Monday, April 18, 2016

Manitoba: no final prediction

Sad news, I have no final prediction for Manitoba. Predictions are things I do for fun and for your entertainment. This contrasts with Projections, which use hard math.

I have no prediction because I've decided to do a Projection. A long time ago, I decided to ignore the 6 "smaller" provinces, due to the fact that there are about 500 or so votes per riding that behave nothing like the math would suggest, due to local concerns (like knowing the candidate personally) and when you are talking about results that look like 4,000 votes VS 3,000 votes, such as you find here in Manitoba, those 500 voters can matter much more than in federal ridings where you can easily see margins of 22,000 to 23,000.

However, Manitoba has one huge advantage: Winnipeg. Cities tend to behave more as the math would suggest as opposed to rural areas. Compare Manitoba to it's neighbour, Saskatchewan. They have two smaller cities (smaller cities make things much worse, think of it like a square root) with 24 seats VS 37 outside the cities. In Manitoba, however, Manitoba has 31 Winnipeg seats, vs 26 non-Winnipeg seats. Manitoba has far more people living in it's large cities, and has cities that are far larger.

As such, if the math is to work in any province that has under 4 million people, Manitoba is the best bet. So. Lets dive in:

Recent Polls have actually - when averaged - indicated the Liberal "dip" might not be as low as we had thought. Of course (at the time of writing) the two most recent polls indicate a strong recovery and a massive drop respectively. On the whole, when averaged, the Liberal vote has stabilized at somewhere around 17%.

You'll notice a few X's. These are ridings where I've decided to over-ride the math. First, a few ridings where I chose not to:

Brandon East: The Liberal campaign here has had some problems, including some anti-Brandon comments made by the candidate in Brandon West. If the Liberals were doing even better, I'd consider putting them higher, but as it stands, I just can not project a Liberal win in Brandon East based on these numbers.

Fort Rouge: This is where the Liberal Leader is running against Wab Kinew. Why? Simply; the math suggests a massive Liberal lead here. I do think Bokhari will drive voters away, but the lead is so large (even the NDP+PC vote only beats the Liberal vote by 0.3%) that I can not see this gap being overcome, even with a popular candidate. This does not even get into possible strategic voting from Tories eager to split the left-wing vote by voting Liberal.

St. Boniface: This is where Greg Selinger is running. The math behind a victory for him is very strong, and his chances of losing are extraordinarily low. Excuse the allcaps, but; IF the Tory candidate was potentially able to tap into the francophone vote by being known by the community, then MAYBE I would consider this. I am NOT saying that I'm putting the NDP in the column because they are running a white guy, I am saying that I'm not even willing to consider making such a huge math leap when the candidate can not be considered a "star" in any way.

And now where I have overturned the results

Burrows: I have this riding going Liberal even though the math indicates it should be a 3-way race with a narrow NDP lead over the Tories. Why? The simple answer is Lamoureux and Lamoureux. First is Kevin, who represents part of this riding Federally, and has built a small Liberal base of support. The second is the Provincial candidate, who has the Lamoureux name. I believe that these two factors can combine into the few extra points needed to win a close 3-way race.

Kewatinook: The math suggests the Tories can win, but I have very strong doubts. The poll by poll results from 2011 indicate pockets of PC strength, pockets that, in the 2015 Federal election, frequently saw the CPC candidate beat by the Libertarians. It's no surprise that the 2011 candidate came from a reserve in the riding, the reserve he won heavily. The former candidate in 2007 was also from a reserve within the riding. The current PC Candidate is not. When you get this far north, things like that matter a great deal, and for that reason, I am considering this riding to see the largest increase in NDP vote in the province.

As such, that is my projection for the election tomorrow. I will only be doing an update if a we get a poll from one of the form that has not published in two weeks publishes a new poll. Given the margins between the Tories and NDP, the only real change from an update would be the possibility of the loss of Liberal seats.


The two most vulnerable Liberal ridings are The Maples, in the far north of Winnipeg, and Logan, in the middle of the "S" which could be lost to the PC Party and the NDP respectively.

No comments:

Post a Comment