Tuesday, September 13, 2016

More STV

Of the other ridings named, I want to start by addressing the "PEI" riding, as this is the shortest. We'll end up with a single province-wide riding. This means the threshold for winning a seat is 20%+1. One thing to keep in mind when you look at real STV results is that 100% of voters do not always stay within a party. In the 2015 election, the Liberals did extremely well. Normally we see a more balanced result. Lets make up a fake future election based on that. This is what we may see in a future election.

I will presume 84,999 ballots cast, a reasonable number for PEI. This places the threshold at exactly 17,000.

In elections where one party does very well, we could see one of the main parties (Liberal, Conservative) running 3 candidates, but outside of that, we are far far more likely to see each only run 2. The NDP is smart to only run 1, as are the smaller parties. In the last 5 elections, the CHP has run, and in the election before that, the CHP candidate in 4 of those 5 elections, ran as an Independent, due to registration problems. As such I will be using the 4 national parties, and the CHP in this example. I will use names from the 2015 election for the names of candidates, though no guarantee any of these people will want to run for another term.

Here is a possible result.

17,181 - Easter - L  ELECTED
17,094 - Shea - C  ELECTED
16,838 - Pataki - C
16,822 - MacAulay - L
13,263 - Cann - N
3304 - Viau - G
497 - Squires - H

you may wonder how exactly the main parties manage such quality vote management. There are a few ways to do this. You could tell everyone born on an odd numbered day to vote for one and everyone even for another. You could ask everyone with a name that start with A-M to vote for one, etc etc, but the most logical way, and the usual way parties do this, is geographically. The Liberals, in this example, would likely have said something to the effect of "If you lived in the ridings of Egmont or Malpeque, vote for Easter, and if you lived in Cardigan or Charlottetown, vote for MacAulay." This could produce a result like what we see. Had the Tories done the same they would have been more imbalanced, so they could fix this by asking everyone living in Charlottetown, Winsloe, and on or east of Winslow Road to vote for Pataki, while asking anyone who lived east of Charlottetown, and on or east of Rustico Road to vote for Shea. This means that in general, voters are considering either Easter/Shea or MacAulay/Pataki, but that some voters in the Brackley/Stanhope area, are considering Easter/Pataki. The ability of the parties to pick a dividing line is what makes this work. Parties that pick the right dividing line do very well indeed, while picking the wrong one, will leave you terribly unbalanced. However, lets look at how this would play out.

First we start at the top. We distribute Easter votes.181 of them. Most of them, lets say 150, will go straight to MacAulay. So what of the other 31? Well 10 of these folks are voting for "progressives" so they voted NDP second. 8 of them liked Easter in cabinet, but also liked Shea in cabinet, so they voted for her. 6 of them care about the environment, so they voted Green. 4 of them have met Easter personally, as well as Pataki, and like them both, so that's where they go. 2 of them consider themselves Christian, so go CHP. And 1 of them really don't like any politicians except Easter, on the grounds that they are related, he's 18, and was pressured to vote, so they did not rank anyone else. This becomes, effectively, a dead vote.

17,000 - Easter - L  ELECTED
17,104 - Shea - C  ELECTED
16,972 - MacAulay - L
16,842 - Pataki - C
13,273 - Cann - N
3310 - Viau - G
499 - Squires - H
1 - Dead

Now we distribute the 104 Shea votes. Much like above, all of them have a host of different reasons for voting how they did. A few of them voted Shea #1, and Easter #2. With Easter eliminated due to winning, those votes need to go to their #3 choice. As such, the 3rd round looks like this:

17,000 - Easter - L  ELECTED
17,000 - Shea - C  ELECTED
16,977 - MacAulay - L
16,929 - Pataki - C
13,276 - Cann - N
3315 - Viau - G
502 - Squires - H
2 - Dead

After this we need to eliminate the last placed candidate. However, many people voted CHP precisely because they didn't like the other 4 parties; as such, this may well be the next round:

17,129 - Pataki - C ELECTED
17,004 - MacAulay - L ELECTED
17,000 - Easter - L  ELECTED
17,000 - Shea - C  ELECTED
13,296 - Cann - N
3325 - Viau - G
247 - Dead

With 4 candidates elected, the count ends.

This is the last time I directly look at counts. In a few hours, more ridings.

No comments:

Post a Comment