Ah. I guess the term we're looking for is "physiologically meaningful"; I wasn't talking about "statistical significance". Something can be statistically significant but have a clinically worthless effect size.

FYI the standard null hypothesis is that differences are not significant until shown otherwise. E.g. no doctor will prescribe 50mg metformin to a prediabetic even if it might have an effect; it will be physiologically meaningless.

> the numbers you are fixated on are for athletes or neurotropic nerds

If you have other numbers/studies to back "0.5g of creatine, I feel, is a useful amount", then I'm willing to "fixate" on those numbers.

Ok.