Cool move!

[Ovyron]

Instead 3...Qd7, maybe 3...Qf6 it´s better, because at least
blacks takes the white pawn in f and stay with pawns in g and h
passed.

Paulo Soares

Thanks, the line with Qf6 goes like this:

1... Nxg4 2. hxg4 Bxg4 3. Bd5 Qf6 4. Nh2 Qxf2 5. Rxf2 Bxf2 6. Kxf2 Bxd1 7. Ndf3 a5 8. Bg5 Bc2

Maybe it's better, but without queens on the board I think white has an easier game.

[PauloSoare]

Instead 3...Qd7, maybe 3...Qf6 it´s better, because at least
blacks takes the white pawn in f and stay with pawns in g and h
passed.

Paulo Soares

Thanks, the line with Qf6 goes like this:

1... Nxg4 2. hxg4 Bxg4 3. Bd5 Qf6 4. Nh2 Qxf2 5. Rxf2 Bxf2 6. Kxf2 Bxd1 7. Ndf3 a5 8. Bg5 Bc2

Maybe it's better, but without queens on the board I think white has an easier game.

You are analyzing without the help of engines, you must be a good player OTB.
Rybka analyses of the position were different from yours, but I believe
that it is very difficult for black to win the endgame.

[Event "?"]
[Site "?"]
[Date "2008.03.02"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[Annotator ",Usuario"]
[SetUp "1"]
[FEN "r1bq1r1k/bpp2ppp/p1np3n/4p3/4P1P1/1BPP1N1P/PP1NRP2/R1BQ2K1 b - - 0 1"]
[PlyCount "13"]
[EventDate "2008.03.02"]
[SourceDate "2008.03.02"]

1... Nxg4 2. hxg4 Bxg4 3. Bd5 Qf6 4. Nh2 Qh4 5. Qe1 (5. Nxg4 {?!}) 5... Bxe2 6.
Qxe2 Qg3+ 7. Kh1 Bxf2 *

[Ovyron]

You are analyzing without the help of engines, you must be a good player OTB.

No, I've been actually using engines for my analysis and am very weak on the board (About 1500), but I like analyzing positions and this one is interesting as I'm trying to find if black does indeed have a winning game. Sorry if I gave the wrong impression.

Rybka analyses of the position were different from yours, but I believe
that it is very difficult for black to win the endgame.

1... Nxg4 2. hxg4 Bxg4 3. Bd5 Qf6 4. Nh2 Qh4 5. Qe1 (5. Nxg4 {?!}) 5... Bxe2 6.
Qxe2 Qg3+ 7. Kh1 Bxf2 *

I overlooked 4... Qh4! and the positions are looking so good for black that maybe giving away the rook like that isn't a very good idea, and I think that 3. Qf1 is better.

However, if even this crazy idea with 3. Bd5 could hold the draw, I think that we may [partially] conclude that Nxg4 doesn't 'win outright' (Even when it's the best move.)

[F. Bluemers]

Yes, Nxg4 is the best move, but I don't see it "completely winning." My old variations seemed losing, but I think that one key element of the position is that Black's Bishop has at least the same value of White's Rook, so a good freeing move is to give the rook away.

This is my new main line:

1... Nxg4 2. hxg4 Bxg4 3. Bd5 Qd7 4. Nh2 Rab8 5. Kh1 Bxe2 6. Qxe2 Ne7 7. Bb3 a5 8. Ndf3 a4

It has some counter-intuitive moves but that's the best I could find for both sides.

instead of 4..Rab8 4..f5 seems better.
btw: Gandalf 6.0 feels very much at home in these kind of positions
12.. Nxg4 13. hxg4 Bxg4 14. Qf1 Qd7 15 Qg2 f5 16.Bd5 Rf6
17. Re1 Raf8!
Best
Fonzy

[Cubeman]

If Nxg4 does not lead to a win, then can't black play some other sensible move.Togall 3.1.2SE seem to like 1..Ng8 and thinks that Black is slightly better.

[F. Bluemers]

If Nxg4 does not lead to a win, then can't black play some other sensible move.Togall 3.1.2SE seem to like 1..Ng8 and thinks that Black is slightly better.

Most engines dislike Ng4,they want to sit on the material.
black doesn't seem to risk very much with Ng4.
having said that,Ng8 seems very playable.
the whole line doesn't look promising for white.

another line 12... Nxg4 13. hxg4 Bxg4 14. Bd5!? f5 15 ef5 Rf5
16. Re4 Bh5

Best
Fonzy

[Ovyron]

If Nxg4 does not lead to a win, then can't black play some other sensible move.Togall 3.1.2SE seem to like 1..Ng8 and thinks that Black is slightly better.

