I know that you know, but do you know that I know that you know?

I thought I had come across an interesting trap in the Sicilian. Black lets White think he has made a mistake with 10... Ng4, but after the plausible 11.Nd5 Black gives up the queen on b6, but wins  3 pieces in return. Looking at the stats, White 'fell' for this at least 85 times, and while it didn't always  win for Black, a score of 63% for Black showed that is was quite effective.
However it turns out that White can lure Black into this line, but instead of 'winning' the queen, plays 12.Bxg4! Black has no choice but to capture on e3, but then the queens come off and White is perfectly fine. It even looks as though some White players specialise in this variation, as the same names keep turning up o the white side of the board. I suspect the choice is somewhat psychological, in that Black thinks he is doing something clever, only to have it refuted by the opponent.
Nonetheless, White isn't doing so well to render the line unsound. Black can still play it in the hope that White doesn't play the correct line, as in the following game.


Poulsen,Christian - Weil,Wolfgang [B73]
München Schach-Olympia (No FIDE event) Munich (4), 19.08.1936