34.Nh7? what the heck,of all the moves white makes he chooses to isolate his knight and potentially trapping it! and moving it away from the battlefield. now even tho hunting that knight may be trixy,it is still completely offside and quite bad. In the eyes of a grandmaster this would be a critical error that may cost white the game

Hmmmm? This knight is diving into dangerous territory with no back up. What can I say about this: well... with an f7 pawn, Ke5, & Nd6... the White knight is in danger of being dominated... another important endgame priniciple. If White's knight runs out of escape sqaures... he is said to be dominated. Not good!!! It is equivalent to being down that piece. I think it will be more clearly shown in a few moves.

Nf6. I dont think black wants to wander his king down to f8 to capture that knight it may be to risky. but white's knight is still as good as dead.

Yes, Black does want to wander down to capture the knight.

i honestly think white may have thrown the game already

Domination!!! The White knight is dominated, it has no where to go! Do not let your pieces get dominated! Black has a won game! For more on this, buy GM Muller's excellent DVD on the subject.

but the problem is that white can play a4 and may be able to make it so that black's pawn majority will be useless

Can't black just win the knight with 36...Kd6 and 37...Ke7? It doesn't have any escape squares. And if white trades down into the pawn ending he should be lostbecause he's a pawn behind.

I'm not sure if Black has much alternative - if he plays 36.... Kf5 37. Kf3 Ke5 Ke3 and we meet at the same position. I suppose now I think about it more, Black could have played 36... bxc5 37. bxc5 Kd6 and make the same movements as the game. The a-pawn would not have been given time to move from a square the knight will naturally attack later. EDIT: If Black tries to win the knight, White follows him and gets the Black knight for the trouble. Black cannot break through with the queen-side pawns without the king. Black must protect f7... White will swap between f8 and e8 as White brings down his pawn. After the later g6, fxg6, and Kxg6... the Black king on e8 or f8 does not have enough time to get around and break them open. White forces a draw.

l-d-j is right... black needs to go after the dominated knight: 36 ...Kd6 37.Kf4 Ke7 38.Ke5 bxc 39.bxc Ne8 40.Nh7 f6+! (the domination continues! The knight still has no where to go). The White knight is lost! Do not let your pieces get dominated!

and now after bxa4,theoreticaly white and black has same ammount of pawns on their queen side.

this is risky. very risky. however if black wants to do this he musn't hesitate. i'd suggest 38. ... Kf5 39. Kf3 (39. Kd3 Kf4) Ke5 as an attempt to go to d4 but if white's smart he'd play 40. Ke3,as an attempt to win this game without any harm done to black. If this doesn't work then perhaps 38. ... Ng4+ another very risky move with the point of letting black go to d4 (and black can play Nf6 immideately after checking if he wants, especially if white plays 39. Kd3 Nf6 Ke3 and we are back to the start once more)

this is still exciting and unclear what the game will turn out. i'd say this is the part where the winner gets decided. unless a new twist gets added.

firstly what are you talking about? 40. ... Kxf8 41. Kxf6 is anything but stalement since black can move to e8 and g8. However this may loose for black,lets see... 40. ... Kxf8 41. Kxf6 Ke8 42. g3 Kf8 43. g4 Ke8 44. g5 Kf8 45. g6 fxg6 46. Kxg6 Ke7 47. Kf5 Kd6 and now the famous zwichensug move: 48. Kf6 Kd7 (49. Ke5 Ke7 is a draw) 49.Kf7 and now black has two choices. 49. ... Kd8 looses after 50. Ke6 Kc7 51. Ke7 Kc8 52. Kd6 Kb7 53. Kd7 Kb6 54. Kc8 Ka6 55. Kc7 and now we see clearly see that my calculation with the zwichensug wins. now lets look on black's other option: 49. ... Kd6 50. Ke8 Ke5 51. Kd7 Kd4 52. Kxc6 Kxc4 53. Kb6 Kb4 oh that wins for black. no sorry my bad,black will win if he goes 40. ... Kxf8,these are still very interesting variations and if anyone wants to see if there are other calculations that win for white after 40. ... Kxf8 plz comment of course i see that white can obviously get a draw after 47. ... Kd6 with 48. Ke4 but i wanna see if white can win. Now let's move on. In your variation it's still very unclear who will win after 42. ... Ke6 both players have chances but like you said,i think black may be better placed

Oops, no idea why I thought it was a draw. Perhaps I was looking at the board upside down!   Nice analysis but Black can't force a win after Kxf8... if we follow your line, White just uses the opposition to force a draw after 49... Kd6 50. Kf6 Kd7 51. Kf7, etc.

so black decides to try to play for a win by surrendering h7 square with 40. ... Ne8 of course after 41. Nh7 f6 white's knight is again temporary shut which is similar to the variation that you previously gave only that the knight is on e8 instead of g8. still 41. ... Nd6+ wins a pawn so its most likely stronger

As you said before: the tricky knights are difficult to control. Now black has a definitive edge.

