Discussion:
[EM] RCV in SF Mayoral election
Christopher Colosi
2018-06-09 01:24:10 UTC
Permalink
Curious to hear people’s thoughts on some issues.

1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the vote
currently stands, 144 votes separate the top two candidates (<0.1%) and
over 16,000 ballots have been exhausted (all 3 choices eliminated). About
9% of voters have been removed from the pool. It is very possible that the
result would have shifted if they had the opportunity to rank a 4th
candidate, and therefore, it is possible that we won’t elect the person who
truly represents the majority. Is this common? This is probably an
abnormally close race. Thoughts?

2. What are your thoughts on London Breed’s response to being asked if RCV
is fair? She stated “This is the system we are working with. That’s a
discussion we can have at a later time. For now, we’re stuck with it.”
- insinuating
it is not fair. I was quite bothered to have a Dem in a progressive city
insinuate that first past the post is more fair. It also felt divisive.
If Leno wins, will her supporters feel that democracy prevailed, or that
the election was stolen? She also presents herself as a minority candidate
and it is my understanding that RCV gives minority candidates better
chances and causes all candidates to be more likely to campaign to minority
communities. Am I mistaken? Are there any legitimate arguments that FPTP
can be more fair? Thoughts?

Regards,
—Chris
Greg Dennis
2018-06-09 02:21:09 UTC
Permalink
Post by Christopher Colosi
1. May not elect majority candidate
I haven't done an analysis of the cast vote record, but I highly doubt that
additional ranks would have put the winner over a majority of non-blank
ballots. From my experience with past RCV elections, the exhausted ballots
in the final round are largely voters who didn't use all the available
ranks, not those who used all the ranks and saw them all eliminated. If a
voter doesn't use all the ranks, it's an expression of indifference between
the remaining candidates: an abstention from any round not including the
ranked candidates. That's why including all non-blank ballots in the
denominator when determining a "majority" is odd to me: it _excludes_
abstentions from the first round but _includes_ abstentions from the final
round. My guess is you would find the winner _does_ have a majority if you
exclude _all_ abstentions. Happy for someone to do that analysis.

2. What are your thoughts on London Breed’s response to being asked if RCV
Post by Christopher Colosi
is fair?
I don't know, I'd expect most politicians to think systems that benefit
them are "fair" and those that don't as "unfair," so I don't think this
statement carries much of value. My $0.02.
Christopher Colosi
2018-06-09 02:59:55 UTC
Permalink
They separated out ballots “exhausted by overvotes” and “under votes” from
“exhausted ballots”. I assumed that meant there were 16,000 ballots
eliminated after all 3 choices were eliminated, but after closer analysis,
it appears you are correct that this includes voters who did not vote for 3
candidates. However, even if only a fraction of the exhausted ballots were
fully ranked and would have ranked more, it probably still surpasses the
144 vote gap currently deciding the election. Though then you would still
need a substantially larger number to actually sway it 144 votes. Still,
the possibility exists.

At start
Over votes: 339
Under votes: 1,149
Exhausted: 0

At end
Over votes: 401
Under votes: 1,149
Exhausted: 16,025

So, an over vote on a second or third rank clearly didn’t invalidate the
vote until that rank was reached. Happened 62 times on later ranks. Only
401 such voting errors is pretty good

Undervotes didn’t change, meaning that my interpretation was wrong.
Post by Christopher Colosi
1. May not elect majority candidate
I haven't done an analysis of the cast vote record, but I highly doubt
that additional ranks would have put the winner over a majority of
non-blank ballots. From my experience with past RCV elections, the
exhausted ballots in the final round are largely voters who didn't use all
the available ranks, not those who used all the ranks and saw them all
eliminated. If a voter doesn't use all the ranks, it's an expression of
indifference between the remaining candidates: an abstention from any round
not including the ranked candidates. That's why including all non-blank
ballots in the denominator when determining a "majority" is odd to me: it
_excludes_ abstentions from the first round but _includes_ abstentions from
the final round. My guess is you would find the winner _does_ have a
majority if you exclude _all_ abstentions. Happy for someone to do that
analysis.
2. What are your thoughts on London Breed’s response to being asked if RCV
Post by Christopher Colosi
is fair?
I don't know, I'd expect most politicians to think systems that benefit
them are "fair" and those that don't as "unfair," so I don't think this
statement carries much of value. My $0.02.
VoteFair
2018-06-09 05:01:42 UTC
Permalink
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck with it.”
- insinuating it is not fair. I was quite bothered to have a Dem in a
progressive city insinuate that first past the post is more fair. ...
This remark does not imply support for first past the post (FPTP, a.k.a
plurality counting).

There are other ways to count the preference marks on "ranked-choice"
ballots. In particular, pairwise counting could be used instead of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had been
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.

Pairwise counting does not result in any exhausted ballots. Unmarked
choices are an indication that the choices are equally disliked. And
multiple candidates being marked at the same preference level is also no
problem.

In other words, the ballots contain enough information that they can be
counted in other ways, besides instant-runoff counting. Those alternate
counting methods could reveal a clearer outcome.

In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the vote
currently stands, 144 votes separate the top two candidates (<0.1%) and
over 16,000 ballots have been exhausted (all 3 choices eliminated).
About 9% of voters have been removed from the pool. It is very possible
that the result would have shifted if they had the opportunity to rank a
4th candidate, and therefore, it is possible that we won’t elect the
person who truly represents the majority. Is this common? This is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being asked if
RCV is fair? She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck with it.”
- insinuating it is not fair. I was quite bothered to have a Dem in a
progressive city insinuate that first past the post is more fair. It
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also presents
herself as a minority candidate and it is my understanding that RCV
gives minority candidates better chances and causes all candidates to be
more likely to campaign to minority communities. Am I mistaken? Are
there any legitimate arguments that FPTP can be more fair? Thoughts?
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
robert bristow-johnson
2018-06-09 06:07:28 UTC
Permalink
Richard, a few points:



the limitation to only three levels of ranking is a problem.  if someone ranked all three levels and none of the candidates ranked were either London Breed nor Mark Leno, that voter was effectively "disenfranchised" by being unable to weigh in on the final choice of choosing the
mayor.  however, i think the news media made it clear that the race was really gonna be between Leno, Breed, and Kim, so these fringe voters might have a chance to insincerely mark either Leno or Breed as their 3rd choice and betray their *true* third choice and, in doing so, have an effect in
the final round.


ignoring the problem of only 3 ranking levels, it is not possible that London Breed is the Condorcet Winner (a.k.a. "pairwise champion").  it might be the case that Mark Leno or Jane Kim is the Condorcet Winner and if the latter is the case, this is another real indictment against STV
or IRV as a method of tallying RCV.  and your reverse namesake, FairVote, is partially (or mostly) to blame.


i wonder if the City of SF has a file of all of the cast and scanned ballots and the full ranking for each.  if so, and if they release it to the public, we can investigate if there is a Condorcet Winner and if that CW is or is not Mark Leno.  this would be
interesting.
L8r,
 
r b-j 
Post by VoteFair
... She stated &ldquo;This is the system we are working with. That&rsquo;s
a discussion we can have at a later time. For now, we&rsquo;re stuck with it.&rdquo;
- insinuating it is not fair. I was quite bothered to have a Dem in a
progressive city insinuate that first past the post is more fair. ...
This remark does not imply support for first past the post (FPTP, a.k.a
plurality counting).
There are other ways to count the preference marks on "ranked-choice"
ballots. In particular, pairwise counting could be used instead of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had been
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots. Unmarked
choices are an indication that the choices are equally disliked. And
multiple candidates being marked at the same preference level is also no
problem.
In other words, the ballots contain enough information that they can be
counted in other ways, besides instant-runoff counting. Those alternate
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people&rsquo;s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the vote
currently stands, 144 votes separate the top two candidates (<0.1%) and
over 16,000 ballots have been exhausted (all 3 choices eliminated).
About 9% of voters have been removed from the pool. It is very possible
that the result would have shifted if they had the opportunity to rank a
4th candidate, and therefore, it is possible that we won&rsquo;t elect the
person who truly represents the majority. Is this common? This is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed&rsquo;s response to being asked if
RCV is fair? She stated &ldquo;This is the system we are working with. That&rsquo;s
a discussion we can have at a later time. For now, we&rsquo;re stuck with it.&rdquo;
- insinuating it is not fair. I was quite bothered to have a Dem in a
progressive city insinuate that first past the post is more fair. It
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also presents
herself as a minority candidate and it is my understanding that RCV
gives minority candidates better chances and causes all candidates to be
more likely to campaign to minority communities. Am I mistaken? Are
there any legitimate arguments that FPTP can be more fair? Thoughts?
Regards,
&mdash;Chris
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
 
 
 


--



r b-j                         ***@audioimagination.com



"Imagination is more important than knowledge."

 
 
 
 
Greg Dennis
2018-06-09 13:25:36 UTC
Permalink
San Francisco always make the cast vote record public:
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports

Based on the most recent analysis of these numbers that I saw, Leno was
indeed the Condorcet winner, and if Breed were to beat Leno in the final
round, she would then necessarily be the Condorcet winner. The probability
of IRV not elected the Condorcet winner appears to be exceedingly low in
practice. We're up to about ~200 IRV elections conducted nationwide since
2004 and Burlington 2009 is the only case so far.

