As polling details from the Sunday Telegraph poll are placed online by ICM, it shows that the filter mechanism that they use actually lowered the real SNP vote, and increased the Labour vote.
Of the 516 constituents they asked only 303 were willing to say who they would vote for.
Of the 303 who indicated their voting intentions:-
131 would vote Labour.
124 would vote SNP.
28 would vote Conservative.
10 would vote Liberal Democrat
10 would vote for Others
In percentage terms that breaks down as:-
Labour 43.2 %
SNP 40.9 %
Liberal Democrats 3.3%
In other words there is a 2.3 % difference between Labour and the SNP.
So the SNP now just require a swing of over 1.15% to take the seat!
The reason why ICM released the doctored figures to the Sunday Telegraph is to do with the filtering mathematics they use on the poll.
For instance, as it has been a very safe seat in the past ICM have gave Labour the majority of the don’t know vote.
This implies that the don’t know vote will come out again in favour of Labour.
But as I quoted the Labour Campaigner yesterday, the Labour vote is very soft and reluctant. There is no justification for assuming that they will turnout and vote Labour.
The ICM’s adjustments may well be valid for General elections but I think these type of adjustments are misplaced for by-elections.
This doctoring of figures is ICM’s usual technique so we can assume there has been no political bias in their adjustments.
However it does serve as a cautionary tale for those willing to place great store in opinion polls.
Nonetheless the true figures provide a great lift to the SNP campaign and are a warning shot to Labour.
I expect this by election to be very close indeed!