When it comes to attributing a symptom you're feeling to a cause, the number of correct-negatives in COVID testing indicates a lot of noise (not covid) vs signal (true covid symptom). With placebo effects it's the same problem.

So I mean that when 'active placebos' are used, rather than ones that have no side effect (like dry mouth), the effects are larger. So in basic terms, the more noise there is in the system, the more signal error there is. The main point is about the reason they do blind control studies. People are not very good at correct attribution.