Republican momentum against Democratic maths
Ahead of the first presidential debate in Denver, I told Al Jazeera's global audience that for Mitt Romney the three Ms were very important; message, momentum and money.
My argument was that Mitt would try desperately hard to get his message across to the watching millions, hope that would create some momentum for his stuttering campaign, and in turn attract the donors he needed to finance the last big push to election day.
No-one outside of the Romney campaign expected such a turnaround from that first meeting. No-one expected President Barack Obama to be so bad, or for Mitt Romney to energise the Republican base quite so dramatically.
The day after the debate I went to a Romney/Ryan rally in Fishersville, Virginia. The traffic to the event tailed back for kilometre after kilometre, hundreds more people turned out than were expected. Suddenly people were positive about their candidate. Romney was no longer the anti-Obama vote. People actually wanted to support him.
From that first debate the face of the race changed. Romney started to pull back the huge gap he had in the national polls, and in some cases, even pushed ahead.
The story changed from the problems his campaign had through September to a momentum story (Mittmentum one newspaper called it in a terribly tortuous use of language).
And so with just over a week to election day, the Romney campaign is hammering momentum, the 'big mo', the continuing surge in the polls.
Some Democrats are worried and wavering, thinking an election they considered in the bag just six weeks ago might actually be lost.
However, the problem the Republicans have is that this election is more likely to be decided by maths than momentum.
Most of the polls are national but US elections are not decided on the popular vote, but on the electoral college and the individual state races. As I've written before, Obama simply has more ways to the coveted figure of 270 electoral college votes than his Republican rival.
No Republican has ever won the White House without winning Ohio. In 14 of the last 16 polls, Obama has been ahead and tied in the other two.
A top Republican political operative in Nevada said last week that Romney is likely to lose the state. In Iowa, another key swing state, early voting has attracted more female and young voters breaking, according to some estimates, two-one in favour of Obama.
A win in those three states plus the traditional Democratic strongholds, and Obama's your president. If that wasn't enough two recent polls in the fiercely fought state of Virginia shows Obama just slightly ahead.
The Romney campaign says their rivals are obsessed with numbers and are ignoring the realities of a surging Republican vote in Michigan and huge inroads being made in safe Democrat areas, such as Pennsylvania.
It has adopted the theme it used so much during the primary process – that Mitt Romney's success is inevitable. Given the national polls, they believe it is a compelling argument.
Obama's team do obsess over numbers.
They did it when competing against Hilary Clinton for the Democratic nomination in 2008 primary, insisting the data was on their side and the nomination would be won. They were right.
They are also widely accepted as possessing a great "ground game", the ability to get people registered and out to the polls. That will be hugely significant in the final result. The Republicans have improved, but not as much as they would have liked.
Hurricane Sandy – which is currently smashing into the East coast – may yet change the face of this election.
But in eight days we'll know if maths or momentum was more important.