Minneapolis Mayoral Election 2017

Elections - City Councils and Commissions - Policies
MattW
Union Depot
Posts: 371
Joined: June 13th, 2015, 5:05 pm

Re: Minneapolis Mayoral Election 2017

Postby MattW » November 8th, 2017, 1:36 pm

With the current system of three choices and 15 candidates, there are 2730 possible ballots. With six choices there are 3,603,600. Hand counting with six will be a time suck, but very easy for a computer to do.

David Greene
IDS Center
Posts: 4760
Joined: December 4th, 2012, 11:41 am

Re: RE: Re: Minneapolis Mayoral Election 2017

Postby David Greene » November 8th, 2017, 1:47 pm

MattW wrote:With the current system of three choices and 15 candidates, there are 2730 possible ballots. With six choices there are 3,603,600. Hand counting with six will be a time suck, but very easy for a computer to do.
The number of possible combinations of ballots has exactly zero to do with the algorithmic complexity of counting.

The growth in the number of possible ballot combinations is factorial in the number of candidates, even more rapid growth than exponential. The counting is still polynomial.

"Exponential," even used colloquially, gives an impression of a much harder job than it actually is. I get that "polynomial" can turn people off. So I just tell people, "it's not bad at all," and if they want more detail I explain that computer people are professionally trained in this stuff and if you don't trust me (which is fine!) then just go ask another computing professional, preferably someone with a degree.

amiller92
Wells Fargo Center
Posts: 1759
Joined: October 31st, 2014, 12:50 pm

Re: Minneapolis Mayoral Election 2017

Postby amiller92 » November 8th, 2017, 1:52 pm

xandrex wrote:
November 8th, 2017, 12:39 pm
I even know someone who did Frey/Hodges/Hoch!
I don't get the Hoch in this grouping. Well, I do, but personally I think it says something about how little people actually understand issues in city elections. They've heard of Hoch ($$$) and he seems like a good guy, so okay.
I think this actually bodes worse for Betsy, because I think Jacob might be sitting on a larger share of her 2nds than people think.
I think it's entirely possible. Heck, at the caucuses I was Frey/Hodges. Something Frey definitely didn't seem to see as making much sense when I spoke to him beforehand.

That said, someone who voted NLP first is probably an equity voter and thus probably not particularly likely to rank Frey second. Only slightly less true for Dehn.

RailBaronYarr
Capella Tower
Posts: 2702
Joined: September 16th, 2012, 4:31 pm

Re: Minneapolis Mayoral Election 2017

Postby RailBaronYarr » November 8th, 2017, 1:54 pm


jebr
Nicollet Mall
Posts: 171
Joined: April 9th, 2013, 1:04 am
Location: St. Paul (Hamline-Midway)

Re: Minneapolis Mayoral Election 2017

Postby jebr » November 8th, 2017, 1:56 pm

Yeah, the exact number of combinations doesn't really matter. You could pretty much keep the ranking of voters in some sort of table or entry from first to sixth (or whatever) and once the first choice drops out, push their other ranks "ahead" so the "first choice" vote is actually a "current choice" vote. If someone's second/third/whatever choice drops out, keep pushing until you hit a candidate that's still in or you've exhausted the rankings. Rinse and repeat until you get someone who's crossed the threshold or has the most after there's just two candidates left.

It's also not terribly difficult to do by hand (versus only having three, anyways.)

amiller92
Wells Fargo Center
Posts: 1759
Joined: October 31st, 2014, 12:50 pm

Re: Minneapolis Mayoral Election 2017

Postby amiller92 » November 8th, 2017, 2:00 pm

No surprise, then, and it was the Hodges second and thirds that put him over the top.

Hoch spent all that money and wound up fourth. And no one actually getting to 50%-1 and 11914 effectively disenfranchised because we only get to rank 3 (or they chose to rank fewer than 3). ETA: Oops, that exhausted number was just in the last round. Total was 22,835.

ETA: Oh, and Frey getting 9888 votes when Hoch was eliminated, about 2/3 of those reallocated.
Last edited by amiller92 on November 8th, 2017, 2:07 pm, edited 1 time in total.

LakeCharles
Foshay Tower
Posts: 833
Joined: January 16th, 2014, 8:34 am
Location: Kingfield

Re: Minneapolis Mayoral Election 2017

Postby LakeCharles » November 8th, 2017, 2:06 pm

So it was
1. Frey
2. Dehn
3. Hodges
4. Hoch
5. Levy-Pounds

As for Levy-Pounds support, 40% went to Dehn, 30% to Hodges, 20% to Frey, and 10% to Hoch.
Hoch supporters remaining went 60% to Frey, 20% each to Hodges and Dehn.
Hodges supporters remaining were 50% Dehn, 50% Frey.

RailBaronYarr
Capella Tower
Posts: 2702
Joined: September 16th, 2012, 4:31 pm

Re: Minneapolis Mayoral Election 2017

Postby RailBaronYarr » November 8th, 2017, 2:08 pm

Yeah, Betsy's 2nd/3rds went more to Dehn than Frey (barely), Frey's biggest bump (aside from his lead thanks to 1st choice votes) was from Hoch.

