Indiana
Al Giordano: Clinton 55.4, Obama 44.6 (Clinton +10.8)
PsiFighter37: Clinton 53.2, Obama 46.8 (Clinton +6.4)
Poblano: Clinton 51, Obama 49 percent (Clinton +2)
Jerome: Clinton 54, Obama 46 (Clinton +8)
Keep in mind that those guys have a "system", that includes looking at historical results, demographics, barometric pressure -- the works. Me? I have you-know-what. And what I'm pulling out of there for Indiana is:
Clinton: 51.1 percent
Obama: 48.9 percent
That's Clinton by +2.2. As much as it pains me to go against my homeboys SUSA who give Clinton a 10-point lead, I just don't see the Clinton blowout. The trend in Indiana has been in favor of Obama, and it's already a tight race -- 48.6 Clinton versus 44.4 Obama per Pollster.com's Indiana polls composite. That's just 4.2 points apart. So it really comes down to who votes and how undecideds break.
Clinton has been closing the undecideds in recent contests at the rate of 3-1. If she does that in Indiana, and the poll composite is accurate (and it has been fairly good this election cycle) she wins ~53-47. But while all the recent elections have seen Clinton begin to expand her lead the two days prior to the election, in this case it's Obama who seems to be getting the late breakers. So rather than the usual 3-1 break, I'm going to guess (and it is just a guess) that the undecideds break 50-50. So add 3.5 percent to each candidate, and you get 52.1 and 47.9.
As for turnout, there are no big machines working the state (Indianapolis now has a Republican mayor), but Obama has the endorsements of the mayors of Evansville (3rd largest), South Bend (4th largest), Gary (5th largest), and Bloomington (7th largest). Clinton has just Hammond (6th largest) and Anderson (9th largest). Between that and the nearby Chicago machine I assume is helping out in northwestern Indiana, I'll give Obama another random point. Hence my 51.1 and 48.9 numbers.
North Carolina
Al Giordano: Obama 52.9, Clinton 46.1 (Obama +6.8)
PsiFighter37: Obama 58, Clinton 42 (Obama +16)
Poblano: Obama 58.6, Clinton 41.4 percent (Obama +17.2)
Jerome: Obama 53, Clinton 47 (Obama +6)
These are either hyper-optimistic or hyper-pessimistic. Mine split the difference.
Obama: 56.1 percent
Clinton: 43.9 percent
(Obama +12.2)
SUSA shows that among early voters, 25 percent of its sample, Obama is leading 57-41. Zogby, proving that it continues to spit out weird-ass numbers, says those who voted already are going Obama 57-34. What, 9 percent of respondents were undecided over who they voted for? They voted for Gravel? SUSA claims the other three quarters of its sample edge Obama by just two points, giving him a 5-point lead.
I can't believe I'm doing this twice in a day, but I've got to depart from SUSA. Every time I've done that in the past it's come back and bit me in the ass, so it might not be the smart thing to do, but there's no glory in taking the safe bet, even if you end up right.
The Pollster.com NC composite gives Obama a 49.9-42.1 edge. Most of the polling shows Clinton gaining, though Zogby claims to show a late surge toward Obama (perhaps they subscribe to the "glory" theory I posit in the paragraph above?). Hmmm...
There are two theories floating around. The first is that pollsters have consistently underestimated the amount of support Obama gets in heavily African American southern states. And it's true, they have.
The counter theory is that if this was February, that yes, Obama would win NC by 20 points. But it's not. It's May, and the primary season has taken its toll on Obama's white support. Of course, Obama's support in the region has always been poor (24 percent in the three-way battle in SC, 39 percent in MS, 25 percent in Alabama, 30 percent in Louisiana), but perhaps the assumption is that North Carolina is more like Virginia, where Obama got 52 percent of the vote, or Georgia, where he got 43 percent.
I'm going to say that the polls underestimate Obama's support, but not by as much as in previous contests in the region. So a +2-point advantage equals 51.9-40.1. I'm going to guess Clinton does better with the late deciders, a 60-40 split, which brings the numbers to 55.1-44.9. Clinton has the governor's endorsement, but he has no machine. The big city mayors have gone strongly for Obama -- he has at least six of them. The largest of those cities, Charlotte, has a Republican mayor. The other two had not endorsed last I checked. The state's entire political establishment, sans Gov. Easley, seems to have moved over to Obama en masse. Let's say that gives Obama another point.
That gives us my 12-point Obama prediction. If my system sounds stupid, it's because it is. They all are. Ultimately, no one predicts every race accurately, not even the pollsters who actually ask people who they're going to vote for.
So yes, my predictions are wrong, even if they're occasionally right (stopped clocks and all). I'm under no illusions otherwise. I do this because it's fun, not because anything is riding in the predictions themselves. It's not like I'd put money on them.
So have fun and make your own predictions.