With the first primary for six weeks tomorrow, it's time for Freethink to return to the US of A and consider the implications of the various possible scenarios and make some predictions.

In a sense, we already know what will happen. Barack Obama will win in Pittsburgh and Philadelphia, and Hillary Clinton will win the vast majority of what is between. John Kerry managed to pull out a victory in 2004 by winning the two cities at either end of the state and nothing in between, but Obama will find it harder.

Although many polls are showing there are 5 or less points in it, I'm sceptical that it will be that close. I anticipate a Clinton win of between 8-12 points for a couple of reasons. 

Firstly, in recent polls Barack Obama has hit 45% on only two occasions, suggesting a consistent ceiling of support. Second, polls thus far have not added up to 100% because of the 10% or so of primary voters who remain undecided. Critically, those who make up their mind close to the election have tended to break for Clinton in the contests so far. When added to the fact that undecideds seem to be located disproportionately in the rural areas, this suggests that Clinton will get the bigger boost from current poll numbers than Obama will.

A result of between 5-10 points will allow Clinton to carry on, but should she lose in both North Carolina (which she will lose by a significant margin) and Indiana (where she has surrendered a double digit lead and which could go either way) the end will be nigh. A win by anything less than 5 points and her position is unviable. She may decide to fight on, but Obama will be the nominee. A win by over 10 points will allow her to fight on, but the delegate maths remains formidably against her. Given that Howard Dean is clearly keen to wrap things up well before the convention, a Clinton nomination still looks a fairly remote possibility.