The Importance of Balance in Clinical Research

I don't let the significance or the statistics from a double blind study. Like really determined from anything because I think all of research clinical research is inherently flawed in that. Let's just say there's a hundred people that they're checking to see if a supplement works for depression let's say. And you'll it'll show no change and you see the dots all over the map. You'll have a pocket of 20 people or so that it worked better for and then most of them it was like and no real change and some people maybe got worse. Why is that a failure why are we just saying this doesn't work? That makes sense, right? But at the end of the day