You’re the second person to use this hypothetical (unless you were the first one too…). But it’s really not that far off at all. In fact, right now Nate has it as 50.7 to 48.2 (not quite the 2-point margin you cite, but pretty damn close). And he has the most likely EV total for Barack Obama to be… 330.
A 2-point win the the popular vote is a pretty large margin, and will almost certainly translate into a significant EV margin, particularly if it is Obama with the margin (because he currently has more “safe” EVs). GWB won by 2.4% in 2004, for example.
Your other example is right on, of course. If Romney gets a 3-point gain in the national vote he will go from a 2-to-1 underdog to a legitimate favorite.
There’s even currently (according to Silver again) a non-insignificant chance that Obama loses the popular vote but wins the EV - 5.5% or so right now. That’s how solid the state polling has been for Obama vs. the national polling.
Another question: How much do you guys think the GOP convention will be able to influence the election and/or the polls?
I was listening to a few pundits yesterday and they said that Romney could start to turn things in his favor if he delivers a rousing speech at the convention. Still, I don’t know how true that is or if there’s historical precedent for that sort of thing actually happening.
Conventions usually cause a bounce, but the bounce tends to fade back away. Lots of people got swept up in the convention bounce last year, but it was still just a bounce. It was probably fueled by the buzz around Palin. I’m not sure that Romney can (or should) find a buzz-worthy running mate right now.
To be more precise, since 1964 the average convention bounce is around 6 percentage points. Nate Silver had a nice article about this in 2008. I’m sure he will do an updated one for this cycle as well.
He commented during the GOP primary cycle this time that he wishes he had left his “compensation” for the convention bounce in the forecast in 2008, as it actually proved rather accurate - Obama got a quick bounce which was immediately wiped out by Palin + the GOP convention. Then that bounce went away and we went right back to the status quo ante.
It’s fascinating IMO that RCP now has it more likely that Obama will win in Virginia than in Colorado.
Granted, the latest Colorado poll is a bit of an outlier, and at this point I’m still conceding Virginia to Romney and CO to Obama. It could obviously go either way, though, and the whole thing is very interesting to watch.
As an update, the October surprise was Romney’s debate performance full of lies and misinformation. So I was half right (I thought it would be an ad blitz).
Based on current polling, I’m predicting Obama gets all the swing states except Florida (29) & North Carolina (15), and no split in Maine/Nebraska for an Obama win of 303 to 235. The only state I’m wavering on is Colorado.
I was initially annoyed by the lack of partial voting in Nebraska and Maine until I saw a spot to do that in the lower right hand corner. But from what I hear, Nebraska has been reapportioned to prevent the one Democratic seat they got last time, so the point is probably moot.