Can we first agree that this case is a terrible one, in which a man intentionally identified as a woman in order to get the chance to rape women?
There's more cases, but if we can't agree on ground truth in such an obvious case then there probably won't be common ground for calculating statistics.
