John Heilemann reminds us that Romney won 31.6% of the vote in New Hampshire in 2008. He thinks Romney must exceed this percentage to get a decent bump going into South Carolina:

The expectations game no doubt drives some readers crazy, but there is no getting around it. And while the expectations in question are set unscientifically, they are neither wholly subjective nor wildly irrational.

What the press and the rest of the political class are focusing on are a pair of related metrics: the percentage of the vote and the margin of the victory that Romney claims compared to his performance in 2008. In the aftermath of the caucuses in the Hawkeye State, please recall, the credit given to Romney for his victory was tempered slightly (and fairly, I think) because of its astonishing eight-vote narrowness and the fact that he won no higher a percentage than he did four years ago — despite being the national front-runner and a thoroughly known commodity this time around. If something similar (or worse) occurs in New Hampshire, the nagging questions about Romney's strength as a presumptive standard-bearer will persist and might even be exacerbated.

Nate Silver's final projections look good for Romney.