You might remember that after the first Salmond v Darling debate in advance of the referendum, there was just one instant reaction poll with a very small sample size, which showed by a relatively narrow margin that the viewers thought Darling had been the better performer. The entire narrative of a Darling victory was founded solely on that one poll, in spite of its limitations. I wondered at the time what would have happened if the journalists' desired narrative had been frustrated by the poll indicating a narrow Salmond win - which could easily have been the case, given that only a small number of respondents were required to swing the balance in either direction.
We're seeing a scenario of that sort play out tonight, because the media would probably quite like to award a victory to Ed Miliband, not because they support Labour, but because the underdog coming out on top is the better story. However, the ICM instant reaction poll (which admittedly has a bigger sample this time) is stubbornly refusing to give them permission to do so, with Cameron being declared the winner by 54% to 46%. It'll be interesting to see what angle is taken tomorrow - I suppose there's still scope to spin it as the hopeless Miliband exceeding expectations, and almost nicking a plucky draw.
For what it's worth, I do actually think that Miliband came out on top, and that reaffirms the overwhelming importance of us having fought so hard for SNP inclusion in the real leaders' debate next week. There's always a chance that your main opponent might do better than expected, and if you're not even in the room to counteract that, you've got a major problem.
SCOT GOES POP BRICK INDEX
David Cameron 98
Ed Miliband 77
(The Scot Goes Pop Brick Index indicates how much out of a score of 100 I wanted to throw a brick at the TV while the leader in question was speaking. Lower scores are better, therefore Miliband was the "winner".)