Medicine is so often about playing the odds. Give a 60-year-old a new hip and she might enjoy 20 more years of tennis, but give one to a frail 85-year-old and he’ll die on the operating table. A cabinet full of inhalers helps open the lungs of someone with emphysema, at the cost of swollen feet and bruised, shaky hands. If a two-year-old isn’t talking, her parents could start expensive behavioral intervention, or wait six months and see if she comes around.

This is the calculus that doctors do, ideally aided by their own experience, their patient’s preferences, and the collective wisdom of the scientific literature.

But what if there are no studies to turn to, no doctor’s intuition, no odds on