Bizarrely, the Sun have commissioned a YouGov poll with a sample size so enormous that the Scottish subsample is equivalent to a typical full-scale poll. The only remaining question is whether the Scottish figures are properly weighted - perhaps not, but nevertheless they must carry somewhat more credibility than most subsamples. Here are the full figures -
Labour 36% (-3)
SNP 25% (+2)
Conservatives 17% (-2)
Liberal Democrats 15% (-)
Others 8% (+4)
So Labour's lead over the SNP is cut from sixteen to eleven points, while the Nationalists consolidate their second place, which had briefly appeared threatened. Seemingly on the brink of power, the Tories are nevertheless going absolutely nowhere in Scotland, leaving very much open the 'Super Doomsday scenario' Peter Lynch set out on Jeff's blog a few weeks back. Meanwhile, I'd be intrigued to witness Iain Dale's logical contortions in trying to reconcile these figures - suggesting an eight-point drop in Liberal Democrat support since the 2005 election - with his prediction that the Lib Dems will gain four seats, and end up reaching a new high watermark of fifteen!
UPDATE : Just to confuse the issue, the fieldwork for this poll took place between the 5th and 7th of January, meaning it overlapped with - and to some extent even preceded - the previous YouGov subsample. So the 'percentage change' figures I give above are not really meaningful. It also means that some of the fieldwork took place before Labour were hit by Geoff Hoon and Patricia Hewitt's hopelessly bungled coup attempt.