This paper first poses a skeptical challenge to clinical trials in medicine. The efficacy of treatments is measured against placebo; but placebo responses are not constant. They fluctuate with demographic variables, and they seem to be increasing over time. We therefore find ourselves measuring with the equivalent of what Wittgenstein termed an “elastic ruler”.

I then propose a “skeptical solution” to the problem. Elastic rulers are suitable tools for measuring dynamic, floating networks of values, like foreign currency exchanges. We can assuage the skeptical concerns by understanding clinical trials in this way; I suggest several practical guidelines for doing so.