bentarm comments on What a reduction of “probability” probably looks like