Quentin Langley is a PR and social media strategist; author, academic, speaker & entrepreneur.

Alex Salmond thought he had it all his own way: his question; his choice of electorate; no super-majority or turnout requirement; no second vote. He still lost decisively, if not as overwhelmingly as once seemed likely. But could he have played his hand differently? And what are the implications for a referendum on Europe?

I have argued on these pages – these pixels? – before that referendums are special, at least in those countries where they are rare. Holding a referendum communicates that this is a once in a generation choice, at least. It is not like choosing a new government, a choice which can be reversed in four or five years if it doesn’t work out. The implication of this is that last minute deciders swing sharply for the status quo.

Scots who were not convinced of the case for independence opted for the safe course and voted ‘nay’. The same happened in the first Scottish devolution referendum, both polls on Welsh devolution and the EEC and AV referenda. The same pattern can be seen in referendums in France, Ireland, Holland and Denmark – though not in California or Switzerland, where such votes are routine rather than rare, and not reserved for major constitutional changes.

If you are unsure on a question – genuinely undecided – after two years of debate, it makes sense to opt for the cautious path. That’s human nature. Evidence from behavioural economics shows that people are very keen to hold on to what they have, and don’t like risking it for speculative gains. Both sides played to these fears, warning of the dire consequences of the wrong vote. Both sides greatly exaggerated, and got away with it partly because the Scottish Executive wanted an early vote. It did not want to hammer out the details of asset splits and international arrangements in advance of the poll.

It is, of course, concerning to see ruptures of this magnitude settled by a vote which could have been much closer and which had no turnout requirement. With a 60 per cent turnout and a majority of one the Scots could have voted to end the Union, and done so with many critical questions unanswered.

For so grave a change, a more considered approach would seem wise. But a super-majority requirement potentially leaves the winners of the vote dissatisfied. A turnout requirement encourages opponents of change to stay at home rather than vote ‘no’. (This often happens in Italy, where a 50 per cent ‘quorum’ is required in referenda). The best option to secure a considered result and to get round the question of whether to hold the vote before or after the details have been worked out is to hold two votes: the first asking if people, in principle, would like to see Scotland become independent, and the second, a couple of years later, presenting the details of the divorce and asking Scots if they want to proceed on those terms.

At first glance, such an approach would seem to disadvantage the ‘yes’ side. After all, they would have to win both twice, whereas Better Together would only have had to win once. And, yet, I suspect if Salmond had gone for this approach he would have had a better chance of winning.

The first vote would have been an easier one for him. There would probably not have been the surge in turnout with shy unionists avoiding pollsters but rushing to the polling stations. And those cautious undecided people? Well, if you are unsure it makes sense to find out more information rather than to close down the debate. If yesterday’s vote had been only in principle, with another vote to confirm any change, the rational waverer would have voted ‘aye’ to keep the debate going, knowing that she or he could always say ‘nay’ later on, if the terms did not look good.

If the question had been “In principle, do you think Scotland should become an independent country after the details have been worked out?” the Yes campaign would probably have won.

Of course, this would have left Salmond facing another vote in a few years time. This one would be the big one: the scary one. This would be the vote where Better Together would have its best shot of winning. But it would be a harder campaign to win than Thursday’s.

Both sides engaged in fear-mongering. Both claimed that the NHS was at stake. What of the relationship with the EU? What about the pound? How much of the national debt would Scotland inherit? And if it, as a successor state to the old UK, inherits the debt, why not the responsibilities, such as EU membership, too? For all the bluster on both sides, much of this remains unclear.

At the time of a second vote, the situation would be clearer. While the best hopes of the SNP would not have been realised in these negotiations, the worst fears would also have been taken off the table. Independence would still have been a leap in the dark, surrounded by much uncertainty, but not as much uncertainty as when the vote is held before the terms of the break up are agreed.

I am not predicting that Salmond would be in a position to win that second vote, but his chances would be fairly good. With the scariest of the potential consequences removed – let’s face it, talk of border guards was always fatuous – and with a vote in principle already won, he would be facing a realistic prospect of winning the second vote too.

And, yes, these lessons are generalisable. There is no doubt that if David Cameron were to say that instead of holding a referendum on EU membership he would now hold two, Eureseceptics would be furious. Let’s say he proposed a vote in principle in 2016 and a ratification of the terms of departure in 2018, if the first vote went to the Out camp. The Better Off Out camp would certainly see this as betrayal. And yet, if you are proposing major constitutional and economic change, there is a case for saying that you have a better chance of winning one vote than you do of winning two.

Just as in the Scottish debate, both sides will be making predictions in 2017. One side will claim that we will never be able to do business with Europe again. We will need to apply for visas to go on holiday, with nightmarish bureaucratic delays. The border with Ireland will be shut. The other will be claiming that we will slip easily into a free-trade only agreement such as Norway and Switzerland have. Neither would be wholly correct. Any Brexit will not be as bad as Europhiles claim, but not as easy as its advocates suggest either. A second vote, after the details of the departure have been agreed, would not be subject to as much scaremongering.

I have no doubt that any referendum in 2017 will resolve to stay in. That potentially settles the matter for decades to come. But Britain would probably vote, in principle, to withdraw, provided there was the reassurance of a further vote later on. And the second vote? The result of that would probably depend on the terms of exit. So I am not ready to call that one yet.