Tomorrow's "too close to call" presidential election may not be so close after all, a woman I know predicted the other day. There's a good chance, she said, that it will be decided by the "unwanted, unpolled unlikelies." That, of course, would be great news for challenger John Kerry, who has never had more than a marginal lead in the polls among those considered likely voters.

Several things prompt this woman's assessment. To start with, the political polls this cycle have tended to undercount Democrats -- in part, according to some reports, because Republicans are more likely to make themselves available to telephone pollsters. Further, college students -- who may be unusually active this election cycle and who may be more inclined to question the direction of the present administration -- are notoriously fused to their unlisted cell phones, meaning that they are unreachable by polling organizations that try to reach specific voters.

But the big thing is a "groundswell of new voters" -- many of whom, though eligible, didn't vote last time. That makes them, in the lexicon of pollsters, "unlikely" to vote this time.

I've been talking to people from the Association of Community Organizations for Reform Now (ACORN) who have been responsible for some of that groundswell. They predict that those of us who've relied on the polls to call this race a tossup may be very surprised.

"The really untold story is the huge numbers of new registrants -- particularly among low-income Americans and communities of color and particularly in the critical states," Steven Kest, ACORN's executive director, told me from Florida.

Charlene Sinclair, the group's national director, said from Ohio that ACORN has registered more than 1.1 million new voters in the past year, 180,000 of them in Ohio. Both numbers, she says, may be more significant than they sound.

She noted that earlier get-out-the-vote drives among communities unused to voting have sometimes led to far more new registrants than actual new voters.

"What makes the ACORN effort different," she says, "is that it is connected to local grass-roots organizations with a permanent presence in their communities. We have local chapters that meet every month, working on issues from schools to minimum wage to predatory lenders. The people who go door to door are not outsiders but trusted messengers who've been involved in local issues. This is not just about turning out a one-time vote. We are teaching people how to make their voices heard. Our people will actually vote."

So is ACORN predicting a surprisingly large margin for Kerry?

"That's not our interest; we are nonpartisan by vote of our board," says Kest, disingenuously.

It hardly seems credible that ACORN -- which claims 700 neighborhood chapters in 75 cities engaged with 150,000 low- and moderate-income families -- could be working on such issues as raising the minimum wage and be indifferent to whether the Democrat or the Republican wins. The minimum-wage issue is actually on the ballot in Florida and may by itself increase the low-income turnout.

On the other hand, Sinclair and Kest say they worry about what they call "suppression tactics" to hold down the vote. These include efforts by party workers to challenge or "intimidate" certain voters and rules enforced by officials to accomplish the same thing -- for instance, by disallowing even a provisional ballot for a person registered in a different precinct.

It's a legitimate worry. Maybe it's only because the accusations are more likely (since the cliffhanger election of four years ago) to be reported by the media, but it does appear that the Republicans in particular are investing more than the usual effort in finding ways to suppress the vote in areas that are predominantly black and poor and thus likelier to vote Democratic.

Will these voters -- many of them first-timers -- allow themselves to be intimidated out of casting ballots? Will long lines at tactically understaffed polling places discourage others? Will significant numbers of would-be voters disqualify themselves by turning up at the wrong place or marking their ballots improperly or failing to respond to a challenge?

Or will efforts such as ACORN's really produce a groundswell of "unwanted, unpolled unlikelies"?

These new voters could change the outcome of the election. Or not.

Remember, you read it here first.