For the pundits, who arrive at their conclusions from a mix of polling trends, political knowledge and experience, on-the-ground intel and, in some cases, ideological bias, the race has been up in the air since Romney rumbled Obama in the first debate in Denver on October 4. They have been considering questions such as who has the better ground organisation for getting people to the polls, early voting, crowd numbers at rallies, which party is more enthused, which way independents will fall, will the youth vote hold up, will the gender gap be maintained and has Sandy sealed the deal for the President?
With polling data consistently showing Obama with a steady edge across the battleground states, some of the more objective pundits have accepted that he goes in as favourite.
They are largely relying on the chilled dealers in data, such as FiveThirtyEight's Nate Silver and Princeton Election Consortium's Sam Wang, Drew Linzer's Votamatic, Huffington Post Pollster, TPM Poll Tracker, FHQ, and aggregate sites such as RealClearPolitics. Betting sites Intrade and Betfair have Obama pegged as the favourite as well.
It's not just about which approach is more accurate and reliable.
Conflict between conservative strategists and pollsters became part of the campaign story when Romney took a slight lead in national polling and gained traction in Florida, Colorado, Virginia and New Hampshire.
The idea that Romney had maintained momentum after the first debate, that he was confidently surging towards victory, became part of the Republican campaign's strategy. It was dubbed "Mittmentum".
Silver, the most high-profile of the pollsters, became a target for some right-wing commentators - one called him effeminate - for not buying into the 'Big Mo' theory and becoming a soothing balm for scared Democrats.
Dr Tim Stanley, a conservative writer on US politics for the Daily Telegraph, accused Silver of being partisan and picking data that favoured Obama. Stanley claimed the Democrats talked up polling "to create a sense of inevitability, to convince the public to vote for Obama because he's a winner".
He added: "We've witnessed the evolution of polling from an objective gauge of the public mood to a propaganda tool: partisan and inaccurate."
Former Republican Congressman Joe Scarborough told MSNBC, after mentioning Silver's forecast, that: "Anybody that thinks that this race is anything but a toss-up right now ... they're jokes."
As various bloggers have pointed out, Silver's forecasts have been fairly conservative compared to other prediction models - and they all point to the same result.
The Republican sales pitch about momentum to party activists and voters has been maintained by the Romney campaign, although at least a couple of prominent Republicans, Bush strategist Karl Rove and former Mississippi Governor Haley Barbour, have suggested that Superstorm Sandy swamped Mittmentum.
"Obama has temporarily been a bipartisan figure ... He has been the comforter-in-chief and that helps," Rove told the Washington Post last weekend.
"If you hadn't had the storm, there would have been more of a chance for the Romney campaign to talk about the deficit, the debt, the economy. There was a stutter in the campaign."
Such claims are perhaps in preparation for a battle over the narrative should Romney lose.
But before Sandy hit, Silver pricked the momentum balloon by pointing to polling data showing that Romney's bounce actually ended in about mid-October. Obama has gradually made up lost ground in New Hampshire, Colorado and Virginia.
Polling data suggests that Obama's response to Sandy has crucially helped him with independent voters. But Silver's point is that Obama was making progress anyway and other factors such as an improving economy and debate wins could have helped.
Silver's view is that Obama has had a basic edge of about 2 per cent over Romney all year - not counting the bounces - and that national and state polls are now aligning.
On Monday Silver wrote on Twitter: "Obama's win probability peaked at 87 per cent on [October 5] last day before polls reflected Denver debate. Fell to 61 per cent on [October 13]. Now fully recovered. Obama unlikely to win by anything like his post [Democratic National Convention] margins. But Romney has no momentum, Obama's state polling is robust and time is up."
But could the vast polling data simply be wrong and is it really the best way of predicting the election?
Stanley says: "A problem with a lot of polling to date is that it has either over-sampled Democrats or used models of turnout that presume Obama's supporters are as enthused in 2012 as they were in 2008."
New York Times columnist David Brooks wrote: "Obama turns in a bad debate performance. Romney makes offensive comments at a fundraiser. These unquantifiable events change the trajectories of tight campaigns. You can't tell what's about to happen. You certainly can't tell how 100 million people are going to process what's about to happen. You can't calculate odds that capture unknown reactions to unknown events."
Republican consultant Alex Castellanos wrote of a "reticent Republican factor" - that the polls were underestimating Republican support because of a certain reluctance of some people to identify as Republicans.
He said that late polls in 1980 gave Ronald Reagan a 2-3 per cent lead over Jimmy Carter. But Reagan won by nearly 10 per cent.
Then again, Democrats have argued that the cellphone-toting youth vote - which favours Obama - is under-represented in polling.
Nate Cohn of the New Republic summed it up: "The polls are quite consistent and clear in the battleground states ... With Obama above 49 per cent in Nevada, Wisconsin, and Ohio, a wave of undecided voters can't flip the outcome. At this point, the polls must be wrong for Romney to prevail. The polls have been wrong before ..."
Polling data is really the only way to measure the impact of changes in the race as it is under way rather than relying on the wisdom of pundits. They add colour and interest to the campaign but essentially they are just expressing opinion.
For instance, there are major assumptions being made by pundits about turnout - favouring Republicans. Yet when you consider that Republican turnout for John McCain was 45.7 per cent four years ago and Romney would need another 5 per cent on top to win, you realise it won't be an easy task.
All the answers will very soon become clear.
