WE can’t say the warning signs weren’t there.

All of the polls over the closing days of the campaign were suggesting the SNP had slipped below 50 per cent of the constituency vote, and the final YouGov poll put them on "just" 48 per cent – their lowest since before the UK General Election a year ago, and only three per cent better than their performance in 2011.

But perhaps the greatest reason for the ongoing sense of complacency about an SNP majority was the temptation to use that 2011 constituency result as a benchmark for success. Consciously or unconsciously, people tended to assume that as long as the SNP remained north of 45 per cent, a second majority was bound to follow. Unfortunately, there isn’t really any specific target figure on the constituency ballot that can leave a party assured of breaking the system.

In spite of the gains they were always bound to make from Labour, the SNP faced one obvious problem in the constituencies this time around that they didn’t have to worry about five years ago. In heavily No-voting areas, especially where the Conservatives or the Liberal Democrats have a history of strength, the urgency of the constitutional issue was motivating Unionist voters to coalesce around whichever candidate was best-placed to defeat them.

Even if the SNP had racked up 50 per cent or more of the national vote, this factor alone could conceivably still have been enough to prevent them from reaching the formidable target of 65 out of 73 constituency seats needed to win an overall majority prior to any list seats being declared. And as soon as list votes were required, there was always going to be an element of uncertainty about the final outcome. Historically, there doesn’t seem to be much rhyme or reason to the size of the gap between a party’s vote on the constituency ballot and on the list ballot, and opinion polls don’t really offer reliable clues on that point either. In 2011, the polls had suggested that the SNP would lose many of its constituency voters on the list ballot, but in the end the success of the "Alex Salmond for First Minister" campaign saw the party transfer almost its entire constituency support to the list. As it happens, in this year’s election the polling average also overstated the gap between the SNP’s constituency and list support – but only very, very slightly.

Put simply, the reason the SNP lost their majority is that their list vote was a little down on 2011 – and if you fall significantly short of securing a majority on constituency seats alone, it’s the list percentages that will broadly decide the result. The overall composition of parliament is intended to be proportional to how people vote on the list, not on the constituency ballot. There was a lazy tendency among the presenters of the TV results programmes to say “Oh well, of course the SNP haven’t won any list seats in Region X, because they won too many constituency seats”. This reinforced one of the most frustrating myths about the Additional Member System – namely that list votes are guaranteed to be "wasted" if you’re voting for a party that sweeps the constituencies in the region. But any viewers with a memory of at least five years would have known that what they were being told didn’t quite add up – because the SNP were losing list seats in some regions where they either failed to make any constituency gains, or actually suffered constituency losses. The reality is that the party could have won at least one list seat in every region if the list votes had been there in sufficient numbers – but they simply weren’t.

How far the SNP fell short of the necessary votes varied sharply from region to region – in Glasgow they were well short of a list seat, and if a crystal ball had been available pro-independence voters in the city might have concluded that they were powerless to use their list votes to increase the chances of an SNP majority, and perhaps would have plumped for the Greens, who regrettably failed to claim the second seat that could realistically have been within their grasp in the region.

But the whole problem with list vote “strategies” is that the crystal ball is never available – a voter switching to the Greens in Glasgow might just as easily have been wasting their vote on a party that already had enough support for a second list seat, but that had no chance at all of a third after the d’Hondt formula divided its vote by three. In any case, the arguments from supporters of smaller parties that the real problem in this election was wasted SNP list votes can’t possibly make sense, because if the objective was merely a pro-independence majority in parliament, that objective has been comfortably achieved. What we’re mourning is the lack of a single-party SNP majority – and the only possible way of achieving that would have been through a greater number of SNP votes.

We can only speculate about how much of the crucial drop in the SNP’s list vote can be explained by a small number of voters taking heed of the misleading claims that a majority was guaranteed and that the list vote could therefore be safely treated as a ‘bonus’ or "luxury" vote. The Greens attract support from many corners of the political spectrum, including unionists, so it’s obviously not the case that their three per cent increase in support came entirely from tactical voters. Unfortunately, though, the SNP were so close to a majority that we’ll always wonder if any small factor may have swung the balance. One of the few things we know for almost certain is that the SNP would have won Edinburgh Central had it not been for the rare Green intervention in that constituency – and that would have left them requiring just a tantalising one extra list seat.

Whatever the consequences of the SNP’s setback for the cause of independence in the short-to-medium term, I hope the lesson we draw from this experience is that the proportional section of any proportional voting system should be treated with the respect it deserves.

That means treating the list vote as the more important vote, and not as second preference. There’s no way of ensuring that we get the exact election result we want as individuals, but as long as the system is used as intended, we can at least be reasonably confident of getting the result that was collectively desired. The debate over whether that happened in the election of 2016 may well rage for years.

James Kelly is a blogger who runs scotgoespop.blogspot.com

Nicola Sturgeon vows to keep making the case for independence after failing to win an SNP majority