"Easy, the particles are correlated at birth. *they* know what their orientation is, it is fixed at birth. The math says *we* don't know." No. Bell's inequality tells us that there are no "hidden variables". It's not that we don't know the value of the measureable prior to wavefunction collapse...the specific measureable doesn't exist prior to wavefunction collapse. When Bell formulated the testable inequality circa 1980, and then it was experimentally violated (the inequality, that is, not the theory behind it), it became accepted within physics that (as expected) wavefunction collapse determines (right then and there) the value of observables, forcing the universe to choose, according to the probabilities. (Of course, this was basically understood from QM's beginnings, but prior to John Bell's work we couldn't actually test that this was reality.) (I would argue that this was one of the necessary conceptual pieces needed for people to trust the notion of quantum cryptography.) Taking a measurement (whether acidentally or on purpose) forces the quantum system to "choose" instantaneously. Einstein understood this aspect of Quantum Mechanics so well that he and 'P' and 'R' concocted the EPR gedanken to show that this implies what is effectively "action at a distance"...the different pieces of a quantum system, even far removed, spontaneously 'know' about the other parts. (In fact, I would bet that Einstein's original complaints about QM's action-at-a-distance may have been what prompted the reactionary fad of 'well, QM is merely a useful calculational tool...) But in the end, as strange and unreasonable as this action-at-a-distance may be, it's now regularly seen in the laboratory. (Even wierder are the 'quantum eraser' and other bizarre behaviors). -TD