Romney could lose New Hampshire. Here's how.

In my view, Romney needs near 40 percent of the New Hampshire vote to avoid what would prove to be a significant defeat, regardless of the first reaction of political insiders. The key number is not Mitt versus his conservative opponents, but Mitt versus 50 percent-plus-one of delegates, in a GOP system where proportional representation makes this harder than many analysts realize.

I agree with most of the points raised by my colleague A.B. Stoddard in her well-done post yesterday. Let me take the analysis even further, into the convention, with my view that Romney is the most likely nominee but the chances of a brokered convention or a post-Super Tuesday draft of someone like Mitch Daniels are greater than people think.

If Romney keeps winning 28 percent, 32 percent or 36 percent against a divided field, how does he get from there to 50 percent-plus-one of actual delegates? It will include Ron Paul delegates, who would never go to Romney, but who might go to a conservative alternative. Ron locks out his percent from any combination that gets Romney to 50 percent of delegates.

Considering the enormous influence of the conservative base on the GOP, and their intense feelings, the pressure against Gingrich, Santorum or Perry delegates (if Perry has any) will be excruciating against any deal with Mitt. The anger from the right toward any conservative candidate who puts Romney over the top will be bloodthirsty.

The only other candidate who would be naturally predisposed toward Romney would be Huntsman. However, if Huntsman does well in New Hampshire he could become a formidable candidate himself, or a power player who leads a move toward a more conservative alternative to Romney, possibly making Huntsman VP in the deal.

In short, while I believe Romney is more likely than any other to be nominated, a) I believe his odds are less than the consensus does, and b) I believe the odds on a move to draft a new, more conservative candidate, or a brokered convention, are larger than other analysts believe.

Personally I hope Huntsman does well today, which is possible, though I never predict New Hampshire voters before they vote!

Politically, if Romney wins very close to 40 percent today I would call this a major victory for Romney, but if he falls below 37-38 percent I would call this a defeat for Romney, how badly depending on how far below he falls, regardless of how the other candidates shake out.

The key number is how Romney gets to 50 percent-plus-one in actual delegates, in a system where proportional representation makes this much harder than news cycle analysts seem to realize.

And do not believe anyone who tells you anything is decided before Super Tuesday, at the earliest, which in my humble opinion is mathematically impossible.

More in Presidential Campaign

Bush’s New Hampshire problem

Read more »