Quebec Election Tracking

Based on the relative success of my Ontario poll tracking, I'm going to try my hand at New Brunswick's upcoming election. Follow along, it'll be fun!

Ontario's elections take place in October, and I'll be tracking the polls here. I'm using this to practice working with my model in anticipation of better coverage for the Alberta election, and am using a hybrid between proportional and uniform swing models that I've cobbled together.

Last updated: October 1, 2018

First of all, the big picture since the last election:



As well as the trend since the beginning of the year:


I'm also working on a seat projection, again only since the beginning of the year:



This seat projection is using my "Ontario" model, which worked well for Ontario, but not especially well for the New Brunswick election due to inefficient vote distribution. For this graph, the dark zones represent 50% odds of occurrence, and the lighter zones represent 90% odds. Based on this model, I estimate that if an election were today, the CAQ have an 81.2% chance of winning, most likely (43.3%) with a minority.

I'm certainly not the only person to be doing this, and I recommend comparing and contrasting with some of these other fun psephologists:

No comments: