It IS possible to checkmate with two knights

Yes you can force mate with bishop and knight.

With only two knights and king you cannot force checkmate, with a knight and a bishop checkmate can be forced. with two knights we can give mate only if position is so to give force mate or sometimes with one or two pawns on the board blocking the opponent king. For learning the hardest checkmate with a knight and a bishop see Capablanca in his book, Chess Fundamentals. However with two bishops it is very easy to give checkmate.

Thanks

True, but it's not forced. 2 opposite color bishops is forced, just like 1 queen or 1 rook.

It could be forced.

1 king against 1 king and 2 knights will never be forced.

Mate with 1 king against 1 king and 2 knights. It's a forced mate.

nice loophole

snoozyman, you had to do it didja this song is a huge meme

It's not possible unless the position is set up. If there are no other pieces, just two knights, I don't see how it could happen.

Any legal set up position can be reached in a game, so you must be talking about illegal set up positions.

It is obviously possible to checkmate from legal positions with no other pieces, just two knights.

The maximum mate depth from positions with king and two knights against king is one move.

The position I posted in #70 is also a checkmate with no other pieces, just two knights (mate depth zero with the same material, but it can definitely happen).

K+2N can mate a sole king but only if the side with only the king blunders.

Nice

That rather depends on whether or not the starting position is or is not a theoretical draw.

If it is a theoretical draw then no side can mate with any material unless his opponent blunders.

Can you prove that if it is a win for one side no positions such as the initial position shown in #70 or the position displayed in #74 can occur in any winning line without a blunder by the loser (or, alternatively, prove that the starting position is a daw)? There are no blunders subsequent to reaching those positions by the losing side - and obviously no blunders after the endgame is reached, because the side with only the king makes no moves after the endgame is reached.

(The difference between the two is that the endgame is reached in the first example with one move less mate depth than in the second. It may be easier to prove it in the second case ; I haven't looked very deeply into the problem. Even so, only if you count an inaccuracy, possibly gross, as a blunder - I reserve the term for a move that changes the outcome of the game.)

Read my post above.

Your statement is tenable only if you conform with the convention that forcing mate necessarily applies to a sequence of at least three ply.

You must accept that there are positions in KNN v K where the side with the knights can achieve mate and the lone king can do nothing about it, even if you are unwilling to say that the mates are forced in such cases; also that there are positions in other endgames where one side can achieve a mate in the endgame KNN v K with a sequence longer than three ply without his opponent cooperating in any way.

@pranavodave asks in this post about the endgame two knight[s] and King vs one rook and King.

From 3.2% of positions in that endgame (or of positions with ply count 0 under the 50 move rule if that is in force) it is possible for the side with the rook to mate his opponent without any cooperation from said opponent. But, if the opponent does not cooperate, it's necessary in almost all of those cases to first capture a knight, changing the endgame to KN v KR. Would you then say that in almost all cases it's impossible to force mate with a rook against two knights unless your opponent cooperates? (I would be more in sympathy with that statement.)