In an election this close, with so many competing, partisan voices battling to be heard on the subject, it's hard to get a sense of what is most likely to happen in just over a week.
The best approach is to take the opinions of pundits with a sackful of salt, plug out the noise and look at what the data are saying.
The figures suggest that both sides are competitive but that President Barack Obama has an edge.
As of yesterday, his rival, Mitt Romney, was ahead in the RealClearPolitics.com average of national polls by 47.9 per cent to 46.9 per cent.
But the election is decided by the electoral college, with attention focused on a clutch of states which could vote either way.
Of the closest 11 states, Obama leads by a 2 per cent poll average or more in six of them.
He has a 1.4 per cent lead in one - New Hampshire - and two states - Colorado and Virginia - are tied. Romney leads by 2 per cent or more in North Carolina. He also leads by 1.8 per cent in Florida.
If the situation stays static, Obama would cross the 270 threshold to win even without considering the two tied states.
Two per cent doesn't sound like much but it is this close to the election with no set-piece events remaining - apart from next weekend's jobs report - to influence the race one way or the other.
The New York Times' FiveThirtyEight blog looked at the historical reliability of presidential state average poll leads 10 days out from an election.
From 1.9 per cent and higher, the polls were very reliable in indicating the winner. Of 30 polls in this category, only one was wrong. Under 1.9 per cent, anything goes. Of 14 polls in this category, eight states went to a different winner. That suggests that the states really in play at this late stage are Colorado, Virginia, New Hampshire and Florida. Obama can afford to lose all four and still win.
Looking at the poll trends on RCP over the past month, what has happened is that Romney initially gained a good bounce from the first debate on October 4. But a combination of Romney failing to kick-on and Obama coming back meant that the bounce stalled and subsided.
Snap polls showed that the public believed the Republicans won the first debate but the Democrats took the next three.
Romney essentially battered Obama's outer wall of Florida, Virginia, Colorado and New Hampshire without breaking through to the Democrat's inner keep of Ohio, Wisconsin, Michigan, Iowa and Nevada.
Over the past two weeks New Hampshire has crept back to Obama's side and Romney grip's has lessened on Virginia and Colorado.
But the figures aren't all good for Obama. Polls have shown a decline in his share of the white vote and a closing of his lead with women voters. His favourability ratings have also been sliding. The Republicans are highly motivated. And Romney does have that narrow lead in the national vote.
For Obama, there's no room for error.
Organisation on the ground will be important to get out the vote and there the Obama team's experience from four years ago could be crucial. According to the Atlantic Obama has more than 800 offices to about 300 for Romney.
The new GDP growth figure of 2 per cent is a reminder of the wider issues at stake here. Obama, having been gifted a recession and two wars by his Republican predecessor, could well be turfed out just as the economy picks up only to see his Republican successor get the political credit for the recovery.
State of the race
RCP national average
Barack Obama 46.9%, Mitt Romney 47.9%
Obama 49.7% for, 45.2% against
Romney 49.3% for, 42.8% against
Obama 63.5, Romney 36.6
Electoral College 270 to win
Obama 201 (290 no toss-ups)
Romney 191 (248 no toss ups)
Obama leads: Nevada, Ohio, Wisconsin, Pennsylvania, Michigan, New Hampshire, Iowa
Romney leads: Florida, North Carolina
Tied: Virginia, Colorado