Monday, 15 April 2019

Final 338 Alberta Projection: Available Data Points to UCP Majority

[edit: see here how the projection fared]

This is the final 338 Alberta Projection.

I'm tired.

But I have learned so much.

I am a Montrealer who has been to Alberta exactly five times in my life. I have visited Calgary, Edmonton, Banff, Jasper and Lake Louise. And I have stopped for gas in Red Deer.

That's not much, some would say. I know.

But what struck me most about the Albertans I met was how nice and friendly they were.

(No, those who write Facebook comments are not a representative sample of the population. Not in Alberta, not anywhere.)

It makes me want to go back once more. I am pretty sure I will. Soon.

I loved their local beer. And the landscape, my god, the landscape.

Here are the final 338 numbers for Alberta.

Popular Vote Projection


The United Conservative Party is on track to win the popular vote by 8 to 12 point margin. In the last polls published this afternoon, the gap between the UCP and the NDP has ranged from 6 points (Pollara) to 14 points (Léger). The weighted average for the UCP stands at 48.9%, a drop of two points since the start of the campaign.



Rachel's Notley's NDP ends this feisty campaign on a positive note, with an average of 38.5%, which is 6 points higher than when the writ was drawn up a month ago. If those numbers end up accurate, the NDP will suffer a somewhat "honorable loss" by capturing a share of the vote very close to its 2015 result.

The Alberta Party stands at an average of 7.2%.

Here are the popular vote projections with 95% confidence intervals:




Seat Projection


The United Conservative Party is on track to win a majority government. By breaking down the numbers per region and per riding, the UCP averages 60.2 seats per simulation performed by the 338 electoral model. This seat average still puts the UCP well above the majority threshold of 44 seats at the LAA.



The NDP's seat average stands at 26.1. Unsurprisingly, most of those seats will be in and around Edmonton. The NDP could win additional seats in Calgary if it gets its vote out tomorrow, but it probably won't be enough to beat the United conservatives

Here are the seat projection probability densities:






Odds of Winning the Most Seats


With these numbers, the UCP wins the most seats in about twenty-four of of twenty-five simulations (96.2%).



The NDP wins the most seats in only 3.2% of simulations, so the UCP is still, as of this morning, considered a strong favourite to win the election.

The UCP and NDP are tied for the most seats in 0.6% of simulations.


Find Your District


The 338 Alberta page has been updated. You can find your electoral district by clicking the following links:


Find the complete interactive map of this projection here.



In Conclusion


[Excerpt from the following was published by Maclean's earlier today.]

Opinion polls are scientific experiments. However, unlike the cases of stars or electrons, the subjects of these experiments are human beings who have varying moods and may hold contradicting opinions. After all, humans beings are far more complex and mysterious than stars or electrons, which makes the study of their behaviour all the more uncertain.

As a scientist whose career it is to teach the scientific method to young adults, I created the Qc125 and 338Canada models with the scientific method in mind. René Descartes' rules for acquiring knowledge may date back to almost four hundred years, but they still ring true today: don't judge without proper evidence, don't judge based on preconceived ideas, and don't let your judgement go beyond the evidence. While it is true that polls contain uncertainty, when conducted properly, polls also hold precious information about the mood of voters. As a rule, the 338 model sees a poll as but a tiny piece of a giant puzzle, and it is only by assembling these pieces together that we may fully understand the big picture.

And the data collected and made available about Albertan voters this spring is abundantly clear: Jason Kenney will most likely become the 18th Premier of Alberta. Moreover, since the likelihood of third parties winning more than one or two seats is slim, the odds of a majority government are overwhelming.

Does Rachel Notley have a path to victory?

The answer to that question is simple: Yes, she does, but the data would have to be wrong. The NDP would have to sweep Edmonton (which is likely), gain at least three quarters of Calgary seats (which is not), and win pockets of districts in smaller cities such as Lethbridge and Red Deer.

Polling data is right more often than it is wrong the same way meteorologists, in the long run, get their forecasts right more often than not. But the Alberta polling miss of 2012 still hovers over forecasters' minds.

Polls may capture the general mood of an electorate, but they do not cast votes. Only voters do. We shall see which party Albertan voters will have chosen tomorrow to lead their government for the next four years.


Thanks to all readers of this page. I couldn't have done it without your support. 

Time to crack open a cold one. Go Flames. Boo Leafs.



Philippe J. Fournier is the creator of Qc125 and 338Canada. He teaches physics and astronomy at Cégep de Saint-Laurent in Montreal. For information or media request, please write to info@Qc125.com.


Philippe J. Fournier est le créateur de Qc125 et 338Canada. Il est professeur de physique et d'astronomie au Cégep de Saint-Laurent à Montréal. Pour toute information ou pour une demande d'entrevue médiatique, écrivez à info@Qc125.com.