i wrote this a few years ago. how "bayesianism" is actually done.