For the most part no one needs to experiment. There are two sides to it
1. Most medication, inc PEDS, has a known effect. Esp so after approval. Hence the white paper insert. 90% of users will respond thus, 5% thus and so on. MD's will occasionally tweak the dosages within a range. As an example it's rare to see 1000mg of Ibuprofen suggested.
2. Of course we do tend to dose, as Stevesmi points out, PEDs at far higher levels than for any medical use. But how many cycles started with 500mg of test a week and how many users need to exceed 1000mg a week? How many do amazingly on 30mg a day of dbol and so on? Nearly everyone. With, as per my own experience, a little variation to what we respond best too