[The Zerojack]

I would think a correctly designed pseudoint would react to the 72o situation as an aberration. If it was sufficiently developed, and capable of self-programming, it is a virtual certainity it would come to the same conclusion after enough processing time's devoted to it, that it is a move designed to "tilt". Would it react wrongly? The first time, of course. If you've had your bullets cracked by that hand, you've tilted. But over time, you've learnt to factor that into your whole poker-playing arsenal. "This will happen, but it is unlikely to happen, so it's not a major concern, merely an overbetting caution." <--- illustrating what, IMHO, a recursive would apply to this situation. That said, they'd assign it a probability value of nonzero, and if they were permitted to remember its opponents moves (Permitted to remember? Jove, that's telling of a biological bias, isn't it? Hums Thus Began Zarathusra and looks for a rock) It'd be likely to assign a probability flag and some processing time for that situation.

Asleep yet? Your pardon. It's just that I find it quite fascinating we're at a point where we're discussing the qualities of emergement intelligences... In... NL... Poker... It kicks ass, man.