Black is definitely winning!!

my brain figured out that now either way Ke5,Ke4,Kf4,Kg4,Kg5,Nf6,Ng5 black will still have the supreme advantage of passed pawns and capturing a pawn. this is very critical for white and i think black is winning now.

This is maybe why I shouldn't have played a5 earlier. It wouldn't have given the pawn time to move from a3, and I could have continued the same game saving a move by taking on a3. If the pawn moved to a4 (and a5), I can easily chase it for the capture. I wonder if this move would have given me that extra move needed to win the game.

This is an interesting theory that you have but i believe that either way the outcome would have looked pretty much the same.

White cannot save the a4 pawn. With the knight fork, white lost the game.

A very correct move! Perfect play by Black. Again, the theme is domination of the White Nh7.

look,white is just doing desperate mooves. black is probbaly winning this if black plays correctly.

After 44 ...Nb2: With the hold up over on the kingside, it is a perfect opportunity to gain a distinct advantage over on the queenside. 44 ...Nb2 is clearly going after the a4 pawn. Black is 2 pawns up (soon to be 3 pawns), this is still a clear win for Black. I cannot imagine that this game ends in a draw.

Maybe,just maybe white can get a draw here (even tho i doubt it) if he is really lucky. The problem is that white's king is closest to the pawns but i still think black is winning

Has atleast a draw?? That is correct... this is a very solid win for Black! The outside passer on a5 is the key.

i understand how the computer would call this inaccuracy. Now white cant use the advantage of the king (at least for the moment) to block black's pawns,but from the looks of it white is just trying one last risky trick to see if it works. but this may as well be his downfall

I cant believe I am saying this but, the darn machine is correct: The opposition works well in K+P endgames... almost always 100% advantageous but, this is not a K+P endgame, their are knights involved plus, the doubled c pawns (the real reason why 47.Ke5 is bad) present unique problems... the c5&c6 pawns & Ke7 create a wall on d4, d5, & d6 so, White must infiltrate under these squares like... 47.Kd3! However, taking more importance here would be the re-deployment of the offside Ng5 to intercept the outside passer on a5 so, 47.Nf3 or Ne4 is probably best.

ah yes,a brilliant trick by black. but i am still not really sure about 47. ... c4. Black's king looks to much offside and need to be back in the game however the time spent to bring the king back in the game may white use to produce a drawish position. very risky position. Topalov and Kramnik would love to check this position out,i dont doubt that.

c4 allows the white king to approach the pawns. Kd4, Nb6, Kc3 improves white's position, and he has time to walk with the knight.

After 47 ...c4??: The trump card here is the outside passer on a5! i.e. 47 ...N anywhere followed by a5-a4-a3 is hard for White to meet.

Unfortunately, I think Nf3-e1-c2-a1 has a reasonable grip on the two files... and then Black would need his king to help get something. But that is how I saw it when I played it, anyway.

okay so black tried a confusing way to bring his king into the fight. this did however give white the chance to attack with his king as we see here perhaps it was because of 47. ... c4 that black allowed white to become this offensive. but i wonder if black can get away by playing instead 48. ... Nb6!? immideately. this is another interesting move. the point is to push the a-pawn. lets take a look on this. since white's knight obviously can't stop both the a-pawn and the c-pawn at once in this very position we start with 49. Kd4 a4 50. Ne1!? which allows black to play 50. ... a3?!(dubious) but i wanna look at 50. ... Kd6 51. Nc2 c5+ 52. Kc3 Kc6 hmmm lets see.... white has a few options here. lets try 53. Na3 yeah the position looks drawish. I dont know maybe i missed something that could have won for black.

White can easily block the two files with his knight by getting to c2 and attacking a1. He can move between c2 and a1 easily enough to stop them. Black would need his king to come and help faster than White can aid. Unfortunately, the White king will win the race to rid of a pawn.

indeed but this position looks drawish anyways. white's king is nicely placed while black's king isn't which makes it more difficult for black to take advantage of his advantages. if only black could have traded knights then he'd have an easy win. i would blame this on 47. ... c4. giving up the d4 square and making it difficult for black

Noooooooooooooooooooooo!!! The move that throws it all away!! 50 ...Kc7 holds it all together!

kind of a pointless check but then again,white doesn't have much of a choise. However maybe white can play Nc2 here immideately,blocking that c-pawn and covering a3 square,maybe he can bounce to a3 immideately after Nc2 to controll more squares that's closer to the other pieces. this position looks drawish either way. the biggest issue is the simple fact that white's king is good placed - black's king is bad placed. but maybe black still have winning chances if he tries to offer a pawn as a gambit in order to get a better position for his other pieces,somehow. i've tried that in my endgames and a few times it has worked.

After 51.Nd4.... what is wrong with the GK computer?? 51.Nd4 throws the draw away!! The only move to draw is... 51.Kxb6! looks scary but; 51 ...c2 52.Ne5 (the saving check) K anywhere 53.Nd3 & the game is drawn! The rule goes... If the knight can get in front of a passed pawn (in a K+P vs. N endgame) than the pawn can never promote. Yes, there is still a passed a pawn here but, the White King can handle him & the doubled c6 pawn here makes no difference.