Seriously, Nate Silver predictions for senate aren't extremely inacurrate. Nunn isn't going to have more chance to be electec than Hagan.
He wasn't that accurate in 2012 either.
Are we really going to do this again? Silver doesn't make predictions. He can't ever be "right" or "wrong." There is not nearly enough data out there to analyze his probabilities.
If candidates he gave a 5% chance to win never won, his model would be bad. Around once every 20 times, the 5% chancer should win. We need a larger sample size to know for sure, but since Heitkamp's the only senator to ever pull off the 5% upset by his numbers, and there have probably been around 20 races that he gave around a 95% on, it would be expected that one of them would have won.
