You don't have to be a Nate Silver worshipper to realize he's probably right about 2014.

[DanTex]

And by "right" I don't mean that the GOP is going to win the senate, just that, based on the current evidence, they are favored to win.

Nate Silver has a good model for evaluating the odds of election outcomes. Which, by the way, isn't that difficult. You basically take the polls and average them, plus a few other adjustments. In past elections, there have been other people with similar models, and they came up with similar predictions. The data is what the data is, and believing in the accuracy Nate Silver's projections doesn't mean believing he is some kind of super-human psychic, it just means believing that he is capable of building a solid statistical model based on polls, and that he is willing to look at the data and not let his personal beliefs affect his projections. Whatever you think if him, and whatever his political beliefs are, I think he's shown that he is capable of that.

This doesn't mean Nate is right, or even probably right, about anything else. For example, turns out he has hired a climate change denier for his new website, and he has a climate change denial-ish chapter in his book. So he's wrong about that, at least. Why? I don't know. For one, climate science is a lot more complicated than averaging polls. Maybe it's a marketing ploy, he wants to be seen as some kind of maverick. Maybe his ability to outsmart clueless pundits on Meet the Press with rudimentary statistics has inflated his ego to the point where he thinks that a few simple calculations make him more knowledgeable than scientists with decades of research on their side.

For whatever reason, his analysis of non-electoral stuff isn't the same quality as his election models. But his election models are still good. So, yeah, we're behind right now.