On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson <
the limitation to only three levels of ranking is a problem. if someone
ranked all three levels and none of the candidates ranked were either
London Breed nor Mark Leno, that voter was effectively "disenfranchised" by
being unable to weigh in on the final choice of choosing the mayor.
however, i think the news media made it clear that the race was really
gonna be between Leno, Breed, and Kim, so these fringe voters might have a
chance to insincerely mark either Leno or Breed as their 3rd choice and
betray their *true* third choice and, in doing so, have an effect in the
final round.
ignoring the problem of only 3 ranking levels, it is not possible that
London Breed is the Condorcet Winner (a.k.a. "pairwise champion"). it
might be the case that Mark Leno or Jane Kim is the Condorcet Winner and if
the latter is the case, this is another real indictment against STV or IRV
as a method of tallying RCV. and your reverse namesake, FairVote, is
partially (or mostly) to blame.
i wonder if the City of SF has a file of all of the cast and scanned
ballots and the full ranking for each. if so, and if they release it to
the public, we can investigate if there is a Condorcet Winner and if that
CW is or is not Mark Leno. this would be interesting.
L8r,
r b-j
Post by VoteFair
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck with
it.”
Post by VoteFair
- insinuating it is not fair. I was quite bothered to have a Dem in a
progressive city insinuate that first past the post is more fair. ...
This remark does not imply support for first past the post (FPTP, a.k.a
plurality counting).
There are other ways to count the preference marks on "ranked-choice"
ballots. In particular, pairwise counting could be used instead of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had been
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots. Unmarked
choices are an indication that the choices are equally disliked. And
multiple candidates being marked at the same preference level is also no
problem.
In other words, the ballots contain enough information that they can be
counted in other ways, besides instant-runoff counting. Those alternate
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the vote
currently stands, 144 votes separate the top two candidates (<0.1%) and
over 16,000 ballots have been exhausted (all 3 choices eliminated).
About 9% of voters have been removed from the pool. It is very possible
that the result would have shifted if they had the opportunity to rank a
4th candidate, and therefore, it is possible that we won’t elect the
person who truly represents the majority. Is this common? This is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being asked if
RCV is fair? She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck with it.”
- insinuating it is not fair. I was quite bothered to have a Dem in a
progressive city insinuate that first past the post is more fair. It
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also presents
herself as a minority candidate and it is my understanding that RCV
gives minority candidates better chances and causes all candidates to be
more likely to campaign to minority communities. Am I mistaken? Are
there any legitimate arguments that FPTP can be more fair? Thoughts?
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em for list
info
Post by VoteFair
----
Election-Methods mailing list - see http://electorama.com/em for list
info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts

e :: ***@voterchoicema.org
p :: 617.863.0746
w :: voterchoicema.org

:: Follow us on Facebook <https://www.facebook.com/voterchoicema> and
Twitter <https://twitter.com/voterchoicema> ::
VoteFair
2018-06-09 20:36:46 UTC
Permalink
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports

I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is
no data indicating exactly how those ballots were marked, so we lack
enough information to be sure of final results.

Converting instant-runoff counts into pairwise counts might be (probably
is?) possible, but I don't have time to do that analysis.
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only
case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is
rare when the number of ballots exceeds a few hundred.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots
had this issue. Right?

That's not a big number, but a fair counting method -- such as pairwise
counting -- would not have to discard any ballots.

The bigger number is "under votes" and admittedly pairwise counting
cannot compensate for a voter saying "here is the only acceptable
choice" (or two choices in this case).

It's great that these results are getting analyzed by people who do not
drink the FairVote kool-aid.

In haste,
Richard Fobes
"The VoteFair guy"
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
Based on the most recent analysis of these numbers that I saw, Leno was
indeed the Condorcet winner, and if Breed were to beat Leno in the final
round, she would then necessarily be the Condorcet winner. The
probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so far.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor. however, i think the news media made it clear
that the race was really gonna be between Leno, Breed, and Kim, so
these fringe voters might have a chance to insincerely mark either
Leno or Breed as their 3rd choice and betray their *true* third
choice and, in doing so, have an effect in the final round.
ignoring the problem of only 3 ranking levels, it is not possible
that London Breed is the Condorcet Winner (a.k.a. "pairwise
champion"). it might be the case that Mark Leno or Jane Kim is the
Condorcet Winner and if the latter is the case, this is another real
indictment against STV or IRV as a method of tallying RCV. and your
reverse namesake, FairVote, is partially (or mostly) to blame.
i wonder if the City of SF has a file of all of the cast and scanned
ballots and the full ranking for each. if so, and if they release
it to the public, we can investigate if there is a Condorcet Winner
and if that CW is or is not Mark Leno. this would be interesting.
L8r,
r b-j
Post by VoteFair
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by VoteFair
- insinuating it is not fair. I was quite bothered to have a Dem
in a
Post by VoteFair
progressive city insinuate that first past the post is more
fair. ...
Post by VoteFair
This remark does not imply support for first past the post (FPTP,
a.k.a
Post by VoteFair
plurality counting).
There are other ways to count the preference marks on "ranked-choice"
ballots. In particular, pairwise counting could be used instead of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had been
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots. Unmarked
choices are an indication that the choices are equally disliked. And
multiple candidates being marked at the same preference level is
also no
Post by VoteFair
problem.
In other words, the ballots contain enough information that they
can be
Post by VoteFair
counted in other ways, besides instant-runoff counting. Those
alternate
Post by VoteFair
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the vote
currently stands, 144 votes separate the top two candidates
(<0.1%) and
Post by VoteFair
over 16,000 ballots have been exhausted (all 3 choices eliminated).
About 9% of voters have been removed from the pool. It is very
possible
Post by VoteFair
that the result would have shifted if they had the opportunity to
rank a
Post by VoteFair
4th candidate, and therefore, it is possible that we won’t elect the
person who truly represents the majority. Is this common? This is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being
asked if
Post by VoteFair
RCV is fair? She stated “This is the system we are working with.
That’s
Post by VoteFair
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by VoteFair
- insinuating it is not fair. I was quite bothered to have a Dem in a
progressive city insinuate that first past the post is more fair. It
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also
presents
Post by VoteFair
herself as a minority candidate and it is my understanding that RCV
gives minority candidates better chances and causes all
candidates to be
Post by VoteFair
more likely to campaign to minority communities. Am I mistaken? Are
there any legitimate arguments that FPTP can be more fair? Thoughts?
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em for
list info
Post by VoteFair
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts
p :: 617.863.0746 <tel:617.863.0746>
w :: voterchoicema.org <http://voterchoicema.org/>
:: Follow us on Facebook
<https://www.facebook.com/voterchoicema> and Twitter
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
Greg Dennis
2018-06-10 01:33:15 UTC
Permalink
Post by VoteFair
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Actually, all the data you need is available from that page. The "Ballot
Image" file will give you the full cast vote record of every individual
ballot, and the "Master Lookup" is the legend that tells you what each
number means. If you have trouble interpreting the numbers, just ping me!
Post by VoteFair
Post by Greg Dennis
https://sfelections.sfgov.org/june-5-2018-election-results-d
etailed-reports
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Converting instant-runoff counts into pairwise counts might be (probably
is?) possible, but I don't have time to do that analysis.
Post by Greg Dennis
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is
rare when the number of ballots exceeds a few hundred.
Post by Greg Dennis
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots had
this issue. Right?
That's not a big number, but a fair counting method -- such as pairwise
counting -- would not have to discard any ballots.
The bigger number is "under votes" and admittedly pairwise counting cannot
compensate for a voter saying "here is the only acceptable choice" (or two
choices in this case).
It's great that these results are getting analyzed by people who do not
drink the FairVote kool-aid.
In haste,
Richard Fobes
"The VoteFair guy"
Post by Greg Dennis
https://sfelections.sfgov.org/june-5-2018-election-results-d
etailed-reports
Based on the most recent analysis of these numbers that I saw, Leno was
indeed the Condorcet winner, and if Breed were to beat Leno in the final
round, she would then necessarily be the Condorcet winner. The
probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so far.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor. however, i think the news media made it clear
that the race was really gonna be between Leno, Breed, and Kim, so
these fringe voters might have a chance to insincerely mark either
Leno or Breed as their 3rd choice and betray their *true* third
choice and, in doing so, have an effect in the final round.
ignoring the problem of only 3 ranking levels, it is not possible
that London Breed is the Condorcet Winner (a.k.a. "pairwise
champion"). it might be the case that Mark Leno or Jane Kim is the
Condorcet Winner and if the latter is the case, this is another real
indictment against STV or IRV as a method of tallying RCV. and your
reverse namesake, FairVote, is partially (or mostly) to blame.
i wonder if the City of SF has a file of all of the cast and scanned
ballots and the full ranking for each. if so, and if they release
it to the public, we can investigate if there is a Condorcet Winner
and if that CW is or is not Mark Leno. this would be interesting.
L8r,
r b-j
Post by VoteFair
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by VoteFair
- insinuating it is not fair. I was quite bothered to have a Dem
in a
Post by VoteFair
progressive city insinuate that first past the post is more
fair. ...
Post by VoteFair
This remark does not imply support for first past the post (FPTP,
a.k.a
Post by VoteFair
plurality counting).
There are other ways to count the preference marks on
"ranked-choice"
Post by VoteFair
ballots. In particular, pairwise counting could be used instead of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had
been
Post by VoteFair
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots. Unmarked
choices are an indication that the choices are equally disliked. And
multiple candidates being marked at the same preference level is
also no
Post by VoteFair
problem.
In other words, the ballots contain enough information that they
can be
Post by VoteFair
counted in other ways, besides instant-runoff counting. Those
alternate
Post by VoteFair
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the
vote
Post by VoteFair
currently stands, 144 votes separate the top two candidates
(<0.1%) and
Post by VoteFair
over 16,000 ballots have been exhausted (all 3 choices eliminated).
About 9% of voters have been removed from the pool. It is very
possible
Post by VoteFair
that the result would have shifted if they had the opportunity to
rank a
Post by VoteFair
4th candidate, and therefore, it is possible that we won’t elect
the
Post by VoteFair
person who truly represents the majority. Is this common? This is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being
asked if
Post by VoteFair
RCV is fair? She stated “This is the system we are working with.
That’s
Post by VoteFair
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by VoteFair
- insinuating it is not fair. I was quite bothered to have a Dem
in a
Post by VoteFair
progressive city insinuate that first past the post is more fair.
It
Post by VoteFair
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also
presents
Post by VoteFair
herself as a minority candidate and it is my understanding that RCV
gives minority candidates better chances and causes all
candidates to be
Post by VoteFair
more likely to campaign to minority communities. Am I mistaken? Are
there any legitimate arguments that FPTP can be more fair?
Thoughts?
Post by VoteFair
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em for
list info
Post by VoteFair
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts
p :: 617.863.0746 <tel:617.863.0746>
w :: voterchoicema.org <http://voterchoicema.org/>
:: Follow us on Facebook
<https://www.facebook.com/voterchoicema> and Twitter
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
Brian Olson
2018-06-10 13:46:44 UTC
Permalink
I processed the latest data (2018-06-09) and posted the results of the SF
Mayor election using a few algorithms:
https://bolson.org/~bolson/2018/SF_Mayor_20180605.html

