There are essentially two different "modes" for a presidential primary season. In the first mode, a candidate wins decisively in the early set of primaries, and his momentum carries him forward to victory. All of the fighting really takes place in the early primaries. This is the mode that characterizes almost every presidential primary season.
Technically speaking, the winner of the presidential nomination is determined by the number of delegates he wins. Each state has its own rules on how the delegates are chosen: Some are winner-take-all, some are based on districts, some are proportional, some are a combination of different methods. And some states are penalized by the national Republican Party for breaking the calendar.
Normally, little attention is paid to the delegate math. Partly that's because it's difficult to follow, but mostly it's because the delegate math usually doesn't matter. In the "normal mode," the winner is obvious long before people start counting up the delegates.
But in the second mode, the "delegate mode," the outcome of the race isn't obvious after the early states vote. In such a case, the race goes on through every state (or a lot of states), and the candidates must grind out as many delegates as possible.
If Mitt Romney had won South Carolina and gone on to win Florida, the race would be in "normal mode." Arguably, if Romney wins Florida, the race will soon become characterized as one-sided, with Romney going on to win the next several states. As suggested a few days ago, the media will turn against Gingrich as quickly as they turned in his favor after South Carolina.
If Gingrich wins Florida, particularly if it's a strong finish, the race will enter "delegate mode." The race will become a contentious battle until Romney wins enough delegates to clinch the nomination, or until the convention. Since the Republican establishment views Gingrich as an unacceptable, unelectable nominee, there will be no giving up on the part of Gingrich's rival(s).
A few days after the Florida Primary, the Nevada Caucus will be held (February 4th). Romney is expected to win there, as he did in 2008. He won't get much credit for this firewall, as the Mormon vote will play a big role. But Romney will at least be able to claim a 2-2 tie with Gingrich.
There are four non-binding contests in early February, taking place in Maine, Colorado, Minnesota, and Missouri. Romney is expected to win most or all of these, as he won Maine, Colorado, and Minnesota in 2008, and since Gingrich failed to qualify for the ballot in Missouri. (Gingrich will claim Missouri doesn't count, but Romney will credibly argue that Gingrich's failure to get on the ballot is significant in and of itself.) Having done well in these contests, Romney will regain some momentum lost after defeats in South Carolina and Florida. If Romney sweeps Nevada and the four non-binding contests, his campaign will begin pushing the narrative that Gingrich can only win in the South.
On February 28, there will be two primaries, in Arizona and Michigan. The situation in Arizona won't be totally clear for some time, but it should be a reasonably good state for Romney. Michigan is expected to be a strong state for Romney, since it's his home state and the state where his father served as governor.
Thus, although Gingrich's chance of winning the nomination would increase substantially upon winning Florida, his momentum would be blunted by Romney's wins in the February contests. That indicates Gingrich's best scenario is a protracted battle, since an outright win is not possible under these conditions. At most, Gingrich would be considered even with Romney; at worst, the narrative would be that the momentum has shifted against him again. By contrast, an outright win is possible for Romney if he wins Florida.
