As we've discussed a number of times now, the main reason the SNP fell just short of a second overall majority is that their vote went up on the constituency ballot, but down on the list ballot - and ultimately it's the list that decides the overall composition of parliament. But the SNP weren't the only party to suffer disproportionately on the list . It shouldn't be forgotten that Labour actually narrowly defeated the Tories in the popular vote on the constituency ballot, before taking a big hit on the list.
Constituency ballot :
Regional list ballot :
In past elections, we might have assumed that Labour's 3.5% drop from one ballot to the other could be explained by voters drifting off to other left-of-centre parties on the list. But that doesn't fully make sense in this case, because support for the two radical socialist parties was negligible, and the Green vote was only up by 2.2% - much of which can presumably already be accounted for by the drop in the SNP's list support.
On the face of it, the most logical alternative explanation is that a significant number of Labour constituency voters switched to the Tories on the list. That would certainly help to explain why the Tories' list support was 1% higher on the list than on the constituency ballot - a counterintuitive outcome when you consider that they must have lost some votes on the list to UKIP, and other assorted fringe right-wing nutjob outfits. This doesn't necessarily mean that Labour constituency voters misunderstood the system and gave their more important vote to their second preference party - it could be that these were people who preferred the Tories anyway, but were giving a tactical constituency vote to Labour in the vast swathes of Scotland where the Tories and Lib Dems are also-rans.
The huge Labour-to-Tory swings in many SNP/Labour battleground seats may superficially appear to disprove that theory, but that's not really the case - some of the tactical voting for Labour may have been going on under the radar, meaning that the swing would have been even bigger without it.
Incidentally, we've heard quite a bit in recent days about "the roar of Middle Scotland". Assuming that the representatives of "Middle Scotland" are the Tories, Labour and the Liberal Democrats, it should be noted that those three parties in combination won 60 seats - which is just three more than the 57 they won five years ago. Nobody would deny that represents a modest recovery from an all-time low, but a "roar"? Hmmm. Maybe more of a meow.