"So whattaya think?"

I'm getting that question a lot, as Pennsylvania Democrats - old and new - head to the polls today in expected record numbers. And I generally respond like this:

"Beats the heck out of me."

Naturally that is not viewed as a sufficient response, but, given the events of January, when Hillary Clinton foiled the predictors and won New Hampshire, it seems wise not to bloviate excessively about the unknowable. This is an election season like no other in my long memory, and we alleged seers have been chastened too often already.

So I'll confine myself to discussing a couple factors that could well shape the Pennsylvania results:

The army of the undecideds. The final round of polls report that roughly 10 percent of the Pennsylvania voters had not yet decided between Clinton and Barack Obama. That's a sizeable number of people; if, as widely expected, this primary draws a record two million voters (or 50 percent of the Democratic registration), this means that 200,000 Democrats haven't made up their minds.

And if the past is prologue, this translates into a sizeable advantage for Clinton - one that could arguably add several percentage points to a Clinton victory.

Notwithstanding Obama's successes in 2008, the inescapable fact is that he has been a poor closer. In most of the primaries thus far, he has been spurned by those voters who withheld their choice until the eleventh hour. The late undecideds have broken for Clinton in almost every contest, opting to go with the known quantity instead of taking a leap with the new guy.

The exit polls tell the tale. A sampling:

In Ohio, 12 percent of the voters decided on the final day. Clinton won those voters by 11 points, and the overall contest by 10.

In Texas, 11 percent decided on the final day. Clinton won those voters by nine points, and the overall contest by four.

In Massachusetts, 18 percent decided on the final day. Despite Ted Kennedy's ballyhooed Obama endorsement, Clinton won those voters by a whopping 20 points, and the overall contest by 15.

In New Mexico, 14 percent decided on the final day. Clinton won them by 11 points, and the overall contest by one.

In California (a state where an Obama win would have shoved Clinton toward the exit door), 14 percent waited until the last day, and Clinton won them by eight points, cementing her victory.

Even in states where Obama was victorious, the undecideds trimmed his margins. In Virginia, 10 percent decided on the final day, and he split them with Clinton, roughly 50-50. In Wisconsin, 12 percent held back until the final day, and they too split roughly 50-50.

It's hard to imagine that undecided Pennsylvanians will break for Obama today; the state's political culture has long preferred familiar brands to the flavor of the month. And the latest surveys indicate that the undecideds are heavily concentrated on Hillary-friendly turf. A poll sponsored by MSNBC, McClatchy and the Pittsburgh Post-Gazette reports a Clinton lead of only five points statewide, but finds that 11 percent of the folks in the so-called "T" region (between Pittsburgh and Philadelphia) were still undecided on primary eve.

Logic suggests that the late-deciders will stick with the person they know, rather than take a risk on Obama. Clinton's eleventh-hour TV ad, the one where she touts herself as the candidate best able to handle everything from Osama bin Laden to energy crises, seems aimed squarely at these voters. If she wins tonight, the undecideds could be crucial in padding her margin and helping her spin the bragging rights to maximum advantage.

And yet, there is also...

The army of the newbies. Roughly 307,000 new Democrats (potential first-time voters and party-switchers) have signed up for this primary. Assuming an overall record turnout of two million, the newbies could be roughly 14 percent of the total. And by every measure, Obama appears poised to win the newbies by a landslide. Nearly half of the new registrants hail from Obama territory - Philadelphia and its suburban counties (Bucks, Chester, Montgomery, Delaware, plus Lehigh and Berks); and roughly a third hail from counties with big college populations (Centre has Penn State, and Union has Bucknell). Some pollsters think that Obama will get anywhere from 60 to 80 percent of the newbies statewide.

So maybe the newbies trump the undecideds, in terms of sheer numbers and greater motivation to vote. Or not. The bottom line, for Clinton, is that her needs tonight are greater than Obama's. She needs to roll up a huge popular-vote victory - say, a 200,000-vote margin (attainable via a 10-point victory with two million people voting), in order to slash deeply into Obama's national lead of 700,000. She needs something of that magnitude to sell to the unpledged superdelegates. The newbies and the undecideds will help determine her future.

-------

Just finished a gig as guest online chat host at The Washington Post. It's all about the primary. The transcript is here.