An Obama Landslide? Watch These States.

Call them the Icing States. The candidates aren't visiting. The reporters aren't calling. And the rest of the country barely knows they exist.

With six days until Nov. 4, the political world is focusing on traditional battlegrounds like Florida, where both Barack Obama and John McCain are campaigning todayand understandably so. But as Obama's edge in the Electoral College has expanded over the past few weeksthe latest RealClear Politics map shows him collecting 311 electoral votes solely from states in which he outpolls McCain by more than five pointsa subtler shift has taken place under the radar. Since the start of the month, Obama has quietly, methodically reduced his rival's long-held leads in a quartet of red statesMontana, North Dakota, Georgia and, most surprisingly, Arizonato the point where they could conceivably flip on Election Day.

Crazy? Only a little. George W. Bush won Montana by 20 points in 2004 and 25 points in 2000; now RCP shows McCain leading by a mere 3.4 percentage points. It's worth noting, however, that the RCP average includes three polls taken before Oct. 16. Restrict your results to surveys released in the last two weeksMontana State (Obama +4) and NBC/Mason-Dixon (McCain +4)and you suddenly have a dead heat. That's a dangerous position for McCain in a place where the GOP is in disarray and Democrats Jon Tester, Brian Schweitzer and Max Baucus have a monopoly on statewide office. Which is probably why the RNC just decided to start airing ads on local television.

It's a similar story in neighboring North Dakota, where Bush beat Kerry 63 percent to 36 percent four years ago. RCP shows McCain leading by four points, 47.7 to 43.7. But that's only because they include a Rasmussen sounding from Sept. 8 that gave McCain a 14-point advantage. Limit your sample to polls released since the start of OctoberForum/MSUM (Obama +2), North Dakota UTU (Obama +3) and Research 2000 (tie)and Obama is actually up by 1.67 percent on average. Given that the Sept. 14 Research 2000 poll showed McCain clobbering his opponent by 13 points, this is a sign of real movement.

Neither Georgia nor Arizona appears to be quite as close. But there's been a notable narrowing in recent weeks. A Sept. 29 Rasmussen poll showed McCain winning his home state by 21 points. Now the firm has him ahead by five. Overall, McCain's average Arizona advantage has plummeted from double-digits to about five percentage pointsjust barely beyond the margin of error. The latest Arizona State survey gives him a mere two-point edge, down from a seven points a month ago and 10 points this summer. In Georgia, McCain's average cushion has sunk from about 15 percent in mid-September to 5.2 percent today, and one pollsterthe Atlanta-based Insider Advantagesays it's a one-point race. Add in a 10 percent boost in early voting among blacks, and that's too close for McCain's comfort.

Am I suggesting that Obama will win in Montana, North Dakota, Arizona and Georgia? Hardly. Five points is still a steep climbespecially in Republican strongholds like Arizona and Georgia where undecideds will likely break for McCainand sparsely populated (and traditionally uncompetitive) states like Montana and North Dakota are notoriously difficult to poll.

That said, it's worth consideringfor perspective's sakehow much stronger Obama seems to be in these "reach" states than McCain seems to be in states where he's still said to be competitive. Take Ohiothe king of all battlegrounds. According to RCP, Obama currently leads there by 6.3 percent. Or, say, Virginia, where a Democrat hasn't won since 1964--RCP has Obama up 7.4 percent. Or, for that matter, Pennsylvaniathe last blue state on McCain's target list. Obama's latest margin? 10.5 percentmore than double McCain's lead in Georgia. Judging by the most recent polls, in fact, the odds of Obama winning Montana, North Dakota, Arizona or Georgia are better than the odds of McCain winning either Ohio, Virginia, Colorado (Obama +8.3), Nevada (+7.5), Iowa (11.4), New Mexico (8.4) or any single Kerry state. Them's just the numbers. If you want to say that McCain has a shot in Ohio, you have to admit that Obama has a shot in Arizona. And so on.

None of these prizes, of course, will tip the election to Obama. If he wins Georgia, he's already won Virginia and North Carolina; if he wins Montana or North Dakota, he's already won Colorado; and if he wins Arizona, he's already won New Mexico and Nevada. But if Obama succeeds in "expanding the electorate" in Ohio, Florida, Virginia and the rest of the swing stateswhich, it should be noted, is hardly a sure thingthose gains would undoubtedly extend beyond the battleground's traditional borders. In that case, Arizona, Georgia, Montana and/or North Dakota appear to be the states most likely to boost Obama from a "resounding victory" (say, 375 electorate votes) to a "massive landslide" (more than 400). In other words, they'd be the icing on the cake.