Randomness beats faulty prediction

I've signed up for an Aeon feed and so far am finding the occasional interesting article.  Here's one that examines how seemingly meaningless augury techniques might be a good way to break people of the habit of using affirmatively harmful predictors.  As I see it, though, the real trick is the double-blind study:  one group uses the proposed predictors while the other relies on a random draw, then you figure out which produced better results.  Randomness for the sake of randomness doesn't have the appeal for me that it seems to have for the author.  But as a way of deciding whether that herb supplement is worth the money and risk, sure.