As a statistician, when he is facing an unknown phenomenon or even the whole world, his highest rationality should be that he knows nothing, and everything to him is in randomness thus nothing is certainty. This is the only assumption that he can establish when he begins his cognition to the whole world.