Isn't the whole point of a prior that it doesn't have to be very close to the real probability? It's literally supposed to be some value you're uncertain about and then you refine it when you get more data...
Quote Tweet
Conversation
Replying to
that may be a bad way of describing it, but my understanding is that priors can vary in how "informative" they are, so just invoking Bayes' theorem and saying it's a prior is meaningless
1
anyway my prior is that if someone believes race is genetically determined, they probably don't actually know what that even means