How could I have played better in this N vs. N endgame?

NotYourAveragePlayer

I failed to convert the following knight vs knight endgame in which at I had an extra pawn. I am not at all sure where I went wrong but I suspect that the first mistake was 58.Ne3+ where I had missed my opponents (potentially brilliant) reply 58...Ke4, giving up another pawn.

I think I have a decent understanding about N + P vs. N endgames but I am not quite sure about N + 2P vs. N, specifically knight's and rook's pawn. Is it generally a draw or a win and is there anything in particular to keep in mind as the attacker?

I appreciate any suggestions and advice.

Optimissed

Hi again, it's the great endings non-expert again. That was a really complex game full of shuffling about and all of a sudden, black blundered.

I don't like endings but I would have played 65. Kf5 heading for g4. Then the black knight has to go to g7 and you play Nf5. If it goes to g3, you push the pawn. Looks like a win to me.

RALRAL3333

@Optimissed you may very well be right. The computer suggests a very complicated but possible way for white to win with 65 Kf5. Also, you said, "all of a sudden, black blundered". Where was that?

Optimissed

Just where black dropped the pawn. I thought maybe black pushed pawns too fast and lost control very quickly.

Optimissed

Instinctively I thought Kf5 would have won so I would have played it without any thought after I saw that black has only one knight move and white can immediately drive it further away and the pawn should queen. But I didn't calculate; I just thought it wins.

NotYourAveragePlayer

Thank you @Optimissed for your input!

After having cleared my mind and looked at the ending again I think I missed a win when I played 67.Kf4??, which in hindsight is an obvious blunder. Had I played 67.g6 then I think White should win. For example, 67.g6 Kxh2 68.Nd5 Kh3 69.Kf6 Ne8+ 70.Kf7 Nd6 71.Kf8 Nf5 72.Ne7 Nh6 73.g7 Kg4 74.Ng8! Nf5 75.Nf6!+ 1-0

My conclusion is that knight endgames are like pawn endgames with regards to the objective evaluation however they are harder to play than pawn endgames.

Optimissed

67 Ne4 looks ok too. Holding the pawn and the N can't get near it if you don't play Kf4.

TooManyBlunders

That is a very interesting ending. @NotYourAveragePlayer you did well attacking and winning the d-pawn.

I would have considered 58. Nc3 for White with the intention of advancing the d-pawn, as the knight and pawn control the e4 and f4 squares respectively (so the Black king cannot invade to attack the g3 and h2 pawns).

I checked the position with a tablebase after 60. Nxh6 and the position was initially a draw. With endgames with so few pawns on the board, the position of the other pieces becomes important. The main thing is that the Black king and knight are well placed to stop the pawns, while the White king is too far away.

Black blundered with 62. ...Nf4 (62. ...Nf6 is the only move that draws, as g5 can be met by Ne4+ and h3 can be met by Kg3).

Both 67. g6 and 67. Ne4 are winning for White. However, the win is not easy (accurate play is required) and there are many chances to go wrong.

 

 

st0ckfish

you should have asked me during the game 

Kingdom_of_Macedon

You could start by reading chess books to improve your strategies and tactics for starters

NotYourAveragePlayer
TooManyBlunders wrote:

I would have considered 58. Nc3 for White with the intention of advancing the d-pawn, as the knight and pawn control the e4 and f4 squares respectively (so the Black king cannot invade to attack the g3 and h2 pawns).

I checked the position with a tablebase after 60. Nxh6 and the position was initially a draw. With endgames with so few pawns on the board, the position of the other pieces becomes important. The main thing is that the Black king and knight are well placed to stop the pawns, while the White king is too far away.

Black blundered with 62. ...Nf4 (62. ...Nf6 is the only move that draws, as g5 can be met by Ne4+ and h3 can be met by Kg3).

Both 67. g6 and 67. Ne4 are winning for White. However, the win is not easy (accurate play is required) and there are many chances to go wrong.

 

 

 

58.Nf4 looks very good and I did not consider it. It is probably better to keep as many pawns on the board for winning chances and advance the d-pawn. I was too greedy.

I think your comment is spot on: "With endgames with so few pawns on the board, the position of the other pieces becomes important. The main thing is that the Black king and knight are well placed to stop the pawns, while the White king is too far away."  This was exactly @RALRAL3333 comment after game as well.

Once again, the conclusion I can draw is that there aren't many general guidelines in knight endgames, apart from knowing some basic patterns. Two extra pawns should win in the majority of cases. However, it is extremely important to be able to calculate accurately (e.g. not make the relatively trivial mistake I did with 67.Kf4??). The problem is that knight endgames usually appear at a later stage of the game when mental fatigue starts to become a factor. Even though this game was a daily game, I was tired of the position and could not force myself to calculate properly. I felt that I had done so much work in winning the d-pawn and wanted it to be enough effort to win the game.

Note to myself: whenever a knight endgame appears on the board I should stand-up, walk around, get another cup of coffee and then focus intensely (assuming time permits).

TooManyBlunders

@NotYourAveragePlayer it is quite interesting to hear your thoughts. I would agree that the work you put in to win the d-pawn would have been enough to win you that game.

I definitely agree that fatigue can play a part in endgames, especially when your opponent is putting up a stubborn defense and the ending goes on for weeks. However, it is definitely important not to get careless (although this is easier said than done). I can understand why you went for the pawn, as sometimes one pawn ahead is not enough to win, but two pawns ahead is often enough to win.

I would say that for knight endings, being a pawn up is more significant than in bishop, rook or queen endings. With the other endings, I would say that there are more possibilities of a drawing resource  such as a blockade or perpetual check.

One thing I have come across from my experience (both online and OTB) is that with only a few pieces left, it is quite common for the weaker side to have a strong defensive resource (generally with the idea of sacrificing the piece for the remaining pawns).

NotYourAveragePlayer

It makes a lot of sense what you say @TooManyBlunders With more pieces the attacker has more winning options e.g. trade down, not trade down, keep some pieces, etc.

An important thing I need to work on in my games is recognizing when the character of the game changes. Being cognizant of this and able to see in advance where the game is heading is generally useful I think. As in the case of this knight endgame, if I had prepared myself mentally before going into I might have been able to take a deep breath and finish the job.

Winning the d-pawn although it required some thinking did not require a lot of calculation. It was more of strategic endgame skills. When I entered into the knight endgame I was still in the same mental state and wanted to finish the game without having to calculate. Unfortunately, chess doesn't work quite like that!

Optimissed

I like the endgames you keep posting. Got any more?? I like some kind of chess problem when I'm bored but not quite tired enough to sleep.

zurkhaneh10

Excellent attack