The Condorcet win is now 97436 to 91740 for Leno over Breed.
The IRV final round is still just 94783 to 94393.


I'm using my software posted at https://github.com/brianolson/voteutil

commands (needs maven installed for compiling Java, and needs Python3):

curl -O
http://www.sfelections.org/results/20180605/data/20180609/20180609_ballotimage.txt
curl -O
http://www.sfelections.org/results/20180605/data/20180609/20180609_masterlookup.txt

(mkdir -p ~/psrc && cd ~/psrc && git clone
https://github.com/brianolson/voteutil.git && cd ~/psrc/voteutil/java &&
mvn package)
python3 ~/psrc/voteutil/python/rcvToNameEq.py -m 20180609_masterlookup.txt
-b 20180609_ballotimage.txt -o 20180609_%s.nameq
java -jar ~/psrc/voteutil/java/target/voteutil-1.0.0.jar --rankings
--full-html --explain -i 20180609_Mayor.nameq >/tmp/a.html
Post by VoteFair
I quickly looked at the vote data and saw that lots of ballots are
Post by VoteFair
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Actually, all the data you need is available from that page. The "Ballot
Image" file will give you the full cast vote record of every individual
ballot, and the "Master Lookup" is the legend that tells you what each
number means. If you have trouble interpreting the numbers, just ping me!
Post by VoteFair
Post by Greg Dennis
https://sfelections.sfgov.org/june-5-2018-election-results-d
etailed-reports
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Converting instant-runoff counts into pairwise counts might be (probably
is?) possible, but I don't have time to do that analysis.
Post by Greg Dennis
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is
rare when the number of ballots exceeds a few hundred.
Post by Greg Dennis
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots
had this issue. Right?
That's not a big number, but a fair counting method -- such as pairwise
counting -- would not have to discard any ballots.
The bigger number is "under votes" and admittedly pairwise counting
cannot compensate for a voter saying "here is the only acceptable choice"
(or two choices in this case).
It's great that these results are getting analyzed by people who do not
drink the FairVote kool-aid.
In haste,
Richard Fobes
"The VoteFair guy"
Post by Greg Dennis
https://sfelections.sfgov.org/june-5-2018-election-results-d
etailed-reports
Based on the most recent analysis of these numbers that I saw, Leno was
indeed the Condorcet winner, and if Breed were to beat Leno in the final
round, she would then necessarily be the Condorcet winner. The
probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so far.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor. however, i think the news media made it clear
that the race was really gonna be between Leno, Breed, and Kim, so
these fringe voters might have a chance to insincerely mark either
Leno or Breed as their 3rd choice and betray their *true* third
choice and, in doing so, have an effect in the final round.
ignoring the problem of only 3 ranking levels, it is not possible
that London Breed is the Condorcet Winner (a.k.a. "pairwise
champion"). it might be the case that Mark Leno or Jane Kim is the
Condorcet Winner and if the latter is the case, this is another real
indictment against STV or IRV as a method of tallying RCV. and your
reverse namesake, FairVote, is partially (or mostly) to blame.
i wonder if the City of SF has a file of all of the cast and scanned
ballots and the full ranking for each. if so, and if they release
it to the public, we can investigate if there is a Condorcet Winner
and if that CW is or is not Mark Leno. this would be interesting.
L8r,
r b-j
Post by VoteFair
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by VoteFair
- insinuating it is not fair. I was quite bothered to have a Dem
in a
Post by VoteFair
progressive city insinuate that first past the post is more
fair. ...
Post by VoteFair
This remark does not imply support for first past the post (FPTP,
a.k.a
Post by VoteFair
plurality counting).
There are other ways to count the preference marks on
"ranked-choice"
Post by VoteFair
ballots. In particular, pairwise counting could be used instead of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had
been
Post by VoteFair
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots.
Unmarked
Post by VoteFair
choices are an indication that the choices are equally disliked.
And
Post by VoteFair
multiple candidates being marked at the same preference level is
also no
Post by VoteFair
problem.
In other words, the ballots contain enough information that they
can be
Post by VoteFair
counted in other ways, besides instant-runoff counting. Those
alternate
Post by VoteFair
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the
vote
Post by VoteFair
currently stands, 144 votes separate the top two candidates
(<0.1%) and
Post by VoteFair
over 16,000 ballots have been exhausted (all 3 choices
eliminated).
Post by VoteFair
About 9% of voters have been removed from the pool. It is very
possible
Post by VoteFair
that the result would have shifted if they had the opportunity to
rank a
Post by VoteFair
4th candidate, and therefore, it is possible that we won’t elect
the
Post by VoteFair
person who truly represents the majority. Is this common? This is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being
asked if
Post by VoteFair
RCV is fair? She stated “This is the system we are working with.
That’s
Post by VoteFair
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by VoteFair
- insinuating it is not fair. I was quite bothered to have a Dem
in a
Post by VoteFair
progressive city insinuate that first past the post is more fair.
It
Post by VoteFair
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also
presents
Post by VoteFair
herself as a minority candidate and it is my understanding that
RCV
Post by VoteFair
gives minority candidates better chances and causes all
candidates to be
Post by VoteFair
more likely to campaign to minority communities. Am I mistaken?
Are
Post by VoteFair
there any legitimate arguments that FPTP can be more fair?
Thoughts?
Post by VoteFair
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em for
list info
Post by VoteFair
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts
p :: 617.863.0746 <tel:617.863.0746>
w :: voterchoicema.org <http://voterchoicema.org/>
:: Follow us on Facebook
<https://www.facebook.com/voterchoicema> and Twitter
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
Greg Dennis
2018-06-10 14:08:28 UTC
Permalink
Brian, how is it possible that those differ? Since all the other candidates
are eliminated in the final round, shouldn't that necessarily be the same
as the pairwise contest between those two?
Post by Brian Olson
I processed the latest data (2018-06-09) and posted the results of the SF
https://bolson.org/~bolson/2018/SF_Mayor_20180605.html
The Condorcet win is now 97436 to 91740 for Leno over Breed.
The IRV final round is still just 94783 to 94393.
I'm using my software posted at https://github.com/brianolson/voteutil
curl -O
http://www.sfelections.org/results/20180605/data/20180609/20180609_ballotimage.txt
curl -O
http://www.sfelections.org/results/20180605/data/20180609/20180609_masterlookup.txt
(mkdir -p ~/psrc && cd ~/psrc && git clone
https://github.com/brianolson/voteutil.git && cd ~/psrc/voteutil/java &&
mvn package)
python3 ~/psrc/voteutil/python/rcvToNameEq.py -m 20180609_masterlookup.txt
-b 20180609_ballotimage.txt -o 20180609_%s.nameq
java -jar ~/psrc/voteutil/java/target/voteutil-1.0.0.jar --rankings
--full-html --explain -i 20180609_Mayor.nameq >/tmp/a.html
Post by VoteFair
I quickly looked at the vote data and saw that lots of ballots are
Post by VoteFair
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Actually, all the data you need is available from that page. The "Ballot
Image" file will give you the full cast vote record of every individual
ballot, and the "Master Lookup" is the legend that tells you what each
number means. If you have trouble interpreting the numbers, just ping me!
Post by VoteFair
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Converting instant-runoff counts into pairwise counts might be (probably
is?) possible, but I don't have time to do that analysis.
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is
rare when the number of ballots exceeds a few hundred.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice
of
choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots
had this issue. Right?
That's not a big number, but a fair counting method -- such as pairwise
counting -- would not have to discard any ballots.
The bigger number is "under votes" and admittedly pairwise counting
cannot compensate for a voter saying "here is the only acceptable choice"
(or two choices in this case).
It's great that these results are getting analyzed by people who do not
drink the FairVote kool-aid.
In haste,
Richard Fobes
"The VoteFair guy"
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
Based on the most recent analysis of these numbers that I saw, Leno was
indeed the Condorcet winner, and if Breed were to beat Leno in the final
round, she would then necessarily be the Condorcet winner. The
probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so far.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor. however, i think the news media made it clear
that the race was really gonna be between Leno, Breed, and Kim, so
these fringe voters might have a chance to insincerely mark either
Leno or Breed as their 3rd choice and betray their *true* third
choice and, in doing so, have an effect in the final round.
ignoring the problem of only 3 ranking levels, it is not possible
that London Breed is the Condorcet Winner (a.k.a. "pairwise
champion"). it might be the case that Mark Leno or Jane Kim is the
Condorcet Winner and if the latter is the case, this is another real
indictment against STV or IRV as a method of tallying RCV. and your
reverse namesake, FairVote, is partially (or mostly) to blame.
i wonder if the City of SF has a file of all of the cast and scanned
ballots and the full ranking for each. if so, and if they release
it to the public, we can investigate if there is a Condorcet Winner
and if that CW is or is not Mark Leno. this would be interesting.
L8r,
r b-j
Post by VoteFair
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by VoteFair
- insinuating it is not fair. I was quite bothered to have a Dem
in a
Post by VoteFair
progressive city insinuate that first past the post is more
fair. ...
