White resigned
Ingenious Draw


I felt 6...Ke1 would lead to a perpetual 7. Nf3+ Kf1 etc. But 7. Kd1 may be worth a peek.


I think it really is just a win for black. I have the book as well, and it surprised me, but it's more of an optical illusion than anything else: Instinctively, you want to keep your king close to the kingside, yet this actually allows black to gain time with his pawn, and in any case it can be no more than useless since black gets control of the queening square. But really, all you have to do is get something to control g1 and nothing more, which the knight does well. Thus you switch spots with the white knight! 2 Ke2 completes the illusion by even denying Ne2 for a move! But of course, white gets there in time. I guess 2 Ke3 draws as well, though not nearly as simply.
As I am new to starting forums and my PGN efforts seem to disappear, I will put them on separate posts below. The position below is from Jeremy Silman's Complete Endgame Course, and was originally from a game Akiko Uto-Tran, Paris 2005. White resigned, failing to find a solution. The first PGN diagram shows Silman's drawing solution. At first, I thought that black had other tries, an early push of his "a" pawn forcing white's knight to capture and thus preventing the knight from policing the g pawn, and even walking his king to f2 and thus preventing white's crucial Ne2. I have found no winning line for black, but I find the position worthy of study and understanding.