ComRes subsample : OK, this one probably doesn't count

I'm normally fairly scathing about people who say, in Snow-esque fashion, that the Scottish subsamples of UK-wide polls are "just a bit of fun". My normal reply is that they certainly need to treated with considerable caution, but if they really were totally meaningless the figures and placings wouldn't be as relatively stable as they generally are. But the latest subsample from ComRes seems to be the exception that proves the rule. Not much comment is really needed on these figures -

Labour 50% (+11)
Liberal Democrats 19% (+4)
SNP 18% (-3)
Conservatives 9% (-11)
Others 4% (-1)

Of course, the real story from this poll is the narrowing of the gap between the Tories and Labour at UK-wide level to well within hung parliament territory. What's fascinating is the way that a number of Tory supporters are pretending this is somehow a good thing for their party, because the closeness of the race will "concentrate minds". It's a handy truism that the best way to judge if a statement has any credibility is to turn it on its head and see if it intuitively makes much sense. In this instance that would mean Labour trying to work out what would be the most effective way of getting back into the game - and concluding that what they really need is for the Tories to maintain massive poll leads right up to election day. Yep, that makes perfect sense.

Bad night for the Tories.