Biggus - well, you can view it like that, but I was applying psephology to it. All too-often, people read the headlines and make direct correlation between support and what will happen without looking at the 'hidden wiring' - the stuff which means that with the Tories and Labour each on 32%, Labour wins 70 more seats; the stuff which means that UKIP's possessing double the support there is for the Lib Dems translates to UKIP perhaps getting half the number of seats at most (see, Nick, losing the AV referendum wasn't all bad...)
First past the post makes trebling the number of seats difficult. Not impossible, but difficult, even allowing for the SNP's vastly-improving electoral fortunes over the years.
The polling average I cited takes all of this into account; this isn't an exact science by any means, but it tends to be a reasonably accurate indicator of outcomes, and has been for over half a century.
It's not a case of saying that the SNP won't be a significant factor, merely that historical trends in elections suggest that their chances are not as great as some suggest assuming that the current polling figures remain the same.
And there are variables even in this - first, is the local MP a popular MP? If they are, then you often see them keeping their seat. Second, how is the support distributed? If the surge in SNP support comes from Labour's potential vote collapsing in constituencies already held by the SNP - i.e. they were going to win the seat anyway - then while the headline figures are great, that translates into a smaller number of seats than the polls might suggest if you do a raw calculation. Third - how many of those now supporting the SNP will (a) turn out or (b) change their mind in the polling booth? Precedent - again, an imprecise and unreliable thing - suggests (no more than that) that (b) often happens. After the referendum, (a) would appear less of a factor.
This is why - despite the press reports - I'd rather not completely throw precedent and polling prediction (prediction based upon a significant reassessment of the minutiae of elections worked after 1992 made the pollsters look very silly) and offered the view that based on all of this 'it seems fairly unlikely' (as per my original)
I'm not airily writing the SNP off, merely attempting to suggest that confident perorations about what their demands to be coalition partners would be are perhaps a little misplaced at the moment and Ms Sturgeon needs to be working hard to ensure that the 47% support does translate into 40 seats once most/all the variables have played out.
And I am not making a prediction beyond pointing out what the reasonably sophisticated psephological approach now in use says even in the face of the SNP's remarkable surge in support.
I'd not claim to be an expert in this, but having taught British politics (to undergraduate historians rather than those doing the politicians' training course that PPE was becoming), I'm always wary of headlines...
Home was meant to lose by a street in 1964. Wilson got a majority of 4 (3 after the Patrick Gordon Walker fiasco cost him a seat)
Wilson was meant to win in 1970 by a reasonable margin. He lost.
Major was meant to lose in 1992 by some way. He got a majority of 21.
The press assumed that Cameron would have a substantial majority given Brown's unpopularity in 2010. And it didn't quite work like that, as we are experiencing.
On the flip side, Blair's majority in 1997 was greater than generally predicted - he was going to win comfortably, but the extent of victory was greater than the polling suggested.
Put in a less long-winded manner, perhaps I should've written - 'Analysis of election results since the late 1950s suggests that Ms Sturgeon is not quite as likely to be in a position to demand anything as a coalition partner as she seems to think'...
It could happen - which is why the process is so interesting - but I'm simply urging caution, that's all.