Remarks on the use of objective probabilities in Bell-CHSH inequalities