(Note: The figures in this post are based on the 2010 national exit poll for House races, available here.)
Here's one way to look at the intensity gap: in 2008, 53% of voters cast ballots for Barack Obama and 46% cast ballots for John McCain. In 2010, there were the same number of Obama voters as McCain voters -- 45% each. Here's the chart showing how voters in the 2010 midterms voted in 2008:
Keep in mind that this isn't a comparison to other mid-term elections (I'll do a post on that later today), it's a comparison between 2008 and 2010. And the point it makes is clear: more Obama voters than McCain voters stayed home on Tuesday. It's worth debating why that's the case (e.g.: were Obama voters depressed, or were McCain voters just super excited), but there's no debating that it was the case.
It's also true that in addition to being less likely to turn out than McCain voters, Obama voters were more likely to vote Republican than McCain voters were to vote Democratic. 13% of Obama voters cast ballots for Republican candidates while 7% of McCain voters cast ballots for Democratic candidates.
As for voters who either voted for a third party or didn't vote in 2008, they broke 3:2 for the GOP.
So you have a situation where (a) Obama voters were less likely to show up than McCain voters; (b) Obama voters were more likely to flip to the GOP than McCain voters were to flip Democratic; and (c) voters who didn't vote for either McCain or Obama in 2008 voted Republican.
Obviously, that's a bad combination for Democrats. But what was the relative importance of each factor for Democrats? I'll take a look at that in my next post.