Expertinent is a regular Stumper column featuring interviews with experts on the news of the day.

Nate Silver is channeling Nostradamus. As a University of Chicago econ alum slogging through an undemanding post-collegiate consulting gig, he developed a comprehensive historical database that could project the future performance of any pro baseball player by matching him to a comparable predecessor--down to, like, his height, his weight, his career singles and the size of his home stadium. Called PECOTA, it's now recognized as the most accurate baseball forecasting system on the market.

But apparently Silver wasn't satisfied. In October, he started posting anonymously (nom d'écran: "Poblano") at DailyKos, where his detailed district-by-district projections of who would win each Democratic contest (and by how much) soon earned him a massive following. Relying on demographic data from previous primaries and ignoring the usual mishmash of polls, Silver ultimately came within 20 delegates of the final split on Super Tuesday (out of nearly 1,700) and nailed the margins in Indiana, North Carolina, Pennsylvania, Oregon, Montana and Kentucky. Taking a break from blogging at FiveThirtyEight.com--not to mention his new duties writing for The Guardian and The New Republic--Silver (whom I profiled in this week's dead-tree Newsweek and who supports Obama) spoke to Stumper about where the McCain-Obama match-up is going next. Excerpts:

So we'll start with an easy question: Who's going to win the election?
Well, if you look at where the polls are right now, it's going to be just as close as it was the last couple of cycles. That said, if anyone wins by a large margin, it's more liable to be Obama. Right now, he's tied with McCain, even though he's losing 20 percent of Democrats to McCain. That number is pretty unprecedented-it's usually about 10 percent. If Obama can get that number down to 15 percent, then that represents a bump of four points overall: you're basically taking two points away from McCain and giving two to Obama. That would represent a landslide compared to recent elections.

The fact is, the fundamentals favor Obama. He is tying McCain among Independents, and there's four-to-three ratio of Democrats to Republicans in terms of voter ID. If those numbers stay the same, he just has to hold his own among Democrats-maybe even losing a few more Democratic voters than John Kerry might have because of race or because some Hillary supporters stay bitter-and he wins. That's especially important states like Ohio where you had big swings at the state level to Democrats, and Hillary picked up the bulk of that support.

Can he win them back? A lot of Clinton supporters say no-they're going to vote for McCain.
I think by the time it gets to September, Hillary die-hard supporters might still not love Obama, but they won't love McCain either. At that point McCain's going to be advertising to his base-Republicans-and Democrats are going to be advertising to theirs. So he's going to look unacceptable to Democrats, even if they supported Hillary. McCain's goal is to win 60 percent of the Independent vote to counteract by the Democrats' edge in party ID, which he'll do by saying Obama is too liberal.

We are talking about this in sort of a macro sense, but it's the electoral college that will decide the election.  It's state by state. Now, I know on your site is keeping track of every state-by-state poll that comes down the pike. How is the map shaping up?  It seems to me that if Obama can pick up New Mexico, Iowa and Colorado, he doesn't have to worry as much about places like Ohio and Florida where he isn't quite as popular as Hillary among Democrats.
Right. Or, if you take the Kerry states, add Ohio-which I think Obama's almost certainly going to win--and then add Colorado and Iowa, then that's a winning map by a few electoral votes for him. Actually, I think it's one of the more likely maps.

If you had to predict a couple of surprises, what would they be?
You want to get exotic? If Obama just won Colorado and Iowa-not Ohio--and picked up one electoral vote from Omaha, in Nebraska, then he would tie 269 to 269. Now, McCain is going to win Nebraska. He'll totally clobber Obama. Republicans usually win the state by 20 points or more. But the thing is, Nebraska divides its electors up by congressional district, and there are two districts out of three where Obama might be competitive. One is the city of Omaha, where Obama is running about 10 points better than the state overall, and the other is focused around Lincoln and the University of Nebraska and borders Iowa, where people seem to love Obama for some reason. He could win one of those districts, and with things being so close, that could be the technicality that decides the election. In light of what happened in Florida in 2000, it would almost be fitting, no? Tying 269-269 by winning the city of Omaha, then having it go to the House, where the Democrats would vote for him. Farfetched, but possible. That one electoral vote could matter.

Any other possible surprises?
Indiana is an interesting state. It's always a state that maybe shouldn't be as red as it is. Like, the Bill Clinton maps from '92 and '96--Bill won in a lot of states where Democrats don't often win. Indiana is a little lake of red in the middle of a lot of blue. It's unusual that it hasn't gone more Democratic. I think part of the reason is that some of these states haven't really had a democrat campaign there for years.  Indiana, they've just written off. It's usually voted Republican, but it's also a state that has a huge manufacturing industry. It has the same concerns that Ohio has. Which is why, I think, our current projections show Obama losing by less than four. North Carolina is another state that might be reasonably competitive; McCain's up by about six in our system. I think in the course of the long Democratic primary campaign, Obama discovered a couple of hidden swing states where he can compete that he otherwise would've written off.

What about Virginia?  That's another one that people toss out there.
I think in four years Virginia is going to kind of a solid blue state, because it's becoming more northern. So while Virginia is a good opportunity for Obama, whether it gets there right away, I don't know.  The western panhandle of Virginia reaches into Appalachia, where he's going to lose pretty badly. But Virginia-along with Missouri, where Obama is also polling pretty well-are going to be the only Southern tossups. But I think he's not likely to win a state like Mississippi, which the campaign says it's targeting. They talk about turning out the African-American vote out in Mississippi, but even if you do that, it's not going to happen. Turnout among African-Americans is pretty good already. 

That said, I'm watching Georgia.  I'm fascinated by it. The polls show Obama trailing McCain by about ten points, but they also have former Congressman Bob Barr drawing six to eight percent. So if Obama can increase African-American turnout and inch up a few points, it seems like he might have a shot.
Don't forget: it's also one of the youngest states in the country. Alaska could be interesting, too.  Thanks to its harsh climate and hard-core industries, it attracts a very hearty, very male population. Historically, there's a lot of affection for third-party candidates in Alaska, so it's a state where Barr could get five or six or seven percent of the vote-which is about how far behind Obama is in the latest polls. It might be enough to tip it. In which case, we'll be up until 4:00 in the morning waiting for the returns from Juneau.

What about the reverse?  We have been talking optimistically about Obama, but what about McCain?  He's more competitive in places like Pennsylvania or New Jersey or Michigan than any other Republican would be.
There are a lot of Independents in Michigan, and they certainly seem to have a lot of affection for John McCain.  They certainly did in 2000, and right now he's leading in our projections by a tiny, tiny margin-one-tenth of one percent. I think that the fact that Obama didn't get to mobilize there in the primary campaign probably hurts him, but that at the end of the day, Michigan leans Democrat. Still, McCain's could win. He's got a special relationship with the state. Wouldn't be interesting if Obama wins Ohio and loses Michigan? That's what our map shows him doing right now. It's one of millions of permutations, but it's possible.

Overall, though, Obama is in a better position than John Kerry was.  In Pennsylvania, Obama has pulled enough ahead now where it's not likely to be supercompetitive. And in states where McCain otherwise might be especially strong-- like in the West-any "native son," regional advantage is probably outweighed at this point by changes in demographics and the local political cultures. Arizona won't be competitive, of course, but those neighboring states-New Mexico, Colorado, Nevada-will be.

Where McCain does better than Bush did, actually, is on the East Coast. He might stay within eight or ten points in Massachusetts, whereas Bush lost by 20. Or Connecticut might be within in seven points But he's not going to do well enough to actually win then. Because it's his strongest region, he gains popular votes--but not necessarily electoral votes.