Saturday, September 12, 2020

New Brunswick Prediction

 Lets start with the map!

As you can see on the map, I've made a few changes.

The map now includes ridings that are "outside chances". This means the amount of ridings each party can win has been expanded. This is clearest by the Green gains around Fredericton. 

One assumption I've been forced to make is that PANB is not doing as well as some polls think they are. Why is this?

Many pollsters will "prompt". That means they'll say "Who are you voting for, PC, Lib, Grn, NDP, PANB, or other". When you "prompt" you tend to get higher numbers of unpopular parties. Contrast this with someone asking "PC, Lib, Grn, or other". In the latter, you are likely to see much lower NDP and PANB numbers in your poll. The only difficulty is trying to guess if the pollsters 'should have prompted' as the small party is going to do well, or, 'should not have prompted' as the party is going to do poorly.

It is my assumption that the pollsters should not be prompting for PANB. However. I could be wrong. 

One reason PANB may be doing well in prompted polls but not in unprompted polls is that people are still not fully willing to give Higgs a majority, but are at a loss as to how to vote as a result. If this is true, then PANB will indeed take 10%+ of the vote, as, once people take a physical look at their ballot, they'll know exactly what to do. 

If, however, people are not really "considering" PANB in their minds as a viable option, then seeing it no the ballot paper will not matter. Historically, this happened a lot to the NDP in places like NB or PEI, where prompted polls will always show a higher number than the election. 

Regardless, I'll be posting updates on twitter, and those wanting the latest info from me, can find it there. 

No comments:

Post a Comment