Polls aplenty over the weekend show the SNP clearly out in front as we enter the last three days. No-one actually believes their accuracy, which is further questioned by reading details of today’s article in the Herald, which claims as many as 46% have yet to make up their mind. Here in East Lothian, the numbers are not as large but still hover around the 30%. Given a voter roll of 55,000, and turnout around 60%, that would mean 10,000 voters who do intend to vote are still unsure.

That is only one of several curious statistics about this election. That the SNP voters seem mainly solid is encouraging and, while there is clear Tory bedrock in certain areas, many others are being persuaded by the SNP’s strong showing of a unique chance to vote SNP to end Labour’s unimaginative carping. What we can’t seem to find is a Lib-Dem vote. It’s similar to Scots Tories in the nineties who were too embarrassed to admit it; such Cleggophiles as remain have gone underground.

We have made two analyses of the likely result that agree to a remarkable extent. From the thousands of voters we have contacted, we extrapolate the column ‘by Canvass’ shown in the table below. Taking the 2007 results, adjusting for boundary changes and comparing this weekend’s SoS poll national share of vote with 2007, we come up with the column ‘by Polls’. Given a typical turnout of 60%, that means the SNP are a nose ahead of Labour here, with both parties around 11,500 votes and the other two nowhere. So the 10,000 still-undecideds will determine this—unless they stay at home in such numbers that we have a lower turnout and the cliffhanger predicted by the table. I hope they seize this opportunity to make history.