(Not, of course, that there aren't situations in which the best course is to take a gamble. But in situations where there's no evidence, not even a coherent line of reasoning, that the cure is better than the disease? I don't think so.)