Which.

LakeCharles
Foshay Tower
Posts: 833
Joined: January 16th, 2014, 8:34 am
Location: Kingfield

Re: Minneapolis Mayoral Election 2017

Postby LakeCharles » November 8th, 2017, 2:13 pm

amiller92 wrote:
November 8th, 2017, 2:00 pm
And no one actually getting to 50%-1 and 11914 effectively disenfranchised because we only get to rank 3 (or they chose to rank fewer than 3). ETA: Oops, that exhausted number was just in the last round. Total was 22,835.
Most of this was people not putting a 2nd or 3rd. Meaning that adding more on the back end would not enfranchise many people.

tmart
Union Depot
Posts: 348
Joined: October 6th, 2017, 10:05 am
Location: Expat

Re: RE: Re: Minneapolis Mayoral Election 2017

Postby tmart » November 8th, 2017, 2:15 pm

David Greene wrote:
November 8th, 2017, 1:47 pm
MattW wrote:With the current system of three choices and 15 candidates, there are 2730 possible ballots. With six choices there are 3,603,600. Hand counting with six will be a time suck, but very easy for a computer to do.
The number of possible combinations of ballots has exactly zero to do with the algorithmic complexity of counting.

The growth in the number of possible ballot combinations is factorial in the number of candidates, even more rapid growth than exponential. The counting is still polynomial.

"Exponential," even used colloquially, gives an impression of a much harder job than it actually is. I get that "polynomial" can turn people off. So I just tell people, "it's not bad at all," and if they want more detail I explain that computer people are professionally trained in this stuff and if you don't trust me (which is fine!) then just go ask another computing professional, preferably someone with a degree.
Heh, I'm one of those professionally-trained computer people, whoops! I should have known my audience and preferred the more precise term "factorial" to "exponential," but it was this factorial growth I was alluding to. I agree that the algorithmic complexity of counting is polynomial, but the number of permutations still matters.

My understanding from this page is that the current counting process actually separates the ballots into literal piles and enters their counts, separately, into a spreadsheet. If that's the case, the jump from 3 to 6 would indeed overwhelm the existing process. It is of course possible to devise other ways of counting, though if a hand-count or hand-audit is a requirement, it becomes more challenging.

acs
Wells Fargo Center
Posts: 1378
Joined: March 26th, 2014, 8:41 pm

Re: Minneapolis Mayoral Election 2017

Postby acs » November 8th, 2017, 2:16 pm

Strib is calling it for Frey

twincitizen
Moderator
Posts: 6203
Joined: May 31st, 2012, 7:27 pm
Location: Standish-Ericsson

Re: Minneapolis Mayoral Election 2017

Postby twincitizen » November 8th, 2017, 2:18 pm

Kinda cool that you can (sorta) see how many voters went "bullet ballot" (i.e. chose only their favorite candidate, did not rank others) by moving through the elimination rounds.

2,630 voters did not rank any of the top 5 candidates, only ranking among the 14 joke candidates.
2,119 voters did not rank any of the top 4, only ranking Levy-Pounds (*and possibly previously eliminated candidates)
6,172 voters did not rank any of the top 3, only ranking Hoch (*and possibly previously eliminated candidates)

Adam, there's no guarantee that many voters were "effectively disenfranchised", because we don't know IF they actually ranked more than one candidate. I'm not saying we shouldn't be able to rank 6 candidates like St. Paul, we definitely should! (EDIT: I see LakeCharles already pointed this out)

amiller92
Wells Fargo Center
Posts: 1759
Joined: October 31st, 2014, 12:50 pm

Re: Minneapolis Mayoral Election 2017

Postby amiller92 » November 8th, 2017, 2:22 pm

LakeCharles wrote:
November 8th, 2017, 2:13 pm
Most of this was people not putting a 2nd or 3rd. Meaning that adding more on the back end would not enfranchise many people.
Do we know that? I don't think so. There were 2630 exhausted in the first round of elimination. Presumably most of those had no second or third, but some of them could have included 3 minor candidates. I don't think we can guess about any of the later rounds.

amiller92
Wells Fargo Center
Posts: 1759
Joined: October 31st, 2014, 12:50 pm

Re: Minneapolis Mayoral Election 2017

Postby amiller92 » November 8th, 2017, 2:24 pm

twincitizen wrote:
November 8th, 2017, 2:18 pm
Kinda cool that you can (sorta) see how many voters went "bullet ballot" (i.e. chose only their favorite candidate, did not rank others) by moving through the elimination rounds.

