zlacker

The problem is that placebo effect is usually close to nil, when it even exists. Most of the time what is actually happening is regression to the mean, not placebo efficacy.