I'm big on jinxes, so I hesitate to post this. But here it goes:
I've been following the state polls somewhat obsessively. Kerry's had some momentum, but it seems to have slowed right as he draws even (in the national numbers) in the last day or so.
Here's my prediction: I think the difference will be large enough tomorrow that we won't have a drawn-out fight about who won. That sliver of undecideds will break one way, and all the close states will tumble in that candidate's direction, making for a sizable electoral college victory.
There is considerable history that suggests undecideds break for the challenger. I'm not sure if I'm buying that this year. It may be that people who think that W. has screwed just about everything up (budget, Iraq) but think he's stronger on terror (as most do) fall in line with him, despite their objections to every other aspect of his presidency. (Listening to WBZ coming back from Boston yesterday, Bush owned all the commercial air time and every spot - every one - was a variation on the theme "terrorists will kill you if you don't vote Bush." That's the only area where he has a clear polling advantage over Kerry, and he is hitting it hard.)
There is considerable polling data in this election that shows that independents - whom Gore lost despite his slight victory in total votes - are strongly for Kerry. I believe in nearly every election, the winner has caried the Independents. (Don't have a cite handy for that, though - could be wrong.)
There is some evidence that more Republicans are "staying home" and sticking with their candidate than Democrats (even though overwhelming majorities of both parties are backing their guy), but Dems typically comprise about 4% more of the electorate than Republicans. This is also offset by the plurality Kerry seems to have among independents.
Because Republicans have been the minority party (in sheer numbers) for a lot longer, they are far better at identifying their supporters and turning them out. Thus, I hold no great hope that the massive GOTV efforts by the liberal 527s will make the biggest difference. However, most peripheral voters (or "ringers" as one Ohio GOP muckamuck referred to them) identify with Democrats. So if the 527s are as good as the GOP, that's a net plus for Dems.
Looks like rain in Florida (although mainly in the panhandle, GOP turf) and all of Ohio today. Lower turnout favors Republicans. However, lots of buzz in the black social networks about the voter registration challenges in inner cities in Ohio and elsewhere, and regardless of whether they are justified, that could motivate black voters to do whatever they can to get counted. The NAACP announcing a few days ago that the IRS is investigating it doesn't hurt. (Stories of folks waiting five hours in the sun to vote early in Florida anecdotally suggest a strong motivation.)
Most polls show that Kerry is leading among early voters, but there is no telling whether that indicates anything about the trend for the electorate as a whole, or even about the GOTV efforts of either side.
Assuming both sides do a good job of getting out their voters, it's going to be the way those undecideds break that makes the difference tomorrow.
Based on the latest tracking poll, Fox's, which uses the last two days rather than the other major tracker's previous three days, and which shows a 48-45 Kerry edge (up from 47-45 yesterday), I'll say that that poll suggests the undecideds are breaking in their historic pattern - toward the challenger - and that that pattern will be followed in the close states. Kerry takes Florida, Ohio, Michigan and Pennsylvania but drops Wisconsin. Roughly 290-300 electoral votes, with the over/under depending on what happens in NH and HI.
Further, Republicans hold a narrow edge in the Senate as Bunning barely teeters across the line, maintain a safe majority in the House, and the attempt to crush Kerry like a bug immediately begins ...