At the start of the week Lord Ashcroft published his latest batch of polls, this time from Scotland. Perhaps the biggest shocker was the 29 point deficit to the SNP faced by LibDem Danny Alexander, in his seat in Inverness, Nairn, Badenoch and Strathspey.
Today it’s the LibDem leader’s turn. Lord Ashcroft announced today that it had come to light that the poll he had published for Sheffield Hallam included a mistake in the data. The corrected data means that rather than having a three-point lead Nick Clegg should have been three points behind Labour.
The problem is that this poll was done in November, so whether the data was right or wrong, it’s completely out of date. Lord Ashcroft is an erstwhile activist so as I posted earlier in the week there are good reasons for treating his announcements with caution. On the 27 November the day that Ashcroft first announced TNS BMRB put the LibDems on 6% of the national vote, Today a YouGov poll puts them on 9%.
In search of a headline however The Telegraph, The Independent and Channel Four News all carry the line that Nick Clegg is on course to lose his seat. Perhaps he is, it looks as though he was, but a November poll is no basis for a February headline.
Ballot Blogs 
I think your premise is flawed. You can’t cherry pick polls – the Lib Dem polling average in late November was 8%. It is still 8%.
You are absolutely right, but it isn’t a premise. It just shows that polls vary nationally and locally both in accuracy and over time. We have no way of knowing that the 8% who currently say they’ll bote LibDem are the same 8% as in November. The premise here is that it is completely wrong to correct data from November and suggest that it is now valid as an indication of current preference. All of that said I think it is quite possible that Nick Clegg will lose in Sheffield Hallam.