
Nate Silver @NateSilver538
@DanRosenheck I don't know if it's symmetrical, but I think there are arguments that the asymmetries run in either direction. I think the best prior is probably a symmetrical distribution with a wide range and fat tails. — PolitiTweet.org