2,630 voters did not rank any of the top 5 candidates, only ranking among the 14 joke candidates.
2,119 voters did not rank any of the top 4, only ranking Levy-Pounds (*and possibly previously eliminated candidates)
6,172 voters did not rank any of the top 3, only ranking Hoch (*and possibly previously eliminated candidates)

Adam, there's no guarantee that many voters were "effectively disenfranchised", because we don't know IF they actually ranked more than one candidate. I'm not saying we shouldn't be able to rank 6 candidates like St. Paul, we definitely should! (EDIT: I see LakeCharles already pointed this out)
First, that's only roughly half of the exhausted ballots. Seconds, that's why I said "(or they chose to rank fewer than 3)"

David Greene
IDS Center
Posts: 4760
Joined: December 4th, 2012, 11:41 am

Re: Minneapolis Mayoral Election 2017

Postby David Greene » November 8th, 2017, 2:28 pm

jebr wrote:
November 8th, 2017, 1:56 pm
Yeah, the exact number of combinations doesn't really matter. You could pretty much keep the ranking of voters in some sort of table or entry from first to sixth (or whatever) and once the first choice drops out, push their other ranks "ahead" so the "first choice" vote is actually a "current choice" vote. If someone's second/third/whatever choice drops out, keep pushing until you hit a candidate that's still in or you've exhausted the rankings. Rinse and repeat until you get someone who's crossed the threshold or has the most after there's just two candidates left.
Yes, that's essentially the algorithm I and a colleague worked out in about 20 minutes on Facebook (complete with pseudocode!). There may be even faster algorithms. It turns out that for this algorithm, the majority of the time is spent preparing the data (sorting, etc.) for the fastest possible processing. A better ballot design might even eliminate the need for that.

LakeCharles
Foshay Tower
Posts: 833
Joined: January 16th, 2014, 8:34 am
Location: Kingfield

Re: Minneapolis Mayoral Election 2017

Postby LakeCharles » November 8th, 2017, 2:36 pm

amiller92 wrote:
November 8th, 2017, 2:22 pm
LakeCharles wrote:
November 8th, 2017, 2:13 pm
Most of this was people not putting a 2nd or 3rd. Meaning that adding more on the back end would not enfranchise many people.
Do we know that? I don't think so. There were 2630 exhausted in the first round of elimination. Presumably most of those had no second or third, but some of them could have included 3 minor candidates. I don't think we can guess about any of the later rounds.


We know that there were 104,297 first place votes in total, there were 92,273 second place votes in total, and 78,188 third place votes in total.

So 12,024 people elected not to put a second choice, and 26,109 elected not to put a third choice. This represents the majority of the exhausted ballots.

Unless all of those people voted for Dehn or Frey first.
Last edited by LakeCharles on November 8th, 2017, 2:37 pm, edited 1 time in total.

David Greene
IDS Center
Posts: 4760
Joined: December 4th, 2012, 11:41 am

Re: RE: Re: Minneapolis Mayoral Election 2017

Postby David Greene » November 8th, 2017, 2:36 pm

tmart wrote:
November 8th, 2017, 2:15 pm
My understanding from this page is that the current counting process actually separates the ballots into literal piles and enters their counts, separately, into a spreadsheet. If that's the case, the jump from 3 to 6 would indeed overwhelm the existing process. It is of course possible to devise other ways of counting, though if a hand-count or hand-audit is a requirement, it becomes more challenging.
Yeah, that seems on the surface like a bad algorithm, but I haven't put any thought into the best way to process ballots by hand.

What's frustrating about this is a computer scans the ballots initially. Literally all that's needed is for the computer to remember each ballot and a pointer to it's "current" ranked candidate. Then it's simply a matter of counting and updating pointers as candidates are eliminated. Easy to do in a computer, probably much harder by hand.

The bottom line for me is that there's been a collossal dropping of the ball by someone(s) somewhere to not have certified counting software after EIGHT FRICKING YEARS!

amiller92
Wells Fargo Center
Posts: 1759
Joined: October 31st, 2014, 12:50 pm

Re: Minneapolis Mayoral Election 2017

Postby amiller92 » November 8th, 2017, 2:43 pm

LakeCharles wrote:
November 8th, 2017, 2:36 pm

Unless all of those people voted for Dehn or Frey first.
Right, I was confused briefly as to how there could have been more "missing" third place ballots than exhausted ballots, but the answer is some number of bullet voters for Frey and Dehn.

Okay, I agree that it doesn't look like getting to rank more would make much difference in the exhausted ballots.

David Greene
IDS Center
Posts: 4760
Joined: December 4th, 2012, 11:41 am

Re: Minneapolis Mayoral Election 2017

Postby David Greene » November 8th, 2017, 2:46 pm

None of my ranked candidates won and while my fourth choice would have been Frey, I'm still pretty uneasy that I didn't get to have a voice in that.

amiller92
Wells Fargo Center
Posts: 1759
Joined: October 31st, 2014, 12:50 pm

Re: Minneapolis Mayoral Election 2017

Postby amiller92 » November 8th, 2017, 2:49 pm

Didn't you rank Dehn first? How was your voice not counted (I ranked him third, so mine was).


Return to “Local Politics and Governance”

Who is online

Users browsing this forum: No registered users and 3 guests