Abstract

Probabilistic local realism for two correlated systems as formulated by Clauser and Horne in 1974 is shown to be necessarily based on a perfect specification of the state and on an individual definition of probability. All known realistic formulations of probability calculus are instead defined in terms of relative frequencies, and perfect specifications of states are impossible. We reformulate probabilistic local realism by using the relative frequency definition only and show that the Einstein, Podolsky, and Rosen paradox still obtains