When Carlsen sensationally beat Kramnik in the first round of the London Classic, many computer-armed viewers noticed a weird move  that would apparently have justified Kramnik’s play.

The questions on chess forums afterwards were, why? and could any human player have seen it? Andrey Terekhov, analysing at Crestbook.com, added this note on the idea: (originally posted here)

32.Ne2 Bf5?? A fatal loss of time
Rybka claims that after the inhuman 32…Kf8!! there’s a mathematically exact draw on the board! Those wanting to verify that bold claim can check it out on their own :) Explaining precisely why it works isn’t possible in a few words – the position deserves separate investigation. For now I’ll just note that the hidden sense of the move 32…Kf8 is that in some variations black has the ability to play Bb3, not fearing it being taken with check. What can we say about this? For some reason or other it seems to me that no contemporary chess player would be up to playing the move even with sufficient time for reflection, and therefore I’d file the remark that the move 32…Kf8 would allow black to save the game under the category of “pure knowledge”, of no interest to anyone except, perhaps, true enthusiasts of advanced chess.

That’s probably got more than a grain of truth to it, but I don’t think it would actually be that hard to find if you kept calculating lines and found that the king on g8 stops them from working. It does look crazy, though, and it’s not something that’s going to jump out at you during time trouble…