Never in the field of Canadian elections have polls meant so little to so many.
A governing party trailing in second place, maybe even third, and with its leader posting dismal approval ratings, would normally set off alarm bells.
But this is Alberta, and old habits die hard.
After 44 years in power and 12 consecutive election victories, anything but a win by Jim Prentice's Progressive Conservatives is dismissed by many as nearly fantastical.
Even the pollsters are couching their analyses with warnings of the changes likely to come. The Tories may be in dire straits now, but the election is many weeks away. The inevitable will still, probably, come to pass — right?
But re-alignments can happen swiftly. Are the polls presaging a historic defeat for the Alberta PCs, or is there good reason to believe that what the numbers show now will not be the same on May 5, when Albertans cast their ballots?
The case for the polls
ThreeHundredEight.com's latest aggregation of the polls released Monday show the Wildrose Party holds the lead with about 30 per cent support, followed closely by the Progressive Conservatives at 28 per cent and the New Democrats at 26 per cent. The Alberta Liberals bring up the rear with 12 per cent support.
This outcome would likely deliver a minority government headed by either Wildrose or the Tories, with the edge narrowly going to Wildrose. It is too close to say anything more definitive.
If these results were repeated on election day, it would be the best showing for the NDP since 1989 and the worst for the PCs since 1967.
This alone should raise some red flags. But the polls reported in the media have been remarkably consistent. In the three polls conducted by different polling firms since the beginning of April, Wildrose has registered between 30 and 31 per cent, the PCs between 25 and 27 per cent, and the NDP between 26 and 28 per cent. What first looked like a fluke at the end of March has been confirmed again and again.
Other indicators back up these numbers.
Prentice's approval ratings have dropped precipitously. From more than 50 per cent in polls conducted in December, Prentice has registered an approval rating of between 22 and 29 per cent in two recent polls, with his disapproval rating topping out at 60 per cent or higher.
In 2012, when the PCs were running behind Wildrose, the polls never showed such low numbers for then-premier Alison Redford. Only in the weeks shortly before her resignation did her numbers turn so sour.
The budget seems to have been a catalyst for this decline. Polling by Insights West suggested that 78 per cent of Albertans thought the budget would have a negative effect on their households, and a majority did not think that the fall in oil prices justified it.
Fears of change amid an unstable economical climate also appear low. ThinkHQ found that just 30 per cent of Albertans agreed that a change of government would make things worse on that front.
There are also signs that the NDP's stellar climb into contention is no anomaly. NDP Leader Rachel Notley boasts, by a wide margin, the best approval ratings in Alberta. The most recent poll by Forum Research pegged her approval rating at 42 per cent, with just 21 per cent disapproving of her.
The case for another PC victory
The experience of the 2012 election, however, looms large over this campaign. Every poll published during the writ period suggested that Wildrose would win. Instead, the Tories were re-elected with a large majority of seats. A warranted skepticism has defined Albertans' views of the polls since.
But leaving aside the possibility that the polls are just wrong (and the record of the polls in 2012 is far more nuanced than that), there are good reasons to believe the results on election night will differ markedly from where the numbers are now.
The most important one may be the incumbency factor, which plays into the hands of every governing party. Unexpected victories by the incumbent have happened recently in provincial elections from British Columbia to Quebec. And in 2012, the Alberta PCs were trailing from the first week of the campaign.
The incumbency factor could play an even larger role in this campaign because of the relative obscurity of Brian Jean, Wildrose's recently named leader. The latest Forum poll showed that 47 per cent of Albertans had no opinion of him, leaving a lot of minds left to be made up. And considering that 44 per cent of respondents told ThinkHQ they found Wildrose "too extreme," those minds could be made up to Jean's detriment.
Notley, too, is still unknown to more than a third of Albertans.
And then there are the Liberals, who are on track to have candidates on the ballot in only half of the province's ridings. If that happens, they will certainly not get the 12 per cent the polls are currently giving them. Where will those orphaned voters go? To the PCs to block Wildrose, the reason for the Liberals' collapse in 2012, or to the NDP, the other left-of-centre option?
Added to all of this is the reality that a majority of Albertans say they have yet to be convinced to vote for any party. In ThinkHQ's most recent poll, fully 58 per cent of respondents said they could change their mind.
That makes for a lot of voters up for grabs, and the likelihood that the result on election night will resemble the polls today slim. But that doesn't necessarily mean Jim Prentice's PCs will end up on top.
The poll by Forum Research was conducted between April 7 and 9, 2015, interviewing 1,661 Albertans via interactive voice response. The margin of error associated with the poll is +/- 2 per cent, 19 times out of 20.
The poll by ThinkHQ was conducted between April 2 and 6, 2015, interviewing 1,835 Albertans via the Internet. As the poll was conducted online, a margin of error does not apply.
The poll by Insights West was conducted between March 27 and 30, 2015, interviewing 602 Albertans via the Internet. As the poll was conducted online, a margin of error does not apply.
ThreeHundredEight.com's vote projection model aggregates all publicly released polls, weighing them by sample size, date and the polling firm's accuracy record. The projections are subject to the margins of error of the opinion polls included in the model, as well as the unpredictable nature of politics at the riding level. The polls included in the model vary in size, date and method, and have not been individually verified by the CBC. You can read the full methodology here.