I think I did ok! () except for completely missing the win for black on 34() and various other inaccuracies.

My analysis with Firebird's corrections

I looked at the Rxg3+ line and I think it's good for Black but White can hold on to drawing chances because of the opposite color bishops. (Firebird: Rxg3+ is winning but it's tactical... human analysis wandered down a drawish sideline...) Here's Rxg3+ analysis backed up to move 34 showing where you (and I, and everyone else) missed a fairly simple winning tactical shot (in the notes in the diagram below I call it a "double pin" which it is not.)

I also think 15...Bxe5 was a critical moment, and Bxe5 was no good. [Firebird agrees. Bxe5 was not a good move strategic/positionally ... you'd played very accurately up to this point, and had a definite advantage but your plus took a big hit here.] Although It's very tempting to exchange his strongly centrally posted N -- the continuation after Nf1 shows why it's not a good line [Firebird -- Bxe5 wasn't good, but Qxe5 plain bad... you went from slightly winning to slightly losing with Qxe5] and doesn't win a pawn -- rather than make that exchange it I thought you could kick the N with f6. [Firebird: meh... f6 isn't so great].  Beyond that... I thought it made sense to keep your magnificent center locking down his b2 bishop (which is why you might have considered keeping your N out off the e3 outpost until you could get some real leverage from it -- you didn't want your d4 pawn coming off the board! [Firebird: don't listen to this fish, 14...Ne3 was clearly your best move -- it was the bad Qxe5 plan that caused the problem.].

I thought after an f6 kick you had kingside attacking chances: His bishop on b2 was terrible, the white squares on the kingside were weak, the pawns extended -- I thought something dramatic was in order: maybe g5!?   In fact, maybe you didn't need f6 at all, maybe just an immediate g5!? [Firebird 15...g5! Hey, the blind squirrel found a nut! 15...g5! was by far the best move on the board.] With hair-raising complications to ensue -- fun for the whole family. Open files against his king for your rooks -- your pieces were well-posted and coordinated for a Kingside attack IMO and your queenside pawns weren't looking good for the endgame.

---

You wouldn't believe how much time went into the above not-very-impressive bit of analysis! But I did come up with 15...g5! *pats self on back*  It keeps the initiative and it really does lead to difficult interesting positions. 

My initial thought after Qc1 was g5.  With black's b2-bishop useless and your bishops dominating, white couldn't do much about a gxf4->kh8->rg8 plan.  My (somewhat outdated) version of Shredder agrees, and suggests that white either sac the exchange on e3 or one of his kingside pawns.  The defense OTB looks difficult, though.

Edit:  Oh hey, JG27Pyth.  Way to post my idea while I'm double-checking it with Shredder :-D

OMG do you realize what kind of nervous breakdown I would have had if you'd posted a minute or two ahead of me after the time I just spent on this position. LOL. You didn't want that on your conscience!

Thanks, JG, that is one heck of an analysis.

34. .. Qxe4. It's almost obvious once you see it on the board. Niether defender can recapture, one is pinned and the other defending mate.

15. .. g5. I think this is a key move where I can really learn something. Going all the way back to omnipaul's post, Bxe5 didn't seem right to anybody, but finding an alternative had been elusive as well.

There is a natural hesitation to play a move like 15. ... g5 because it has a sense of danger. Black feels like the positional plus is so good he should be able to win comfortably. But that's just not the case.

It seems the correct evaluation of the position is that with the bishop on b2 locked out and the queen on c1, black's pieces are much better coordinated for the dangerous position.

There is a quiet plan with 15. ... Bxe5 16. fxe5 Bd5 and ... Re8, but it's not as strong as the sharp plan with 15. ... g5.

Hi I'm new to this site and so far I love it.

On the 35 why did you trade Qs with white? you could of check white with RxG3+ then when white moves out move your Q and still have a mate in that corner if you play it right.

I not as rated as you are so if there is a reason you did not do this please say thanks

by the way nice game

I would like to know more about JG27Pyth analysis why in white doesn't play 17.)Rxe6?

Edit: Where the notation is blue in the analysis board.

? I don't think I posted any analysis around that move -- and 17.Rxe6 is not a legal move, did you mean sac'ing the exchange 17.Rxe3. If that's the move you mean I don't see how it's anything but losing after 17....Qxe3+

I don't think myself worthy of telling you how you should play the position, but I agree with Tenna. 16. ... Bd5 with the idea of playing Rae8 and Rxe5, and if Nf1, ...Nxf1 when ...Qd7 targets those weak light squares. The pawn cannot be protected by its brothers, and it creates no real threats so long as Black is careful not to play ...Rfe8 instead of ...Rae8, and White's kingside is hopelessly weak, and dominated by the Bd5. The position after move 19 in the actual game looks pretty horrible for Black: I'd rather have the pawn still on d4, the knights traded off on f1 instead, the e file blocked by White's pawn, and the positions of the Black queen and bishop reversed. I doubt that position could be said to be winning, but White would have little reason to claim equality.