2 Knights Endgame Puzzle


I hate to say it but these don't really work as puzzles; not only don't they follow the top engine line but there's hundreds of possible ways to win

2. Kb3 does not work at that time, as black can play kc1. White needs the same position in both parts of the puzzle, with black to move. 2 consecutive triangulations.

Oh I see what you mean. I mistakenly had black play ka1 first. I meant kb1, you're right. I was wondering why it was 20 moves instead of 19, let me fix it.

It's not a puzzle though, there isn't a single pathway to victory; it's just a 'try to win with 2 knights' position.

It's not a puzzle though, there isn't a single pathway to victory; it's just a 'try to win with 2 knights' position.
Wrong. The original position, white needs to achieve with black to move, so it requires a series of triangulation. Next white needs to achieve the same opposition pattern, but with the king on the 1st rank rather than the a file, which requires a repositioning. Then, white needs to once again make the position black to move to force the king into the corner, which requires the same triangulation as when the king was on the edge. After the king is in the corner, the knight goes to a2 to trap it on 2 squares. Fifth and finally, the other knight and can join..forced mate in 19!

Wrong. The original position, white needs to achieve with black to move, so it requires a series of triangulation. Next white needs to achieve the same opposition pattern, but with the king on the 1st rank rather than the a file, which requires a repositioning. Then, white needs to once again make the position black to move to force the king into the corner, which requires the same triangulation as when the king was on the edge. After the king is in the corner, the knight goes to a2 to trap it on 2 squares. Fifth and finally, the other knight and can join..forced mate in 19!
True, there are a couple of required moves in the beginning, but as the game goes on you get more and more variability- it's a little improbably most of the different endings will be achieved, but according to the tablebase there's 81,225,600 possible ways to win for white in this position
True, there are a couple of required moves in the beginning, but as the game goes on you get more and more variability- it's a little improbably most of the different endings will be achieved, but according to the tablebase there's 81,225,600 possible ways to win for white in this position
No, not from this position but from all positions of the 2N vs P endgame. But solving this puzzle is almost impossible for a human being. I think it comes from chess.com as a training exercise!

Actually I didn't have much trouble understanding how to win in that position. It's basically a 4 or 5 step process that just takes alot of moves. We don't say a queen vs king endgame is hard even though it can take 10-20 moves to force the king to the corner. If the chessboard were 1000x1000 squares a KQK endgame would still be easy, it would just take longer.
Actually I didn't have much trouble understanding how to win in that position.
Understanding it is not that hard, doing it is another matter. Today almost nobody outside the composition domain ever solves a puzzle. They all cheat by letting the computer play the opponent's moves. Which is truly cheating since you need to find all the moves in your mind before you can play one on the board. Just as in a real game you only get one opportunity to choose the right move. Which you can only achieve by visualizing the whole sequence for both sides in your head.
Then again, I don't think this sequence was designed to solve - just to watch and learn - and hope you make less errors than your opponents when meeting similar challenges in your chess games!