Heading into Election Day, Gov. Rick Scott and Charlie Crist are tied at 42 percent each, according to the averages of 20 public polls released in October, when absentee-ballot and then in-person early voting began.*
Technically, Scott has a lead of .07 percentage points (42.03 to Crist's 41.96). That's not really a lead at all. Assuming the undecideds don't vote, Scott gets 45.9 percent to Crist's 45.8. Libertarian Adrian Wyllie could pull 8 percent of the vote, about the same percentage as the undecided, who are an X-factor in all of this.**
Unlike prior poll-unskewing attempts, this exercise uses polls taken as people were casting pre-Election Day ballots, 3.1 million as of this morning. The GOP led Democrats in ballots cast by 98,000, 43-39 percent (3.3 percentage points). But independents account for about 18 percent of those voters, and they're leaning Crist by an average of about 6 percentage points.
Put it all together, and Scott's lead in early votes is 42.6 percent to Crist's 42 percent.
That's a 0.6 percentage-point lead.
That's not much of a lead at all, considering Democrats have a registered voter edge of about 455,000 over Republicans (obviously, not all vote). Scott's lead could be higher or lower because this analysis includes the undecided. Take out the undecided, Scott's lead remains at 0.6 percent.
Still, a lead is a lead.
What does Crist need to do to have a real shot at wining under this scenario? Have Democrats today turnout by 2 percentage points more than Republicans (assuming there's 49 percent turnout). Democrats would need to cast 42 percent of the Election Day ballots, Republicans 40 percent and independents 18 percent. If that happens, Republicans would still wind up casting slightly more overall votes in the election (including early and absentee ballots).
Probable? Not judging by history.