Last Monday, as the results of the Quebec election unfolded, they made the predictions I had laboriously concocted the night before look like amateur work. I was, I admit, a little befuddled.
After all, as I try to make sense of this brave new world of Quebec politics at the dawn of the Mario Dumont period, I start with multiple handicaps: I live in Montreal and most of us on this increasingly insular island seem to have missed the boat of this campaign.
As a political scientist I run the risk of being perceived as an "intellectual," which makes me inherently suspect; as a citizen I tend to have firm opinions, which makes me an untypical Quebecer. With hindsight, however, I can see the picture starting to emerge more distinctly ... and it's clearly fuzzy.
Since my fancy political science concepts didn't help me much in predicting the results, I'll base my analysis on a memorable line from my favourite political comedy, L'aventure c'est l'aventure, a hilarious 1970s French classic: "Politics is clarity in confusion . . . and confusion in clarity."
For example, when Stephen Harper confronted a dithering Paul Martin, who had more priorities than one could count, Harper replied with clarity: five priorities, easy to understand and achievable.
Since then, he seems to have reverted to the other mode, countering Stéphane Dion's direct style with a masterful use osf ambiguity and confusion that serves him well in the naturally confused world of a minority Parliament.
In the case of Dumont, it's important to note that his political guru when he entered politics as a young Liberal was Robert Bourassa, who took the future ADQ leader under his wing.
In Quebec politics, if there ever was a master in the use of ambiguity to generate confusion when things seemed too clear, or to make an inherently confused policy position seem clear, that person was Bourassa.
The late Liberal premier liked his young protege, and he knew that Dumont was destined to move up. But Dumont was too impatient on the constitutional front, and he didn't seem to understand that Bourassa never believed in the ideal constitutional schemes peddled by his own party but rather used them to attract soft nationalists and gain time.
Bourassa's lessons took some time to sink in.
When Dumont joined the Yes side in 1995, he later admitted he did not support sovereignty, but believed that post-victory negotiations would ultimately lead to his fantasy vision of a Canadian federation stretched to its ultimate limits.
Twelve years later, I doubt he still believes that this fantasy can ever materialize. But the master's lessons seem to have sunk in, finally.
Dumont knows that, in a political landscape defined by two starkly contrasted options – Jean Charest's passionate federalism vs. the PQ's commitment to make Quebec a sovereign country – the pool of nationalist voters that can be attracted by an ambiguous constitutional posture is vast.
Arguably, Dumont's constitutional position was just as ambiguous in 2003 as it is now and he already had mastered the art of dodging the question of what side of the Quebec political divide he stands on. He is neither a sovereignist nor a federalist ... bien au contraire!
Federalists claim that Quebecers who do not want a referendum reject it because they do not want sovereignty, while sovereignists argue that many wish to avoid a referendum because they don't want to lose it.
Dumont, for his part, knows that most of those who don't want a referendum reject it simply because they don't want to have to choose between sovereignty and federalism.
Since his ambition is to cast his net ever wider in that large pool of voters, he is going to continue to dodge that question.
Canadians in other provinces should rest assured, however: The new opposition leader is a federalist.
Indeed, whatever he says, the result of his actions can only be that Quebec will remain a province of Canada under the existing Constitution and the federation would change only to the extent that other governments would agree.
Just don't expect Dumont to say it. That would break the spell.
Why, then, did Dumont surge in 2007 and not in 2003? Of course, there was the unpopularity of Charest and Boisclair, but there's also the politics of clarity and confusion.
In 2003, Dumont knew he could win with a confused constitutional position. But he somehow lived with the illusion that he would have to balance it with a clear right-wing program, including things like school vouchers and a flat tax.
That was a mistake.
In 2007, the ADQ still is a party of the right but its policies are wrapped in populist language that makes them sound as though an ADQ government could be all things to all people, just like Dumont's constitutional position.
In the end, the distribution of the vote in Quebec makes sense. One-third of voters are centre-left sovereignists; one-third are centre-right federalists; one-third (and counting?) simply do not want to choose a side.
Dumont always has been an opportunist, and this election gave him what he craved: A golden opportunity to attract nearly all the votes from the latter group.
In the coming months, as Quebec's new opposition leader tries to carry on his momentum, we'll see how well he learned his master's craft.
Clearly, some of us might get confused.

Pierre Martin is a professor of political science at the Université de Montréal.