I think Gwern understands many of the limits of his experimental design. He's has written extensively on a wide variety of statistical and scientific topics. H

I think Gwern understands many of the limits of his experimental design. He's has written extensively on a wide variety of statistical and scientific topics. Here's an example related to placebo that's linked to from his blog. http://www.bluelight.org/vb/threads/689936-My-Paper-quot-Noopept-amp-The-Placebo-Effect-quot?p=11910708&viewfull=1#post11910708 http://www.bluelight.org/vb/threads/689936-My-Paper-quot-Noo... See also: https://www.gwern.net/Statistical%20notes https://www.gwern.net/Statistical%20notes I'm not saying he's a perfect researcher, but give him some credit instead of assuming he doesn't understand what he's talking about.