Post by VoteFair
This remark does not imply support for first past the post (FPTP,
a.k.a
Post by VoteFair
plurality counting).
There are other ways to count the preference marks on
"ranked-choice"
Post by VoteFair
ballots. In particular, pairwise counting could be used instead of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had
been
Post by VoteFair
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots.
Unmarked
Post by VoteFair
choices are an indication that the choices are equally disliked.
And
Post by VoteFair
multiple candidates being marked at the same preference level is
also no
Post by VoteFair
problem.
In other words, the ballots contain enough information that they
can be
Post by VoteFair
counted in other ways, besides instant-runoff counting. Those
alternate
Post by VoteFair
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the
vote
Post by VoteFair
currently stands, 144 votes separate the top two candidates
(<0.1%) and
Post by VoteFair
over 16,000 ballots have been exhausted (all 3 choices
eliminated).
Post by VoteFair
About 9% of voters have been removed from the pool. It is very
possible
Post by VoteFair
that the result would have shifted if they had the opportunity to
rank a
Post by VoteFair
4th candidate, and therefore, it is possible that we won’t elect
the
Post by VoteFair
person who truly represents the majority. Is this common? This is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being
asked if
Post by VoteFair
RCV is fair? She stated “This is the system we are working with.
That’s
Post by VoteFair
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by VoteFair
- insinuating it is not fair. I was quite bothered to have a Dem
in a
Post by VoteFair
progressive city insinuate that first past the post is more
fair. It
Post by VoteFair
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also
presents
Post by VoteFair
herself as a minority candidate and it is my understanding that
RCV
Post by VoteFair
gives minority candidates better chances and causes all
candidates to be
Post by VoteFair
more likely to campaign to minority communities. Am I mistaken?
Are
Post by VoteFair
there any legitimate arguments that FPTP can be more fair?
Thoughts?
Post by VoteFair
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em for
list info
Post by VoteFair
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts
p :: 617.863.0746 <tel:617.863.0746>
w :: voterchoicema.org <http://voterchoicema.org/>
:: Follow us on Facebook
<https://www.facebook.com/voterchoicema> and Twitter
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
Brian Olson
2018-06-10 14:46:31 UTC
Permalink
Ok, a few lines of Python poking the raw data shows I must have some bug in
my Condorcet implementation. Digging into that...
Post by Greg Dennis
Brian, how is it possible that those differ? Since all the other
candidates are eliminated in the final round, shouldn't that necessarily be
the same as the pairwise contest between those two?
Post by Brian Olson
I processed the latest data (2018-06-09) and posted the results of the SF
https://bolson.org/~bolson/2018/SF_Mayor_20180605.html
The Condorcet win is now 97436 to 91740 for Leno over Breed.
The IRV final round is still just 94783 to 94393.
I'm using my software posted at https://github.com/brianolson/voteutil
curl -O http://www.sfelections.org/results/20180605/data/
20180609/20180609_ballotimage.txt
curl -O http://www.sfelections.org/results/20180605/data/
20180609/20180609_masterlookup.txt
(mkdir -p ~/psrc && cd ~/psrc && git clone https://github.com/brianolson/
voteutil.git && cd ~/psrc/voteutil/java && mvn package)
python3 ~/psrc/voteutil/python/rcvToNameEq.py -m
20180609_masterlookup.txt -b 20180609_ballotimage.txt -o 20180609_%s.nameq
java -jar ~/psrc/voteutil/java/target/voteutil-1.0.0.jar --rankings
--full-html --explain -i 20180609_Mayor.nameq >/tmp/a.html
On Sat, Jun 9, 2018 at 9:33 PM, Greg Dennis <
Post by VoteFair
I quickly looked at the vote data and saw that lots of ballots are
Post by VoteFair
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Actually, all the data you need is available from that page. The "Ballot
Image" file will give you the full cast vote record of every individual
ballot, and the "Master Lookup" is the legend that tells you what each
number means. If you have trouble interpreting the numbers, just ping me!
Post by VoteFair
Post by Greg Dennis
https://sfelections.sfgov.org/june-5-2018-election-results-
detailed-reports
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Converting instant-runoff counts into pairwise counts might be
(probably is?) possible, but I don't have time to do that analysis.
Post by Greg Dennis
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is
rare when the number of ballots exceeds a few hundred.
Post by Greg Dennis
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice
of
Post by Greg Dennis
choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots
had this issue. Right?
That's not a big number, but a fair counting method -- such as pairwise
counting -- would not have to discard any ballots.
The bigger number is "under votes" and admittedly pairwise counting
cannot compensate for a voter saying "here is the only acceptable choice"
(or two choices in this case).
It's great that these results are getting analyzed by people who do not
drink the FairVote kool-aid.
In haste,
Richard Fobes
"The VoteFair guy"
Post by Greg Dennis
https://sfelections.sfgov.org/june-5-2018-election-results-
detailed-reports
Based on the most recent analysis of these numbers that I saw, Leno was
indeed the Condorcet winner, and if Breed were to beat Leno in the final
round, she would then necessarily be the Condorcet winner. The
probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so far.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor. however, i think the news media made it clear
that the race was really gonna be between Leno, Breed, and Kim, so
these fringe voters might have a chance to insincerely mark either
Leno or Breed as their 3rd choice and betray their *true* third
choice and, in doing so, have an effect in the final round.
ignoring the problem of only 3 ranking levels, it is not possible
that London Breed is the Condorcet Winner (a.k.a. "pairwise
champion"). it might be the case that Mark Leno or Jane Kim is the
Condorcet Winner and if the latter is the case, this is another real
indictment against STV or IRV as a method of tallying RCV. and your
reverse namesake, FairVote, is partially (or mostly) to blame.
i wonder if the City of SF has a file of all of the cast and scanned
ballots and the full ranking for each. if so, and if they release
it to the public, we can investigate if there is a Condorcet Winner
and if that CW is or is not Mark Leno. this would be interesting.
L8r,
r b-j
Post by VoteFair
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by VoteFair
- insinuating it is not fair. I was quite bothered to have a
Dem
in a
Post by VoteFair
progressive city insinuate that first past the post is more
fair. ...
Post by VoteFair
This remark does not imply support for first past the post (FPTP,
a.k.a
Post by VoteFair
plurality counting).
There are other ways to count the preference marks on
"ranked-choice"
Post by VoteFair
ballots. In particular, pairwise counting could be used instead
of
Post by VoteFair
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had
been
Post by VoteFair
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots.
Unmarked
Post by VoteFair
choices are an indication that the choices are equally disliked.
And
Post by VoteFair
multiple candidates being marked at the same preference level is
also no
Post by VoteFair
problem.
In other words, the ballots contain enough information that they
can be
Post by VoteFair
counted in other ways, besides instant-runoff counting. Those
alternate
Post by VoteFair
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the
vote
Post by VoteFair
currently stands, 144 votes separate the top two candidates
(<0.1%) and
Post by VoteFair
over 16,000 ballots have been exhausted (all 3 choices
eliminated).
Post by VoteFair
About 9% of voters have been removed from the pool. It is very
possible
Post by VoteFair
that the result would have shifted if they had the opportunity
to
rank a
Post by VoteFair
4th candidate, and therefore, it is possible that we won’t
elect the
Post by VoteFair
person who truly represents the majority. Is this common? This
is
Post by VoteFair
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being
asked if
Post by VoteFair
RCV is fair? She stated “This is the system we are working with.
That’s
Post by VoteFair
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by VoteFair
- insinuating it is not fair. I was quite bothered to have a
Dem in a
Post by VoteFair
progressive city insinuate that first past the post is more
fair. It
Post by VoteFair
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also
presents
Post by VoteFair
herself as a minority candidate and it is my understanding that
RCV
Post by VoteFair
gives minority candidates better chances and causes all
candidates to be
Post by VoteFair
more likely to campaign to minority communities. Am I mistaken?
Are
Post by VoteFair
there any legitimate arguments that FPTP can be more fair?
Thoughts?
Post by VoteFair
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em
for
list info
Post by VoteFair
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts
voterchoicema.org>
p :: 617.863.0746 <tel:617.863.0746>
w :: voterchoicema.org <http://voterchoicema.org/>
:: Follow us on Facebook
<https://www.facebook.com/voterchoicema> and Twitter
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
robert bristow-johnson
2018-06-09 05:12:54 UTC
Permalink
I wonder if what happened in San Francisco last Tuesday is what happened to "the San Francisco of New England" 9 years ago?
No way to tell without getting individual ballot data for 160K ballots.
So, specifically, I wonder if Jane Kim might have been the Condorcet Winner and was eliminated prematurely and might have beaten both London Breed and Mark Leno on a mano-a-mano match?  Jane Kim might want to have that investigated.
It all depends on who the surviving 2nd-ranked votes were on the Breed and Leno ballots and that remains opaque to us unless the full ballot data is released like the vermont city of Burlington did in 2009.
Maybe Jane Kim is SF's Andy Montroll.
--
r b-j 

--r b-j                     ***@audioimagination.com
"Imagination is more important than knowledge."




