Now that the list of candidates for the municipal election is out, your correspondent can finally make statistics from reality. And with actual opinion polls come the ability to vote strategically, at least in the race for council. Because voting is like horse races: a lot more fun if your pick actually wins.
First, let's see the results of the first opinion poll.
(Click on it for the full-size version.)
So what does all this mean?
First of all, Ken Latour is almost certainly getting re-elected. The low rank / high rank columns work like this: "low rank" is how low a candidate would place if s/he polls at the bottom of his/her confidence interval and everyone else polls at the top of theirs. So in this case we're showing Latour placing eighth or better, 19 times out of 20. That's how I get pretty confident that he'll still be here October 16.
I'm sure anyone who follows town politics is not at all surprised to see the next three names. Jameson, Jungkind and Mapes are big names; they're sure to get plenty of support on that basis alone. I would also count Vince McKay in that group. It's within the measurement uncertainty. What's more surprising is the bottom of the list. In this case we have the opposite of the Latour scenario: guys who would not get elected even if they poll the top of their confidence interval and everyone else polls the bottom of theirs. McPherson and St. John barely made it out of the "not getting elected 19 times out of 20" by the time the survey closed. But hey, this is statistics. There's always that 20th time out of 20 – and strategic voting. Candow and Coakwell are marginal. A best place of fourth 19 times out of 20 doesn't rule them out, but they're definitely coming from behind. I have never heard of Roger Candow before, but I'm surprised at Coakwell. I rather thought he'd be in the almost-sure-wins than in the almost-sure-losses.
Finally there are the six in the middle of the pack (seven if you count McKay, but I think he's in fifth), who are neither almost certainly in nor almost certainly out. But since five spots out of eight are already as good as taken, these six are looking at more or less 50/50 odds. So this is where most of your strategizing is going to take place.
The key thing to understand about voting for council is that you can vote for up to eight councillors. But you don't have to. If you vote for eight candidates, they all go up one point, obviously. That means the guys who started at the bottom stay at the bottom. In order for the dark horses to gain ground, you want to vote for as few as possible.
Consider the following. Say that you're BFFs with Candow, Coakwell, McPherson and/or St. John. Your guy doesn't have a lot of votes, so if you add on to the other guys' tallies, you're not doing your candidate any favours. You can vote for Latour, because he's strategically neutral: he's getting elected whether you vote for him or not. (Technically, that's not true. Some people need to actually vote for Latour rather than assuming he doesn't need their votes because he's got so many already.) And you can vote or not vote for the Big Four, as you see fit, because it's not likely they'll get voted off anyway, so that's not who's standing in your candidate's way. The main thing you want to avoid is voting for the six middle guys. Four of them need to lose in order for one of the bottom four to get in. So don't give them any help. Vote for your buddy and no one else.
Likewise if there is one of the middle six that you really really really want to see on council, don't vote for the other five. In the 2009 election, the difference between eighth place (Bernard Dueck) and tenth place (Rashna Bundan) was only 14 votes. How many of those 14 voters were really passionate about Dueck? If 14 more Bundan fans had left Dueck off their ballot, she could have been in council.
Ok, so far we've got:
a) Vote for Latour. Or don't. It won't make much difference. (To anything, really.)
b) Doesn't make much difference whether or not you vote for Jameson, Jungkind, Mapes and/or McKay, either.
c) If you really love one candidate out of the middle six or the bottom four, vote for ONLY THAT ONE out of the ten.
On the other hand, suppose there is someone you really really really DON'T WANT on council – and I'm sure we all know who you're thinking. In this case, you really want to fill your ballot. Because if you had been trying to keep Dueck out in 2009, and you picked only seven and didn't vote for Bundan or Gibb (who finished ninth only 13 votes behind Dueck), it's your fault he got in. Just sayin'. There was a good example in 2003, when only nine candidates ran of whom one was extremely unpopular. The expressed votes were higher than in the next three elections. The first eight candidates got from 939 to 1064 votes, which is very high (in 2009 Dueck got elected with less than half of that). The unpopular one got 298. So you see, you can keep the bad eggs out, if you really try. But because we have so many candidates this time, you want to pick the right eight – the ones most likely to finish ahead of the one you're blackballing. Hence the need for opinion polls, right? Because if this poll is representative of everyone's voting intentions, then you know which eight candidates are most likely to finish ahead of That One.
Where it gets complicated is if there are candidates you really really really want elected AND candidates you really really really don't want. Because the guys you put as filler to keep That One out might also boot one of your pet candidates, especially those candidates who are polling lower than That One. Conversely if you're picking filler from the bottom of the odds list, they might not get high enough to keep That One out.
I'll tell you what I think I'm gonna do – subject to change as the polls fluctuate. Obviously I'm voting for myself. Probably also Kandis Jameson, who will be in anyway and whom I like as a person (to the extent that I know her, which is admittedly not much). And I'm gonna vote for Coakwell, too. I wanted him gone when I thought he was gonna be way ahead of me, but since he's well behind in the polls, I'd actually rather like to have him on council with me. I'm not aware that we have any ideas in common, but then I'm not aware of any of his ideas. That's why I like him at his point: whether he does or does not have ideas, and whether I agree with them or not, at least he leaves room for others to speak.
So far, so good, but here's the thing: just like you, I want to keep someone out of the council. So do I vote for the Big Four plus one of the middle six, which would also work against Coakwell and possibly against me too, or do I leave it alone because a) I'd rather have me and Coakwell AND That One on council than none of the three and b) I trust the other voters' good sense?
