Two Knights vs Pawn Mate

Can some one show me an example of how the two knights and king mate vs one pawn and king on the board? It seems like an interesting idea, but I haven't actually seen those kind of positions before..

Don't even need 2 knights.

Thanks, but I already know about that one. There are a few positions in which the two knights set up a mating net and only the opponent's pawn is free to move..

I don't think they win within the 50 move rule though, or not often.

I understand they can win within 50 moves fairly often if you have the proper technique and the pawn isn't too far advanced... but realistically how often will you encounter this endgame?

http://en.wikipedia.org/wiki/Two_knights_endgame#Troitzky_line

Thanks for the link. I just want to know this endgame because it's really interesting. Infact, there was one game with two knights vs pawn in the Masters section of my recently conluded California State Championship, and the player with two knights actually won. It seems like an interesting idea.

Remarkably, as Botvinnik recounts in his book about the 1941 Soviet Championship, Lilienthal had already gotten that endgame three times by then...and failed to find the win each time!

I believe the opponent's king has to be trapped against the edge of the board already.  Maybe even near the corner is a requirement.  I think you could probably figure it out on your own... except that Andy says Lilienthal failed 3 times lol, so maybe it's not so easy.

Dvoretsky gives only a half page to it.  A study by Troitsky where the enemy king is already in the corner.  He mentions that the win may require "dozens of precise moves" and is very difficult.  Some key points are:

-The enemy pawn must not be too far advanced. 

-A knight must be able to blockade it.

-The remaining knight and king must get the enemy king in or near the corner.

-At that point the blockading knight joins the hunt and mate is delivered.

Moves and annotations from Dvoretsky's Endgame Manual:

waffllemaster wrote:

I believe the opponent's king has to be trapped against the edge of the board already.  Maybe even near the corner is a requirement.  I think you could probably figure it out on your own... except that Andy says Lilienthal failed 3 times lol, so maybe it's not so easy.

You're right, it doesn't look easy , but a 2100 rated player at the Calchess state Championships had two knight and actually won. Thanks for the information!

Neither.

You may have intended http://www.chessgames.com/perl/chessgame?gid=1295765

In the nine moves after the ending was reached (move 61) the players threw away three half points in total.

These are all helpful for White but none is necessary. E.g. White can mate from the following position (and in the position Dvoretsky himself gives, the pawn is hardly on its starting square).

highlevel147 is totally mad

i know him. He is a mental friend of mine

hey, who are you?

i was talking with highlevel!

ND3 is also mate. There are some positions where it takes over 140 moves to force mate. Here's an example: Set white king b2, kngihts d2 and e2, set up black king c5, 2 pawns d5 and e5, mate in 96.

This is a perfect example of why the 50 move rule shouldn't exist. Endgames like this, which can take up to 145 moves to force mate. This is an extremely complex, and subtle endgame, not to mention realistic (irrelevant), and can require well over 50 moves. Whether a person could calculate it or not isn't relevant.

I would need 150  to get it. Great puzzles very informative thread.

If the 50 move rule didn't exist, games would last forever.