-------- Original message --------
From: Christopher Colosi <***@gmail.com>
Date: 6/8/2018 6:24 PM (GMT-08:00)
To: election-***@lists.electorama.com
Subject: [EM] RCV in SF Mayoral election

Curious to hear people’s thoughts on some issues.
1. May not elect majority candidateIn SF, we restrict to 3 choices to simplify the process.  As the vote currently stands, 144 votes separate the top two candidates (<0.1%) and over 16,000 ballots have been exhausted (all 3 choices eliminated).  About 9% of voters have been removed from the pool.  It is very possible that the result would have shifted if they had the opportunity to rank a 4th candidate, and therefore, it is possible that we won’t elect the person who truly represents the majority.  Is this common?  This is probably an abnormally close race.  Thoughts?
2. What are your thoughts on London Breed’s response to being asked if RCV is fair?  She stated “This is the system we are working with. That’s a discussion we can have at a later time. For now, we’re stuck with it.” - insinuating it is not fair.  I was quite bothered to have a Dem in a progressive city insinuate that first past the post is more fair.  It also felt divisive.  If Leno wins, will her supporters feel that democracy prevailed, or that the election was stolen?  She also presents herself as a minority candidate and it is my understanding that RCV gives minority candidates better chances and causes all candidates to be more likely to campaign to minority communities.  Am I mistaken?  Are there any legitimate arguments that FPTP can be more fair?  Thoughts?
Regards,—Chris
Bob Richard
2018-06-09 23:11:49 UTC
Permalink
In SF, we restrict to 3 choices to simplify the process. ...
... The ballot shall allow voters to rank a number of choices in order
of preference equal to the total number of candidates for each office;
provided, however, if the voting system, vote tabulation system or
similar or related equipment used by the City and County cannot
feasibly accommodate choices equal to the total number of candidates
running for each office, then the Director of Elections may limit the
number of choices a voter may rank to no fewer than three....
The city is currently working on a design for new equipment, which will
almost certainly allow more than three choices.

That said, there is nonetheless some truth in Chrisopher's statement.
Those opponents of RCV who claim it is "too complicated" also claim to
believe that limiting the voter to three choices simplifies her task.
The discussion of the effects of truncation in this thread -- including
on strategic voting -- proves these opponents wrong on this point.

--Bob Richard
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process.  As the vote
currently stands, 144 votes separate the top two candidates (<0.1%)
and over 16,000 ballots have been exhausted (all 3 choices
eliminated).  About 9% of voters have been removed from the pool.  It
is very possible that the result would have shifted if they had the
opportunity to rank a 4th candidate, and therefore, it is possible
that we won’t elect the person who truly represents the majority.  Is
this common?  This is probably an abnormally close race.  Thoughts?
2. What are your thoughts on London Breed’s response to being asked if
RCV is fair?  She stated “This is the system we are working with.
That’s a discussion we can have at a later time. For now, we’re stuck
with it.” - insinuating it is not fair.  I was quite bothered to have
a Dem in a progressive city insinuate that first past the post is more
fair.  It also felt divisive.  If Leno wins, will her supporters feel
that democracy prevailed, or that the election was stolen? She also
presents herself as a minority candidate and it is my understanding
that RCV gives minority candidates better chances and causes all
candidates to be more likely to campaign to minority communities.  Am
I mistaken?  Are there any legitimate arguments that FPTP can be more
fair?  Thoughts?
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em for list info
robert bristow-johnson
2018-06-11 02:23:48 UTC
Permalink
Thank you Brian for doing this.  Looks like Leno is the legit winner and STV and Condorcet agree.
However, the Leno-Breed pair wise tally in the defeat matrix should be exactly the same as the STV final round result.  So something is wrong somewhere.
But thanks for doing this.  I was about to code up a MATLAB program to parse and count this thing.

--r b-j                     ***@audioimagination.com
"Imagination is more important than knowledge."




-------- Original message --------
From: Brian Olson <***@bolson.org>
Date: 6/10/2018 7:46 AM (GMT-08:00)
To: EM <election-***@lists.electorama.com>
Subject: Re: [EM] RCV in SF Mayoral election

Ok, a few lines of Python poking the raw data shows I must have some bug in my Condorcet implementation. Digging into that...
On Sun, Jun 10, 2018 at 10:08 AM, Greg Dennis <***@voterchoicema.org> wrote:
Brian, how is it possible that those differ? Since all the other candidates are eliminated in the final round, shouldn't that necessarily be the same as the pairwise contest between those two?

On Sun, Jun 10, 2018, 9:46 AM Brian Olson <***@bolson.org> wrote:
I processed the latest data (2018-06-09) and posted the results of the SF Mayor election using a few algorithms:https://bolson.org/~bolson/2018/SF_Mayor_20180605.html

The Condorcet win is now 97436 to 91740 for Leno over Breed.The IRV final round is still just 94783 to 94393.

I'm using my software posted at https://github.com/brianolson/voteutil
commands (needs maven installed for compiling Java, and needs Python3):
curl -O http://www.sfelections.org/results/20180605/data/20180609/20180609_ballotimage.txtcurl -O http://www.sfelections.org/results/20180605/data/20180609/20180609_masterlookup.txt
(mkdir -p ~/psrc && cd ~/psrc && git clone https://github.com/brianolson/voteutil.git && cd ~/psrc/voteutil/java && mvn package)python3 ~/psrc/voteutil/python/rcvToNameEq.py -m 20180609_masterlookup.txt -b 20180609_ballotimage.txt -o 20180609_%s.nameqjava -jar ~/psrc/voteutil/java/target/voteutil-1.0.0.jar --rankings --full-html --explain -i 20180609_Mayor.nameq >/tmp/a.html

On Sat, Jun 9, 2018 at 9:33 PM, Greg Dennis <***@voterchoicema.org> wrote:
I quickly looked at the vote data and saw that lots of ballots are categorized as "Exhausted by Over Votes" and "Under Votes," but there is no data indicating exactly how those ballots were marked, so we lack enough information to be sure of final results.
Actually, all the data you need is available from that page. The "Ballot Image" file will give you the full cast vote record of every individual ballot, and the "Master Lookup" is the legend that tells you what each number means. If you have trouble interpreting the numbers, just ping me!
Post by Greg Dennis
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
I quickly looked at the vote data and saw that lots of ballots are categorized as "Exhausted by Over Votes" and "Under Votes," but there is no data indicating exactly how those ballots were marked, so we lack enough information to be sure of final results.



Converting instant-runoff counts into pairwise counts might be (probably is?) possible, but I don't have time to do that analysis.
Post by Greg Dennis
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only
case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is rare when the number of ballots exceeds a few hundred.
Post by Greg Dennis
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
     the limitation to only three levels of ranking is a problem.  if
     someone ranked all three levels and none of the candidates ranked
     were either London Breed nor Mark Leno, that voter was effectively
     "disenfranchised" by being unable to weigh in on the final choice of
     choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots had this issue.  Right?



That's not a big number, but a fair counting method -- such as pairwise counting -- would not have to discard any ballots.



The bigger number is "under votes" and admittedly pairwise counting cannot compensate for a voter saying "here is the only acceptable choice" (or two choices in this case).



It's great that these results are getting analyzed by people who do not drink the FairVote kool-aid.



In haste,

Richard Fobes

"The VoteFair guy"





On 6/9/2018 6:25 AM, Greg Dennis wrote:


San Francisco always make the cast vote record public:

https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports



Based on the most recent analysis of these numbers that I saw, Leno was

indeed the Condorcet winner, and if Breed were to beat Leno in the final

round, she would then necessarily be the Condorcet winner. The

probability of IRV not elected the Condorcet winner appears to be

exceedingly low in practice. We're up to about ~200 IRV elections

conducted nationwide since 2004 and Burlington 2009 is the only case so far.



On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson

<***@audioimagination.com <mailto:***@audioimagination.com>> wrote:





    Richard, a few points:



    the limitation to only three levels of ranking is a problem.  if

    someone ranked all three levels and none of the candidates ranked

    were either London Breed nor Mark Leno, that voter was effectively

    "disenfranchised" by being unable to weigh in on the final choice of

    choosing the mayor.  however, i think the news media made it clear

    that the race was really gonna be between Leno, Breed, and Kim, so

    these fringe voters might have a chance to insincerely mark either

    Leno or Breed as their 3rd choice and betray their *true* third

    choice and, in doing so, have an effect in the final round.



    ignoring the problem of only 3 ranking levels, it is not possible

    that London Breed is the Condorcet Winner (a.k.a. "pairwise

    champion").  it might be the case that Mark Leno or Jane Kim is the

    Condorcet Winner and if the latter is the case, this is another real

    indictment against STV or IRV as a method of tallying RCV.  and your

    reverse namesake, FairVote, is partially (or mostly) to blame.



    i wonder if the City of SF has a file of all of the cast and scanned

    ballots and the full ranking for each.  if so, and if they release

    it to the public, we can investigate if there is a Condorcet Winner

    and if that CW is or is not Mark Leno.  this would be interesting.



