Limits to quantum mechanics as a source of magic tricks: Retrodiction and the Bell-Kochen-Specker theorem

A trick discovered by Vaidman, Aharanov, and Albert permitting retrodiction of the outcomes of more measurements than one would naively have thought possible is extended to a case in which the retrodicted observables are forbidden all to have values by a Bell-Kochen-Specker theorem. A rather peculiar analysis shows that an even better trick that retrodicts the outcomes of more informative measurements of these same observables is impossible