Incidentally, IM Jeremy Silman in his books recommends that one not bother studying the ending. He argues that it occurs so rarely that if you devote the time you would have spent on it to, say, studying rook endings, you'll get a better return on your time investment. If you do study it, there's a good chance that you'll never have a chance to use the fruits of your labor.

As I later discovered, winning the ending isn't as complicated as it appears. There is a recurring pattern of moves of knight and bishop that effects the crucial objective of forcing the king out of the wrong corner. Granted, in a real game there's the proverbial "many a slip," particularly since merely by defending *incorrectly*, a defender can throw one off one's pattern, but I am still glad to have learned the essentials. If nothing else, opponents who know I know them tend to resign rather than play out the position and make me prove it. :-)

<perfidious>: You're not wrong to make even a master-strength opponent prove he can win this ending. As GM Epishin's debacle shows, even very strong players sometimes have gaps in their knowledge. And what's the worst that can happen? You can lose after a long struggle rather than resign immediately. More work may be required, but is the result any different?

It would be kind of cool if FIDE enacted the "Botvinnik rule" and stripped any GM who gave up a draw in this ending of his title - at least for a month or so.

Of course, one could enact it with provisions to account for such mitigating circumstances, but this would tend to invoke the curse of the arbitrary and all the challenges that must inevitably arise therefrom.

I have never seen this before. I was just wondering if anyone has.

btw, it turns out I have mad skillz with B+N. I mated in 31 from the bad starting position it gives you. (In the worst-case scenario - well, apart from the positions where the defender has an immediate draw - B+N+K v. K can take 33 moves to win. http://en.wikipedia.org/wiki/Bishop... )

The endgame makes me think of all those games he played against Rybka, tormenting it with minor pieces. That skill actually came in handy here - he was in his element.

That answers my question. I like Naka's game better, but I guess I shouldn't be surprised this has been done before (a naked or nearly naked King being checked by an underpromoted Knight to preserve tempo).

1) Tigran Petrosian: 94
2) Boris Spassky: 86
3) Vasily Ivanchuk: 84
4) Paul Keres: 76
5) Vladimir Kramnik: 64
6) Mikhail Tal: 62
7) Borislav Ivkov: 51
8) Julio Bolbochán: 50
9) Svetosar Gligoric: 46
10) Aleksandar Matanovic: 46

SuperPatzer77:

click for larger view

Instead of Kramnik's 78th move - 78...Kh2, 78...Ba7, 79. Ng5+! Kh2, 80. Bg4! Bb6, 81. Ne4 Bd8, 82. Nd4 Bb6, 83. Nf3+ Kh1 - See diagram below:

click for larger view

Now Black has defense against 84. Nf2# or Ng3# but White finds a winning move - 84. Bh3! (Black has no defense against 85. Bg2#)

SuperPatzer77