Nate Silver @NateSilver538
@DanRosenheck @gelliottmorris @mattyglesias Let's just say I've spent a *lot* of time on this problem. ;-) And I'm pretty sure we have things *roughly* right and our numbers look a *bit* more aggressive than yours. Our method is actually not far from being a blend of an RCP-style average and a polynomial trendline. — PolitiTweet.org