The truth is that [the] whole project is (most of the time) quite inaccurate and error-prone, and often involves way too little data to really make a judgment, despite my best efforts. It also involves my amateur method of “age-adjusting” the results to be comparable. So this whole project is quite inaccurate and shouldn’t be used for serious conclusions. But if you understand the inaccuracies involved, you still may find it interesting.
I can’t think of a more reliable source than testosteronedecline.com
From the website:
“Listen, we know we’re full of shit, but that won’t stop your anti-trans uncle from posting our nonsense on Facebook.”