President Obama's reelection Tuesday looks increasingly likely, but it's not certain. There's still a chance Mitt Romney could surprise pollsters and even some Republicans who've grown grim over the weekend. As The Atlantic Wire's Gabriel Snyder showed, Romney faces long odds, winning election in just 76 of the 512 possible combinations of the nine swing state combinations. That is just 15.8 percent. Here are the four that Republicans think are most likely:
1. Win Ohio
Theory: Romney wins Colorado, Florida, North Carolina, Virginia, Ohio. (275 electoral votes)
Argument: Polls are predicting a much higher turnout for Democratic groups in Ohio than is likely, given the inevitable decline in enthusiasm after Obama's four years in office, not to mention dissatisfaction with his presidency. Republicans have the enthusiasm and "passion," The New York Times' Monica Davey and Michael Wines report. They are united by their dislike of Obama. "Obama can get people as different as white supremacists and Black Panthers to walk in the same direction," George Brunemann, who is head of the Cincinnati Tea Party, told the Times.
Republicans say whites will be at least 74 percent of the voting population, as it was in 2008, while Democrats think it'll be about 72 percent. Romney wins whites by 20 percentage points, while Obama wins minorities by 60 percentage points, NBC News' First Read explains.
In Ohio, 1.6 million have already voted, and 29 percent of them were Democrats, while 23 percent were Republicans. Republicans expect to turn out a huge number of their people on Election Day. The Daily Caller's Nicholas Ballasy thinks a lot of them will be social conservatives, who "delivered the state for George W. Bush in 2004."
Problems: Democrats have gotten more of their people to the polls so far. And Obama's turnout operation is far more sophisticated than Romney's -- the Republicans are using pencils and paper, the Times reports. Obama is leading by an average of 2.9 points, and since there have been a ton of polls in Ohio, it's not just that the polls have to be wrong, it's that a whole bunch of them have to be wrong.
2. Win Wisconsin
Theory: Romney wins Colorado, Florida, North Carolina, Virginia, Wisconsin, plus Iowa or New Hampshire. (273 electoral votes and 271 electoral votes, respectively)
Argument: Though Wisconsin went for Obama by 14 points in 2008, the state had been trending more and more Republican. The failed recall of Gov. Scott Walker helped build up Republicans' turnout operation. Picking Paul Ryan, who represents Janesville, might have helped with voters who want to support a homestate hero. Wisconsin's swing voters tend to be rural, working class, and white, The New York Times' Trip Gabriel reports, and those groups favor Romney. Republican internal polls showed Romney 1 point ahead last week, the National Review reported.
Problems: No poll has showed Romney ahead in Wisconsin since August 19. Plus, Obama is averaging a 2-point lead in New Hampshire and a 3-point lead in Iowa. These states have smaller minority populations, so it's harder to make the case that there will be decreased turnout among blacks and Latinos, a factor not captured by polls and big enough to swing the state.
3. Win Minnesota
Theory: Romney wins Minnesota and, um, a whole bunch of other stuff. (321 electoral votes)
Argument: On ABC's This Week, George Will predicted a huge Romney victory because he'd win Minnesota, which has a marriage amendment on the ballot. This will bring out a huge number of evangelicals, Will says.
Problems: Few people other than Will think Minnesota is within reach for Romney. Minnesota only has 10 electoral votes, so Romney would need states that have leaned toward him, at least until recently -- Florida, Virginia, North Carolina, Colorado -- plus some other stuff that Will does not specify. The map at left shows how much Romney would have to win to get to Will's number, 321 electoral votes. But just to get to 270 votes, Romney would need Minnesota and Iowa or New Hampshire. Plus, Bush pollster Matthew Dowd says gay marriage ballot initiatives did not substantially increase turnout in 2004, when a majority of Americans opposed gay marriage.
4. Win Pennsylvania
Theory: Romney wins Colorado, Florida, North Carolina, Virginia, Pennsylvania. (277 electoral votes)
Argument: There hasn't been much campaigning in Pennsylvania all year, so voters might be more open to changing their minds with just a few ads and campaign stops right before the vote. Romney and pro-Romney groups are spending $12 million on ads in the last days of the campaign. An anonymous Republican detailed how the numbers would have to work to The Daily's Dan Hirschhorn: Obama has to win by less than 375,000 votes in Philadelphia, while Romney has to at least tie in the Philly suburbs.
Problems: There are a million more registered Democrats than Republicans in Pennsylvania. Democratic strategist Neil Oxman told The Daily that since 90 percent of Democrats will vote for Obama, and 90 percent of Republicans will vote for Romney, the Republican is likely toast. Or, if he's not toast, then he probably won't need Pennsylvania: "If the whole things collapses to a turnout model like 2004 or 2010, then it’s collapsing everywhere else… If [Obama] loses Pennsylvania, he’s like Jimmy Carter losing every state."
Biggest problem of all: How it looks right now, according to Real Clear Politics' polling averages: Romney wins Florida, North Carolina. (235 electoral votes).