    L8r,







    r b-j







    > On 6/8/2018 6:24 PM, Christopher Colosi wrote:

    > > ... She stated “This is the system we are working with. That’s

    > > a discussion we can have at a later time. For now, we’re stuck

    with it.”

    > > - insinuating it is not fair. I was quite bothered to have a Dem

    in a

    > > progressive city insinuate that first past the post is more

    fair. ...

    >

    > This remark does not imply support for first past the post (FPTP,

    a.k.a

    > plurality counting).

    >

    > There are other ways to count the preference marks on "ranked-choice"

    > ballots. In particular, pairwise counting could be used instead of

    > instant-runoff counting, and that is fairer than FPTP.

    >

    > > 1. May not elect majority candidate

    > > ...

    > > Is this common? This is

    > > probably an abnormally close race. Thoughts?

    >

    > I doubt the voters would regard this as a close race if they had been

    > able to fully rank all the choices. The 3-choice limitation is

    > simplistic, and complicates the counting.

    >

    > Pairwise counting does not result in any exhausted ballots. Unmarked

    > choices are an indication that the choices are equally disliked. And

    > multiple candidates being marked at the same preference level is

    also no

    > problem.

    >

    > In other words, the ballots contain enough information that they

    can be

    > counted in other ways, besides instant-runoff counting. Those

    alternate

    > counting methods could reveal a clearer outcome.

    >

    > In haste,

    > Richard Fobes

    >

    >

    > On 6/8/2018 6:24 PM, Christopher Colosi wrote:

    >> Curious to hear people’s thoughts on some issues.

    >>

    >> 1. May not elect majority candidate

    >> In SF, we restrict to 3 choices to simplify the process. As the vote

    >> currently stands, 144 votes separate the top two candidates

    (<0.1%) and

    >> over 16,000 ballots have been exhausted (all 3 choices eliminated).

    >> About 9% of voters have been removed from the pool. It is very

    possible

    >> that the result would have shifted if they had the opportunity to

    rank a

    >> 4th candidate, and therefore, it is possible that we won’t elect the

    >> person who truly represents the majority. Is this common? This is

    >> probably an abnormally close race. Thoughts?

    >>

    >> 2. What are your thoughts on London Breed’s response to being

    asked if

    >> RCV is fair? She stated “This is the system we are working with.

    That’s

    >> a discussion we can have at a later time. For now, we’re stuck

    with it.”

    >> - insinuating it is not fair. I was quite bothered to have a Dem in a

    >> progressive city insinuate that first past the post is more fair. It

    >> also felt divisive. If Leno wins, will her supporters feel that

    >> democracy prevailed, or that the election was stolen? She also

    presents

    >> herself as a minority candidate and it is my understanding that RCV

    >> gives minority candidates better chances and causes all

    candidates to be

    >> more likely to campaign to minority communities. Am I mistaken? Are

    >> there any legitimate arguments that FPTP can be more fair? Thoughts?

    >>

    >> Regards,

    >> —Chris

    >>

    >>

    >>

    >> ----

    >> Election-Methods mailing list - see http://electorama.com/em for

    list info

    >>

    > ----

    > Election-Methods mailing list - see http://electorama.com/em for

    list info

    >



    --



    r b-j                         ***@audioimagination.com

    <mailto:***@audioimagination.com>



    "Imagination is more important than knowledge."



















    ----

    Election-Methods mailing list - see http://electorama.com/em for

    list info









--

*Greg Dennis, Ph.D. :: Policy Director*

Voter Choice Massachusetts



e :: ***@voterchoicema.org <mailto:***@voterchoicema.org>

p :: 617.863.0746 <tel:617.863.0746>

w :: voterchoicema.org <http://voterchoicema.org/>



:: Follow us on Facebook

<https://www.facebook.com/voterchoicema> and Twitter

<https://twitter.com/voterchoicema> ::





----

Election-Methods mailing list - see http://electorama.com/em for list info




----

Election-Methods mailing list - see http://electorama.com/em for list info


----

Election-Methods mailing list - see http://electorama.com/em for list info




----

Election-Methods mailing list - see http://electorama.com/em for list info
robert bristow-johnson
2018-06-11 05:42:10 UTC
Permalink
 
 
boy this is embarrassing.  i didn't realize they were still counting and that the lead has changed.  looks like London Breed is ahead and will likely win the STV contest.
and, from Brian's initial analysis, it doesn't look like the Condorcet Winner will
be different from the STV winner.  so no Burlington 2009 situation.
L8r,
r b-j


---------------------------- Original Message ----------------------------

Subject: Re: [EM] RCV in SF Mayoral election

From: "robert bristow-johnson" <***@audioimagination.com>

Date: Sun, June 10, 2018 10:23 pm

To: "EM" <election-***@lists.electorama.com>

Cc: "Brian Olson" <***@bolson.org>

--------------------------------------------------------------------------
Thank you Brian for doing this.  Looks like Leno is the legit winner and STV and Condorcet agree.
However, the Leno-Breed pair wise tally in the defeat matrix should be exactly the same as the STV final round result.  So something is wrong somewhere.
But thanks for doing this.  I was about to code up a MATLAB program to parse and count this thing.
"Imagination is more important than knowledge."
-------- Original message --------
Date: 6/10/2018 7:46 AM (GMT-08:00)
Subject: Re: [EM] RCV in SF Mayoral election
Ok, a few lines of Python poking the raw data shows I must have some bug in my Condorcet implementation. Digging into that...
Brian, how is it possible that those differ? Since all the other candidates are eliminated in the final round, shouldn't that necessarily be the same as the pairwise contest between those two?
I processed the latest data (2018-06-09) and posted the results of the SF Mayor election using a few algorithms:https://bolson.org/~bolson/2018/SF_Mayor_20180605.html
The Condorcet win is now 97436 to 91740 for Leno over Breed.The IRV final round is still just 94783 to 94393.
I'm using my software posted at https://github.com/brianolson/voteutil
curl -O http://www.sfelections.org/results/20180605/data/20180609/20180609_ballotimage.txtcurl -O http://www.sfelections.org/results/20180605/data/20180609/20180609_masterlookup.txt
(mkdir -p ~/psrc && cd ~/psrc && git clone https://github.com/brianolson/voteutil.git && cd ~/psrc/voteutil/java && mvn package)python3 ~/psrc/voteutil/python/rcvToNameEq.py -m 20180609_masterlookup.txt -b 20180609_ballotimage.txt -o 20180609_%s.nameqjava -jar
~/psrc/voteutil/java/target/voteutil-1.0.0.jar --rankings --full-html --explain -i 20180609_Mayor.nameq >/tmp/a.html
I quickly looked at the vote data and saw that lots of ballots are categorized as "Exhausted by Over Votes" and "Under Votes," but there is no data indicating exactly how those ballots were marked, so we lack enough information to be sure of final results.
Actually, all the data you need is available from that page. The "Ballot Image" file will give you the full cast vote record of every individual ballot, and the "Master Lookup" is the legend that tells you what each number means. If you have trouble interpreting the numbers,
just ping me!
Post by Greg Dennis
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
I quickly looked at the vote data and saw that lots of ballots are categorized as "Exhausted by Over Votes" and "Under Votes," but there is no data indicating exactly how those ballots were marked, so we lack enough information to be sure of final results.
Converting instant-runoff counts into pairwise counts might be (probably is?) possible, but I don't have time to do that analysis.
Post by Greg Dennis
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only
case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is rare when the number of ballots exceeds a few hundred.
Post by Greg Dennis
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
     the limitation to only three levels of ranking is a problem.  if
     someone ranked all three levels and none of the candidates ranked
     were either London Breed nor Mark Leno, that voter was effectively
     "disenfranchised" by being unable to weigh in on the final choice of
     choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots had this issue.  Right?
That's not a big number, but a fair counting method -- such as pairwise counting -- would not have to discard any ballots.
The bigger number is "under votes" and admittedly pairwise counting cannot compensate for a voter saying "here is the only acceptable choice" (or two choices in this case).
It's great that these results are getting analyzed by people who do not drink the FairVote kool-aid.
In haste,
Richard Fobes
"The VoteFair guy"
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
Based on the most recent analysis of these numbers that I saw, Leno was
indeed the Condorcet winner, and if Breed were to beat Leno in the final
round, she would then necessarily be the Condorcet winner. The
probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so far.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
    the limitation to only three levels of ranking is a problem.  if
    someone ranked all three levels and none of the candidates ranked
    were either London Breed nor Mark Leno, that voter was effectively
    "disenfranchised" by being unable to weigh in on the final choice of
    choosing the mayor.  however, i think the news media made it clear
    that the race was really gonna be between Leno, Breed, and Kim, so
    these fringe voters might have a chance to insincerely mark either
    Leno or Breed as their 3rd choice and betray their *true* third
    choice and, in doing so, have an effect in the final round.
    ignoring the problem of only 3 ranking levels, it is not possible
    that London Breed is the Condorcet Winner (a.k.a. "pairwise
    champion").  it might be the case that Mark Leno or Jane Kim is the
    Condorcet Winner and if the latter is the case, this is another real
    indictment against STV or IRV as a method of tallying RCV.  and your
    reverse namesake, FairVote, is partially (or mostly) to blame.
    i wonder if the City of SF has a file of all of the cast and scanned
    ballots and the full ranking for each.  if so, and if they release
    it to the public, we can investigate if there is a Condorcet Winner
    and if that CW is or is not Mark Leno.  this would be interesting.
    L8r,
    r b-j
    > > ... She stated &ldquo;This is the system we are working with. That&rsquo;s
    > > a discussion we can have at a later time. For now, we&rsquo;re stuck
    with it.&rdquo;
    > > - insinuating it is not fair. I was quite bothered to have a Dem
    in a
    > > progressive city insinuate that first past the post is more
    fair. ...
    >
    > This remark does not imply support for first past the post (FPTP,
    a.k.a
    > plurality counting).
    >
    > There are other ways to count the preference marks on "ranked-choice"
    > ballots. In particular, pairwise counting could be used instead of
    > instant-runoff counting, and that is fairer than FPTP.
    >
    > > 1. May not elect majority candidate
    > > ...
    > > Is this common? This is
    > > probably an abnormally close race. Thoughts?
    >
    > I doubt the voters would regard this as a close race if they had been
    > able to fully rank all the choices. The 3-choice limitation is
    > simplistic, and complicates the counting.
    >
    > Pairwise counting does not result in any exhausted ballots. Unmarked
    > choices are an indication that the choices are equally disliked. And
    > multiple candidates being marked at the same preference level is
    also no
    > problem.
    >
    > In other words, the ballots contain enough information that they
    can be
    > counted in other ways, besides instant-runoff counting. Those
    alternate
    > counting methods could reveal a clearer outcome.
    >
    > In haste,
    > Richard Fobes
    >
    >
    >> Curious to hear people&rsquo;s thoughts on some issues.
    >>
    >> 1. May not elect majority candidate
    >> In SF, we restrict to 3 choices to simplify the process. As the vote
    >> currently stands, 144 votes separate the top two candidates
    (<0.1%) and
    >> over 16,000 ballots have been exhausted (all 3 choices eliminated).
    >> About 9% of voters have been removed from the pool. It is very
    possible
    >> that the result would have shifted if they had the opportunity to
    rank a
    >> 4th candidate, and therefore, it is possible that we won&rsquo;t elect the
    >> person who truly represents the majority. Is this common? This is
    >> probably an abnormally close race. Thoughts?
    >>
    >> 2. What are your thoughts on London Breed&rsquo;s response to being
    asked if
    >> RCV is fair? She stated &ldquo;This is the system we are working with.
    That&rsquo;s
    >> a discussion we can have at a later time. For now, we&rsquo;re stuck
    with it.&rdquo;
    >> - insinuating it is not fair. I was quite bothered to have a Dem in a
    >> progressive city insinuate that first past the post is more fair. It
    >> also felt divisive. If Leno wins, will her supporters feel that
    >> democracy prevailed, or that the election was stolen? She also
    presents
    >> herself as a minority candidate and it is my understanding that RCV
    >> gives minority candidates better chances and causes all
    candidates to be
    >> more likely to campaign to minority communities. Am I mistaken? Are
    >> there any legitimate arguments that FPTP can be more fair? Thoughts?
    >>
    >> Regards,
    >> &mdash;Chris
    >>
    >>
    >>
    >> ----
    >> Election-Methods mailing list - see http://electorama.com/em for
    list info
    >>
    > ----
    > Election-Methods mailing list - see http://electorama.com/em for
    list info
    >
    --
    "Imagination is more important than knowledge."
    ----
    Election-Methods mailing list - see http://electorama.com/em for
    list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts
p :: 617.863.0746 <tel:617.863.0746>
w :: voterchoicema.org <http://voterchoicema.org/>
:: Follow us on Facebook
<https://www.facebook.com/voterchoicema> and Twitter
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
 
