My bone to pick with this popular stats/probability text is that Mlodinow indulges in the same sloppy examples and logic that marred so many of my math and science classes. If you're going to talk probability, stick with dice. As a nerd from the early days of Dungeons and Dragons, I completely understand the difficulty of generating random numbers on your scientific calculator because the one guy who made polyhedral dice wasn't feeling well and there wasn't a dodecahedral die to be had in all the land. Believe me, nothing pleases me more than an extended discussion of the normal distribution. And I'm happy to hear about Bayesian analysis and its place in the science of probability. False positives in HIV testing? Bring it on. But please, don't confuse the issue with examples about humans and their behavior. To put it another way: Any example that you're going to have to keep qualifying by removing variables isn't a good example. All athletes don't have equal abilities. All of management isn't luck of the draw. The freewheeling omission of factors such as capacity, motivation, and personality doom these examples and muddy rather than clarify the concepts. Mlodinow further obscures his points with anecdotes about writing his child's paper and getting only a 93. What a shocker! In a move repeated throughout the book, Mlodinow confuses two variables. An instructor is not only making a judgement about the excellence of an essay in and of itself, if even at all. In my experience, it's more likely that the instructor is also, or entirely, evaluating whether the student followed instructions for presentation, content, and organization. That's not the same thing. By the way, if I ever catch Mr. Mlodnow's child turning in a paper written by his father for one of my classes, they will find that I regard this less as entertainment and more as a matter for the academic conduct office.