You wouldn't bet against Alex Salmond.

The First Minister understands odds better than the rest of us, and has made a habit of beating them. Unlike most of us, he enjoys a gamble. His definition of a long shot is not the common definition. Whether he can win the biggest prize is therefore, appropriately, a big question.

As Mr Salmond knows, there are three ways to handle a game of chance. In one version, you can play the percentages, risking a little and gaining a little, succeeding, with luck, by tiny increments. For most of the SNP leader's contemporaries, this is otherwise known as politics.

It was also the organising principle of Scottish Nationalism for decades. You might not get rich, electorally speaking, but you wouldn't lose your shirt with the tactic. Year by year, said this version of game theory, the odds would turn in your favour. One fine day, all your little bets would become one big pile of winnings.

Another way to play is to risk everything on one fall of the cards. No professional gambler would ever advise this as a method to keep a fool and his money from parting, but no poker addict ever had to draw to the SNP's hand. The reckless bet depends on the belief that there is nothing to lose. Where independence is concerned that is, in essence, the deal.

The third method is more interesting. It had less to do with cards than with Mr Salmond's fascination with horse racing. In that game, betting on the favourite pays off less often than you might think. If following form was the only thing that mattered, we could all get rich. The smart gambler looks for the horse that the rest of the punters have failed to take seriously.

As things stand, all the tipsters say Mr Salmond is risking everything – career, reputation, even (he might have been joking) his house – on a broken-down nag of a proposition. He might have a small side-bet on devo max and the like, but in essence the First Minister is betting the lot on the runner the pundits say is a faller, time after time, at every fence.

Salmond's Folly hasn't been going too well, say observers in the stands. At each time of asking, it refuses the jump in opinion polls. It shies away from issues involving defence, Europe, pensions, the currency. It refuses to put on a burst of speed no matter how often it is goaded. Those with their money on other runners doubt that it will stay the course.

It therefore becomes a good question. Why is Mr Salmond so sure that he has made the right selection amid all the available choices of horses for courses? He has settled on a particular SNP programme. He sticks with it (more or less) despite consistently bad news in the opinion polls, and despite weekly assaults on this or that policy detail. An instinct to gamble is one thing. What has become mysterious is the basis upon which the grand bet has been made.

Where a currency is concerned, a succession of old-school Nationalists have made perfectly fair points. There can be no independence, in any real sense, if it depends on the non-existent good faith of a London Treasury. There can be no independence, by definition, if you do not control your own economic affairs. From a purely economic standpoint it might be stupid, in any case, to depend on a City of London busy preparing its next disaster.

Mr Salmond doesn't want to know. He has Nobel Prize-winning economists he can cite – he still takes their trade seriously, too – and a good case involving England's self-interest to make. Clearly, the First Minister has a plan he believes will appeal to the community of Scotland in September 2014 and he means to stick to it. The source of this confidence is hard to find, however. The fact that the strategy is failing is obvious.

One story is told by the opinion polls. Conventionally, the theory goes that the SNP has shifted its position on issues such as Nato and sterling in order to "reassure" voters. No doubt that was the plan. But where is the evidence, the single shred of evidence, that this tactic has paid any dividends whatever?

The run of polls says, instead, that the Yes campaign would have been no worse off if the patience of Nationalists had not been tested over Nato, sterling and the rest.

Mr Salmond's gradualism has convinced no-one who was not already convinced. No-one has been won to the idea of independence because of a nuclear alliance or some fiddling around with the theory of a currency union. The 30% or so who mean to vote Yes did not acquire their conviction, meanwhile, thanks to Mr Salmond's watered-down nationalism. That's what they tolerate, not what they embrace.

The First Minister's bet is that he can bridge the gulf between two different views of the world. If he's right, it won't happen because of any story he can tell about EU cross-Border pension rules. But his gradualism, his compromises with what he takes to be political reality, have trapped him in an endless round of skirmishes over policy details. Every detail matters, but not a single one, of itself, will persuade a voter to restore a country's independence.

There are other bets available, of course. One attracting good odds says that voters will be so disgusted with the reality of the Coalition Government by 2014 they will take any escape route offered. I'm less than sure about that. In miserably uncertain times, no one votes for more uncertainty. They vote for ideas and for a belief in ideas. The spectacle of a First Minister rounding up the stats on the reality of a currency union does not do a lot for belief.

Mr Salmond believes he understands the voters. The belief lies at the heart of his bet. He is attending to their concerns, as he sees it, one by one. He is asserting that opinion polls do not reflect the opinions of the people and that, properly reassured, they will follow the path he has marked. His record in these matters has earned him confidence. But confidence alone guarantees nothing.

Most gamblers know that the last bet has nothing to do with the next bet. Mr Salmond and his party won a parliamentary majority under conditions that were supposed to be impossible. It would be naive to think that they could not perform a similar trick in the referendum, or in another Scottish general election. It would be daft, nevertheless, to imagine that rehearsals for 2014 are going well.

People, a majority of them, don't believe. They're not buying it. Part of the reason, I suspect, if that when they hear Mr Salmond involved in fiddling arguments over currency arrangements they conclude that he doesn't quite believe in independence himself. When everything is qualified, conditional, consensual and negotiable, the absence of faith becomes obvious. The debate begins to sound like a lot of fuss over nothing much.

For the 30%, with their ingrained beliefs, it sounds like an argument that is beside the point. Perhaps Mr Salmond should take a bet on their instincts. As things stand, he has nothing else to lose.