Nxg4 is not an immediate win but puts white in severe trouble in where exact move choice may be needed to survive, and I hadn't get too broad/deep to know that black doesn't win for sure.

I'll continue analyzing it when I have the time.

[Cubeman]

I don't see anything I would definitely call a win. However after what I suspect is probably the main line: 12...Nxg4 13.hxg4 Bxg4 14.Qf1 Qd7 15.Qg2 f5 I think White is under a lot of presure. Black can play Rf6, Raf8 and continue to build up his attack and it seems like it might be hard for White to untangle his pieces for the defence. If you set up the position after 15...f5 programs start to see good things for Black. Perhaps White has a better defense on move 14. Anyway I think it would take me a long time to form firm conclusions and I don't plan on looking at it long enough to do so. That is one reason I would like to know the source.

This is all correct - white has no good defence after Rf6!

I assume you mean that if White plays 1..Nxg4 2.hxg4 Bxg4 3.Qf1 Qd7 4.Qg2 f5 5.Bd5! Rf6 6.exf5 then if Black plays 6..Qxf5 7.Be4 seems to at least hold and might even be better for White.Because White has some resouces based on 7..Qd7 8.Nc4 Raf8 9.Bxc6 Qxc6! 10.Qxg4 Rxf3 11.Be3 R3f6 leads to an unclear position again.
There may be improvements for both sides but I think the line up to 4.Qg2 is best for White.

[smirobth]

I don't see anything I would definitely call a win. However after what I suspect is probably the main line: 12...Nxg4 13.hxg4 Bxg4 14.Qf1 Qd7 15.Qg2 f5 I think White is under a lot of presure. Black can play Rf6, Raf8 and continue to build up his attack and it seems like it might be hard for White to untangle his pieces for the defence. If you set up the position after 15...f5 programs start to see good things for Black. Perhaps White has a better defense on move 14. Anyway I think it would take me a long time to form firm conclusions and I don't plan on looking at it long enough to do so. That is one reason I would like to know the source.

This is all correct - white has no good defence after Rf6!

I assume you mean that if White plays 1..Nxg4 2.hxg4 Bxg4 3.Qf1 Qd7 4.Qg2 f5 5.Bd5! Rf6 6.exf5 then if Black plays 6..Qxf5 7.Be4 seems to at least hold and might even be better for White.Because White has some resouces based on 7..Qd7 8.Nc4 Raf8 9.Bxc6 Qxc6! 10.Qxg4 Rxf3 11.Be3 R3f6 leads to an unclear position again.
There may be improvements for both sides but I think the line up to 4.Qg2 is best for White.

After 5.Bd5 Black should probably play 5...f4. This cuts off both of White's bishops from participating effectively in the defense. Then Black can continue building his attack with Rf6, Raf8, Rh6 etc.

[Cubeman]

I don't see anything I would definitely call a win. However after what I suspect is probably the main line: 12...Nxg4 13.hxg4 Bxg4 14.Qf1 Qd7 15.Qg2 f5 I think White is under a lot of presure. Black can play Rf6, Raf8 and continue to build up his attack and it seems like it might be hard for White to untangle his pieces for the defence. If you set up the position after 15...f5 programs start to see good things for Black. Perhaps White has a better defense on move 14. Anyway I think it would take me a long time to form firm conclusions and I don't plan on looking at it long enough to do so. That is one reason I would like to know the source.

This is all correct - white has no good defence after Rf6!

I assume you mean that if White plays 1..Nxg4 2.hxg4 Bxg4 3.Qf1 Qd7 4.Qg2 f5 5.Bd5! Rf6 6.exf5 then if Black plays 6..Qxf5 7.Be4 seems to at least hold and might even be better for White.Because White has some resouces based on 7..Qd7 8.Nc4 Raf8 9.Bxc6 Qxc6! 10.Qxg4 Rxf3 11.Be3 R3f6 leads to an unclear position again.
There may be improvements for both sides but I think the line up to 4.Qg2 is best for White.

After 5.Bd5 Black should probably play 5...f4. This cuts off both of White's bishops from participating effectively in the defense. Then Black can continue building his attack with Rf6, Raf8, Rh6 etc.

Yes I have been analysing the consequences of 5..f4 for the past 60 min and can't find a defence for White.The basic idea is Bh3 followed by Rf6 and Raf8, and everytime I keep losing the white Queen.Even sacing the queen for the rook and bishop doesn't help.So 5.Bd5 is no use.