We analyze a possible loophole to the conclusion of the Bell-KS theorem that quantum mechanics is not compatible with any realistic and noncontextual or local theory. We emphasize that the models discarded by Bell-KS-like arguments possess a property not shared by quantum mechanics, i.e. the capability to make non-trivial statements about the joint statistics of quantum incompatible observables. By ruling out this possibility, apparently nothing seems to prevent from a realistic, noncontextual or local view of quantum mechanics.