It is too damn early to predict an election

This article discusses an election prediction from Moody's. It makes good points, but leaves out the fact that it is too goddamned early to predict the fucking election. Sure, Moody's predicted the last election correctly. So what? Any idiot could have predicted that Obama would trounce Mittens.

Presidential elections in this country are weird beasts. We see trends that involve economics (a sometimes great indicator, but sometimes lousy), social issues, international issues, personalities, the moral climate, the weather (yes, kids, weather influences voting) and too many other factors to list.

Demographics play an outsize role. Right now, the Dems have a leg up due to demographic trends, slight economic improvement, and changing views on social issues. That all equals a slight advantage, not a huge one. 

I also dislike the fact that Moody's thinks, this far out, that they can state that the election comes down to Virginia and Ohio. They are making a huge error calling Florida for the Republicans. Right now, I'd put FL as a toss-up - if the Republicans nominate either Bush or Rubio, FL goes Democratic (I'm going against conventional wisdom here, but I bet I'm right). Ohio could (could) be important, but my models suggest that North Carolina is more likely to be the key this time out.  Virginia is going blue again.

The article in question doesn't care who the candidates are. That's a mistake. A Hillary V JEB! match up is a very different beast than a Bernie V Cruz fight. Let's assume that the candidates are Webb and Walker. In that case, who the fuck knows who wins what?!? Webb is as likely to win Georgia as Walker is to win South Carolina.