Roger. This kind of thinking is the reason that the double blind clinical study was the greatest medical advance of the 20th Century. Yep. Still, marginally effective and ineffective drugs manage to make it through the process.
Perhaps a better appreciation of the placebo effect will be a major medical advance of the 21th century.
Or not.
The placebo effect is exactly what the double blind clinical trial is designed to counter. The plural anecdote is
not data. Clinical trials are not perfect and sweeping conclusions require repeatable results. There is also the tendency of people with a stake in the outcome to fudge results, but it's still the best we got. I am, naturally, dismissive of this kind of nonsense. For instance, patterns of betting on state lotteries do not show any indication of statistically significant precognition.
James Randi understands these things better than most statisticians and put his money where my mouth is.