That’s fine, I’ll easily believe “medical research can make no accurate predictions in this domain” - in fact, that’s my prior. In any case, people shouldn’t go around speaking as if placebo is this established strong phenomenon.
Exactly. Thanks, I had forgotten where I had read that, this is a nice source.
Here's the TL'DR: The placebo effect is maybe not quite as powerful as it is sometimes made out to be. And the reason is spontaneous remission or reversion to the mean.
1. Initially, researchers tested a treatment, and when the patients got healthier, it was assumed that it was the treatment.
2. Then researchers got smarter, and tested a treatment against a placebo (control), ideally double blind and randomised etc., and when the control group got better, it was assumed to be the placebo effect, and when the treatment group got even better, the difference was the treatment effect.
3. Then researchers got even smarter, and tested a treatment against a placebo (control) and against no treatment at all, and when the no-treatment group got better, it was assumed to be reversion to the mean or spontaneous remission; when the control group got even better, that difference was the placebo effect; and when the treatment group got even better, that difference was the treatment effect.
Going from 1. to 2., it was realised that many treatments are not quite as potent as hoped, and that the placebo effect is quite strong. However, going from 2. to 3., it was realised that the placebo effect is not quite as potent either, and that there is considerable spontaneous remission or reversion to the mean.
