@pschwede well there might not be your right. But thats why statiscians run the numbers and see rather than make the assumption in the first place. No matter what bias they have or what they may assume int he end the numbers will set the record straight. If there is bias it will demonstrate the bias is wrong.

But no we disagree, sort of, If by unkown you mean "have no data" then yea, anything we have no data for cant be predicted. If we have data we can, however, make predictions on it, even if what we know or assume to know about that data is false. Because statistics doesnt care about your assumptions.