Any drug with a high efficacy should be seen in small sample sizes. E.g. As a sample size of one, I know that consuming alcohol has a relaxing effect. And that

Any drug with a high efficacy should be seen in small sample sizes. E.g. As a sample size of one, I know that consuming alcohol has a relaxing effect. And that's highly testable, one I wouldn't mind participating in. If n=11 and there was no deviation from the normal expectation, it's possible that there was some undetectable effect, but it's far more likely there was no effect.