It's a doozy of a pickle for sure. All I know so far is, I'm voting for me.
Kthxbai.
First, let's see the results of the first opinion poll.
So what does all this mean?
First of all, Ken Latour is almost certainly getting re-elected. The low rank / high rank columns work like this: "low rank" is how low a candidate would place if s/he polls at the bottom of his/her confidence interval and everyone else polls at the top of theirs. So in this case we're showing Latour placing eighth or better, 19 times out of 20. That's how I get pretty confident that he'll still be here October 16.
I'm sure anyone who follows town politics is not at all surprised to see the next three names. Jameson, Jungkind and Mapes are big names; they're sure to get plenty of support on that basis alone. I would also count Vince McKay in that group. It's within the measurement uncertainty. What's more surprising is the bottom of the list. In this case we have the opposite of the Latour scenario: guys who would not get elected even if they poll the top of their confidence interval and everyone else polls the bottom of theirs. McPherson and St. John barely made it out of the "not getting elected 19 times out of 20" by the time the survey closed. But hey, this is statistics. There's always that 20th time out of 20 – and strategic voting. Candow and Coakwell are marginal. A best place of fourth 19 times out of 20 doesn't rule them out, but they're definitely coming from behind. I have never heard of Roger Candow before, but I'm surprised at Coakwell. I rather thought he'd be in the almost-sure-wins than in the almost-sure-losses.
Finally there are the six in the middle of the pack (seven if you count McKay, but I think he's in fifth), who are neither almost certainly in nor almost certainly out. But since five spots out of eight are already as good as taken, these six are looking at more or less 50/50 odds. So this is where most of your strategizing is going to take place.
The key thing to understand about voting for council is that you can vote for up to eight councillors. But you don't have to. If you vote for eight candidates, they all go up one point, obviously. That means the guys who started at the bottom stay at the bottom. In order for the dark horses to gain ground, you want to vote for as few as possible.
Consider the following. Say that you're BFFs with Candow, Coakwell, McPherson and/or St. John. Your guy doesn't have a lot of votes, so if you add on to the other guys' tallies, you're not doing your candidate any favours. You can vote for Latour, because he's strategically neutral: he's getting elected whether you vote for him or not. (Technically, that's not true. Some people need to actually vote for Latour rather than assuming he doesn't need their votes because he's got so many already.) And you can vote or not vote for the Big Four, as you see fit, because it's not likely they'll get voted off anyway, so that's not who's standing in your candidate's way. The main thing you want to avoid is voting for the six middle guys. Four of them need to lose in order for one of the bottom four to get in. So don't give them any help. Vote for your buddy and no one else.
Likewise if there is one of the middle six that you really really really want to see on council, don't vote for the other five. In the 2009 election, the difference between eighth place (Bernard Dueck) and tenth place (Rashna Bundan) was only 14 votes. How many of those 14 voters were really passionate about Dueck? If 14 more Bundan fans had left Dueck off their ballot, she could have been in council.
Ok, so far we've got:
a) Vote for Latour. Or don't. It won't make much difference. (To anything, really.)
b) Doesn't make much difference whether or not you vote for Jameson, Jungkind, Mapes and/or McKay, either.
c) If you really love one candidate out of the middle six or the bottom four, vote for ONLY THAT ONE out of the ten.
On the other hand, suppose there is someone you really really really DON'T WANT on council – and I'm sure we all know who you're thinking. In this case, you really want to fill your ballot. Because if you had been trying to keep Dueck out in 2009, and you picked only seven and didn't vote for Bundan or Gibb (who finished ninth only 13 votes behind Dueck), it's your fault he got in. Just sayin'. There was a good example in 2003, when only nine candidates ran of whom one was extremely unpopular. The expressed votes were higher than in the next three elections. The first eight candidates got from 939 to 1064 votes, which is very high (in 2009 Dueck got elected with less than half of that). The unpopular one got 298. So you see, you can keep the bad eggs out, if you really try. But because we have so many candidates this time, you want to pick the right eight – the ones most likely to finish ahead of the one you're blackballing. Hence the need for opinion polls, right? Because if this poll is representative of everyone's voting intentions, then you know which eight candidates are most likely to finish ahead of That One.
Where it gets complicated is if there are candidates you really really really want elected AND candidates you really really really don't want. Because the guys you put as filler to keep That One out might also boot one of your pet candidates, especially those candidates who are polling lower than That One. Conversely if you're picking filler from the bottom of the odds list, they might not get high enough to keep That One out.
I'll tell you what I think I'm gonna do – subject to change as the polls fluctuate. Obviously I'm voting for myself. Probably also Kandis Jameson, who will be in anyway and whom I like as a person (to the extent that I know her, which is admittedly not much). And I'm gonna vote for Coakwell, too. I wanted him gone when I thought he was gonna be way ahead of me, but since he's well behind in the polls, I'd actually rather like to have him on council with me. I'm not aware that we have any ideas in common, but then I'm not aware of any of his ideas. That's why I like him at his point: whether he does or does not have ideas, and whether I agree with them or not, at least he leaves room for others to speak.
So far, so good, but here's the thing: just like you, I want to keep someone out of the council. So do I vote for the Big Four plus one of the middle six, which would also work against Coakwell and possibly against me too, or do I leave it alone because a) I'd rather have me and Coakwell AND That One on council than none of the three and b) I trust the other voters' good sense?
It's a doozy of a pickle for sure. All I know so far is, I'm voting for me.
Kthxbai.
No comments:
Post a Comment