 
 


--



r b-j                         ***@audioimagination.com



"Imagination is more important than knowledge."

 
 
 
 
Christopher Colosi
2018-06-11 15:45:37 UTC
Permalink
How is the ballot image file laid out? It seems like a horrible choice for
formatting. Is there any delimiter between ballots, or is this just one
gigantic string of numbers?

On Sun, Jun 10, 2018 at 10:42 PM robert bristow-johnson <
boy this is embarrassing. i didn't realize they were still counting and
that the lead has changed. looks like London Breed is ahead and will
likely win the STV contest.
and, from Brian's initial analysis, it doesn't look like the Condorcet
Winner will be different from the STV winner. so no Burlington 2009
situation.
L8r,
r b-j
---------------------------- Original Message ----------------------------
Subject: Re: [EM] RCV in SF Mayoral election
Date: Sun, June 10, 2018 10:23 pm
--------------------------------------------------------------------------
Thank you Brian for doing this. Looks like Leno is the legit winner and
STV and Condorcet agree.
However, the Leno-Breed pair wise tally in the defeat matrix should be
exactly the same as the STV final round result. So something is wrong
somewhere.
But thanks for doing this. I was about to code up a MATLAB program to
parse and count this thing.
"Imagination is more important than knowledge."
-------- Original message --------
Date: 6/10/2018 7:46 AM (GMT-08:00)
Subject: Re: [EM] RCV in SF Mayoral election
Ok, a few lines of Python poking the raw data shows I must have some bug
in my Condorcet implementation. Digging into that...
On Sun, Jun 10, 2018 at 10:08 AM, Greg Dennis <
Brian, how is it possible that those differ? Since all the other
candidates are eliminated in the final round, shouldn't that necessarily be
the same as the pairwise contest between those two?
I processed the latest data (2018-06-09) and posted the results of the
https://bolson.org/~bolson/2018/SF_Mayor_20180605.html
The Condorcet win is now 97436 to 91740 for Leno over Breed.The IRV
final round is still just 94783 to 94393.
I'm using my software posted at https://github.com/brianolson/voteutil
curl -O
http://www.sfelections.org/results/20180605/data/20180609/20180609_ballotimage.txtcurl
-O
http://www.sfelections.org/results/20180605/data/20180609/20180609_masterlookup.txt
(mkdir -p ~/psrc && cd ~/psrc && git clone
https://github.com/brianolson/voteutil.git && cd ~/psrc/voteutil/java &&
mvn package)python3 ~/psrc/voteutil/python/rcvToNameEq.py -m
20180609_masterlookup.txt -b 20180609_ballotimage.txt -o
20180609_%s.nameqjava -jar ~/psrc/voteutil/java/target/voteutil-1.0.0.jar
--rankings --full-html --explain -i 20180609_Mayor.nameq >/tmp/a.html
On Sat, Jun 9, 2018 at 9:33 PM, Greg Dennis <
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Actually, all the data you need is available from that page. The "Ballot
Image" file will give you the full cast vote record of every individual
ballot, and the "Master Lookup" is the legend that tells you what each
number means. If you have trouble interpreting the numbers, just ping me!
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Converting instant-runoff counts into pairwise counts might be (probably
is?) possible, but I don't have time to do that analysis.
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only
case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is
rare when the number of ballots exceeds a few hundred.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots
had this issue. Right?
That's not a big number, but a fair counting method -- such as pairwise
counting -- would not have to discard any ballots.
The bigger number is "under votes" and admittedly pairwise counting
cannot compensate for a voter saying "here is the only acceptable choice"
(or two choices in this case).
It's great that these results are getting analyzed by people who do not
drink the FairVote kool-aid.
In haste,
Richard Fobes
"The VoteFair guy"
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
Based on the most recent analysis of these numbers that I saw, Leno was
indeed the Condorcet winner, and if Breed were to beat Leno in the final
round, she would then necessarily be the Condorcet winner. The
probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so
far.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor. however, i think the news media made it clear
that the race was really gonna be between Leno, Breed, and Kim, so
these fringe voters might have a chance to insincerely mark either
Leno or Breed as their 3rd choice and betray their *true* third
choice and, in doing so, have an effect in the final round.
ignoring the problem of only 3 ranking levels, it is not possible
that London Breed is the Condorcet Winner (a.k.a. "pairwise
champion"). it might be the case that Mark Leno or Jane Kim is the
Condorcet Winner and if the latter is the case, this is another real
indictment against STV or IRV as a method of tallying RCV. and your
reverse namesake, FairVote, is partially (or mostly) to blame.
i wonder if the City of SF has a file of all of the cast and scanned
ballots and the full ranking for each. if so, and if they release
it to the public, we can investigate if there is a Condorcet Winner
and if that CW is or is not Mark Leno. this would be interesting.
L8r,
r b-j
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
- insinuating it is not fair. I was quite bothered to have a Dem
in a
progressive city insinuate that first past the post is more
fair. ...
This remark does not imply support for first past the post (FPTP,
a.k.a
plurality counting).
There are other ways to count the preference marks on
"ranked-choice"
ballots. In particular, pairwise counting could be used instead of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had
been
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots.
Unmarked
choices are an indication that the choices are equally disliked.
And
multiple candidates being marked at the same preference level is
also no
problem.
In other words, the ballots contain enough information that they
can be
counted in other ways, besides instant-runoff counting. Those
alternate
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the
vote
currently stands, 144 votes separate the top two candidates
(<0.1%) and
over 16,000 ballots have been exhausted (all 3 choices
eliminated).
About 9% of voters have been removed from the pool. It is very
possible
that the result would have shifted if they had the opportunity to
rank a
4th candidate, and therefore, it is possible that we won’t elect
the
person who truly represents the majority. Is this common? This is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being
asked if
RCV is fair? She stated “This is the system we are working with.
That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
- insinuating it is not fair. I was quite bothered to have a Dem
in a
progressive city insinuate that first past the post is more fair.
It
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also
presents
herself as a minority candidate and it is my understanding that
RCV
gives minority candidates better chances and causes all
candidates to be
more likely to campaign to minority communities. Am I mistaken?
Are
there any legitimate arguments that FPTP can be more fair?
Thoughts?
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em for
list info
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts
p :: 617.863.0746 <tel:617.863.0746>
w :: voterchoicema.org <http://voterchoicema.org/>
:: Follow us on Facebook
<https://www.facebook.com/voterchoicema> and Twitter
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
Brian Olson
2018-06-11 17:16:23 UTC
Permalink
I wrote Python code to parse the 'ballot image' and 'master lookup' files.
https://github.com/brianolson/voteutil/blob/master/python/rcvToNameEq.py

