Monday, October 26, 2020

BC election "results"

 The 'results' of the election are as follows:

You may note, some seats are still listed as too close to call. This is due to the mail in ballots, which have yet to be counted

I'll skip right to the business of the post; these the BCL-over-NDP vote gaps in the 6 closest ridings:

Will flip:
188 - Abbotsford Mission
180 - Vernon Monashee

Very likely to flip:
385 - Fraser Nicola

Might flip:
647 - Vancouver Langara
729 - Surrey White Rock
791 - Kamloops North Thompson

Some quick math. 

There are 575,000 or so ballots to be counted, covering the 87 ridings of BC. Polls suggest the NDP has an advantage in these votes. This is, very roughly, a third of all ballots.

Lets start with ballots already counted in an example riding. If it mirrored the province, and had 1000 voters, it would have 422 votes for the NDP, 351 for BCL, 166 for the Greens, and 61 others.

I want to veer into a tangent, and say, I consider the mail in ballots to be "10% for the NDP". Why is that? Lets look at what a riding of 667 counted voters; we get 281 NDP voters, 244 BCL, and 111 Green. Now lets add 333 mail in ballots: 183 NDP, 94 BCL, and 52 Green. Totals are 163 Green votes, 338 BCL, and 464 NDP; alongside 35 others. Now we jump into where this tangent comes into play using more math. 

We are going to mostly ignore the Green and "Other" numbers, as, no Green riding has the NDP within striking distance; and thus, it is only the Liberals that the NDP can take seats from. 

As such, lets assume instead of 351 or 338 votes in the ridings in question, the Liberals have taken exactly 1,000 votes, and, scale the NDP vote up to match it. This gives us 1202 NDP votes, and, 1373 NDP votes respectively. This is a gain of 1.14 for the NDP, comparing the combined Mail-In + Counted result (1373) VS the only-as-counted results (1202), a gain of 14%. 

I've simply decided to round that to 10%. So, lets see this in action with our earlier tangential example. The counted votes stay the same, 281 NDP, 244 BCL, and 111 Green. However, we now reduce the mail-in ballots by 10% (from 333 to 300), and, assume the mail-in ballots match the counted ballots. [after all, this entire mathematical process is to help us guess what the mail in ballots will say! So we must use the counted ballots as a base.] The results are as follows; 127 NDP, 103 BCL, 50 Green, and 20 others. The combined result is 408 NDP, 347 BCL, 161 Green.

Lastly, we add 100% of the earlier excluded votes in, as NDP votes. 33 of them. Thus the NDP goes from 408 to 441. 

The final province-wide popular vote would be 44.1% for the NDP vs 34.7% for the Liberals. Given the polls, this is generous to the Liberals. 

So, how does all that mathematical gobbledygook apply to the 575,000 mail in ballots?

Simple. Take 10% of these ballots (57,500) and divide them by the 87 ridings (661) and you end up with about 500 votes (more than that, 661 actually, and that 661 is already generous to the Liberals, but I'm trying to be super generous here)

And thus, we end up with at least 3, but as many as 6, ridings flipping to the NDP when the results are counted. 

No comments:

Post a Comment