Patrick Ruffini of Echelon sees it this way:
First, Donald Trump is going to win South Carolina. Don't get your hopes up about a miraculous surge for anyone else.
The data I'm seeing suggests that far from being highly fluid, the race is actually settling into a pattern.
Trump is not subject to regional fluctuations in support because his coalition is attitudinal, not ideological.
The percentage of voters who think "F*ck it" is pretty constant from state to state.
This means he gets 35% vs the current field. What's striking is how little difference vs New Hampshire.
At the same time, Trump support isn't necessarily *growing* as we might have expected coming out of NH.
In fact, the pattern from IA>NH>SC is not one of state to state momentum but fixed patterns that are likely to repeat.
That means Trump=25% in caucuses, 30% in low turnout closed primaries, 35% in high turnout primaries like NH and SC.
This means fairly sizable regional fluctuations for Cruz or whomever the establishment candidate that emerges.
As we move to a three man race, Trump will have 40, Cruz will have 30, establishment candidate will have 30 *nationally*
This puts Trump in pole position. Winnowing to 3 won't be enough. Field will need to winnow to 2 to defeat Trump under current dynamic.
It seems like we have a 50/50 shot if the field winnows to two, with very little margin for error.