fischer comments on Techniques for probability estimates