A solid case for hyperventilation

Nate Silver does the knowledge, but with precision. He finds bad news for Obama. Not all bad, but bad enough that the worst could happen:

The most disappointing for result for Obama is probably in Indiana, where SurveyUSA has John McCain pulling into a 6-point lead after having trailed by a single point last month. Why so disappointing? Because Obama has been investing heavily in Indiana while McCain has not. A couple of caveats, though. Firstly, investments in the ground game may not show up in the polls in the first place. And secondly, the partisan leaning of the sample has shifted a fair bit more Republican than the last edition of this poll. It's possible that, as McCain enthusiasm grows (and Bush fatigue wanes), more Republican-leaning independents are now willing to identify themselves as Republican. It's also possible that we're just looking at some static.

It is officially time for Obama to be worried in Minnesota, where SurveyUSA marks the third consecutive poll to show him with a lead of only 2-4 points? Our model says ... maybe not qute yet. There has certainly been a pretty big shift in the raw numbers in the Gopher State, but there aren't really any demographic explanations for it -- Obama hasn't lost much ground in demographically similar states like Wisconsin and Iowa. So our model is going to need a little more coaxing before it considers Minnesota a toss-up. It might be close enough, however, that there is an electoral rationale for McCain to pick Tim Pawlenty.

Amy Klobuchar anyone?