I have this exporting to www-cgi formatting like
Alice=1&Bob=2&Carol=3
Post by Christopher Colosi
How is the ballot image file laid out? It seems like a horrible choice
for formatting. Is there any delimiter between ballots, or is this just
one gigantic string of numbers?
On Sun, Jun 10, 2018 at 10:42 PM robert bristow-johnson <
boy this is embarrassing. i didn't realize they were still counting and
that the lead has changed. looks like London Breed is ahead and will
likely win the STV contest.
and, from Brian's initial analysis, it doesn't look like the Condorcet
Winner will be different from the STV winner. so no Burlington 2009
situation.
L8r,
r b-j
---------------------------- Original Message ----------------------------
Subject: Re: [EM] RCV in SF Mayoral election
Date: Sun, June 10, 2018 10:23 pm
------------------------------------------------------------
--------------
Thank you Brian for doing this. Looks like Leno is the legit winner
and STV and Condorcet agree.
However, the Leno-Breed pair wise tally in the defeat matrix should be
exactly the same as the STV final round result. So something is wrong
somewhere.
But thanks for doing this. I was about to code up a MATLAB program to
parse and count this thing.
"Imagination is more important than knowledge."
-------- Original message --------
Date: 6/10/2018 7:46 AM (GMT-08:00)
Subject: Re: [EM] RCV in SF Mayoral election
Ok, a few lines of Python poking the raw data shows I must have some
bug in my Condorcet implementation. Digging into that...
On Sun, Jun 10, 2018 at 10:08 AM, Greg Dennis <
Brian, how is it possible that those differ? Since all the other
candidates are eliminated in the final round, shouldn't that necessarily be
the same as the pairwise contest between those two?
I processed the latest data (2018-06-09) and posted the results of the
SF Mayor election using a few algorithms:https://bolson.org/
~bolson/2018/SF_Mayor_20180605.html
The Condorcet win is now 97436 to 91740 for Leno over Breed.The IRV
final round is still just 94783 to 94393.
I'm using my software posted at https://github.com/brianolson/voteutil
curl -O http://www.sfelections.org/results/20180605/data/
20180609/20180609_ballotimage.txtcurl -O http://www.sfelections.org/
results/20180605/data/20180609/20180609_masterlookup.txt
(mkdir -p ~/psrc && cd ~/psrc && git clone
https://github.com/brianolson/voteutil.git && cd ~/psrc/voteutil/java &&
mvn package)python3 ~/psrc/voteutil/python/rcvToNameEq.py -m
20180609_masterlookup.txt -b 20180609_ballotimage.txt -o
20180609_%s.nameqjava -jar ~/psrc/voteutil/java/target/voteutil-1.0.0.jar
--rankings --full-html --explain -i 20180609_Mayor.nameq >/tmp/a.html
On Sat, Jun 9, 2018 at 9:33 PM, Greg Dennis <
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Actually, all the data you need is available from that page. The
"Ballot Image" file will give you the full cast vote record of every
individual ballot, and the "Master Lookup" is the legend that tells you
what each number means. If you have trouble interpreting the numbers, just
ping me!
Post by Greg Dennis
https://sfelections.sfgov.org/june-5-2018-election-results-
detailed-reports
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Converting instant-runoff counts into pairwise counts might be
(probably is?) possible, but I don't have time to do that analysis.
Post by Greg Dennis
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only
case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is
rare when the number of ballots exceeds a few hundred.
Post by Greg Dennis
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice
of
Post by Greg Dennis
choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots
had this issue. Right?
That's not a big number, but a fair counting method -- such as pairwise
counting -- would not have to discard any ballots.
The bigger number is "under votes" and admittedly pairwise counting
cannot compensate for a voter saying "here is the only acceptable choice"
(or two choices in this case).
It's great that these results are getting analyzed by people who do not
drink the FairVote kool-aid.
In haste,
Richard Fobes
"The VoteFair guy"
https://sfelections.sfgov.org/june-5-2018-election-results-
detailed-reports
Based on the most recent analysis of these numbers that I saw, Leno was
indeed the Condorcet winner, and if Breed were to beat Leno in the final
round, she would then necessarily be the Condorcet winner. The
probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so
far.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor. however, i think the news media made it clear
that the race was really gonna be between Leno, Breed, and Kim, so
these fringe voters might have a chance to insincerely mark either
Leno or Breed as their 3rd choice and betray their *true* third
choice and, in doing so, have an effect in the final round.
ignoring the problem of only 3 ranking levels, it is not possible
that London Breed is the Condorcet Winner (a.k.a. "pairwise
champion"). it might be the case that Mark Leno or Jane Kim is the
Condorcet Winner and if the latter is the case, this is another real
indictment against STV or IRV as a method of tallying RCV. and your
reverse namesake, FairVote, is partially (or mostly) to blame.
i wonder if the City of SF has a file of all of the cast and scanned
ballots and the full ranking for each. if so, and if they release
it to the public, we can investigate if there is a Condorcet Winner
and if that CW is or is not Mark Leno. this would be interesting.
L8r,
r b-j
Post by Greg Dennis
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by Greg Dennis
- insinuating it is not fair. I was quite bothered to have a Dem
in a
Post by Greg Dennis
progressive city insinuate that first past the post is more
fair. ...
Post by Greg Dennis
This remark does not imply support for first past the post (FPTP,
a.k.a
Post by Greg Dennis
plurality counting).
There are other ways to count the preference marks on
"ranked-choice"
Post by Greg Dennis
ballots. In particular, pairwise counting could be used instead of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had
been
Post by Greg Dennis
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots.
Unmarked
Post by Greg Dennis
choices are an indication that the choices are equally disliked.
And
Post by Greg Dennis
multiple candidates being marked at the same preference level is
also no
Post by Greg Dennis
problem.
In other words, the ballots contain enough information that they
can be
Post by Greg Dennis
counted in other ways, besides instant-runoff counting. Those
alternate
Post by Greg Dennis
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the
vote
Post by Greg Dennis
currently stands, 144 votes separate the top two candidates
(<0.1%) and
Post by Greg Dennis
over 16,000 ballots have been exhausted (all 3 choices
eliminated).
Post by Greg Dennis
About 9% of voters have been removed from the pool. It is very
possible
Post by Greg Dennis
that the result would have shifted if they had the opportunity to
rank a
Post by Greg Dennis
4th candidate, and therefore, it is possible that we won’t elect
the
Post by Greg Dennis
person who truly represents the majority. Is this common? This is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being
asked if
Post by Greg Dennis
RCV is fair? She stated “This is the system we are working with.
That’s
Post by Greg Dennis
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by Greg Dennis
- insinuating it is not fair. I was quite bothered to have a Dem
in a
Post by Greg Dennis
progressive city insinuate that first past the post is more
fair. It
Post by Greg Dennis
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also
presents
Post by Greg Dennis
herself as a minority candidate and it is my understanding that
RCV
Post by Greg Dennis
gives minority candidates better chances and causes all
candidates to be
Post by Greg Dennis
more likely to campaign to minority communities. Am I mistaken?
Are
Post by Greg Dennis
there any legitimate arguments that FPTP can be more fair?
Thoughts?
Post by Greg Dennis
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em for
list info
Post by Greg Dennis
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts
voterchoicema.org>
p :: 617.863.0746 <tel:617.863.0746>
w :: voterchoicema.org <http://voterchoicema.org/>
:: Follow us on Facebook
<https://www.facebook.com/voterchoicema> and Twitter
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
Rob Lanphier
2018-06-16 06:08:03 UTC
Permalink
Post by Brian Olson
I wrote Python code to parse the 'ballot image' and 'master lookup' files.
https://github.com/brianolson/voteutil/blob/master/python/rcvToNameEq.py
I have this exporting to www-cgi formatting like
Alice=1&Bob=2&Carol=3
Cool, you're obviously a lot further along than I am. I've got my own (as
yet unpublished) script for parsing out the ballot image files, but I don't
have my script doing any tallying yet.

I downloaded your voteutil code, ran rcvToNameEq against SF's report14
results, and then tried running that through countvotes.py, getting it to
cough up a matrix. I don't want to say yet who my output shows as having
won a particular pairwise contest after parsing report14 (124004 to
105292), but it's kinda interesting (and probably a sign I need to debug
something rather than finding something truly interesting).

Rob
Rob Lanphier
2018-06-17 21:30:28 UTC
Permalink
I'm still sorting through this. Rather than give up on my parsing script,
I decided to use it to create a parsing tool for parsing the sfgov.org
published results, which I just published as a standalone script
(sfballotparse.py):
https://gist.github.com/robla/7664d03372e6a80f1372869c09472b60

Now to nerd out just a little bit: Brian Olson's output format that he uses
(one urlencoded line per ballot-contest) can be condensed into a vaguely
readable form if piped through "sort | uniq -c". Which I did, then
published here:
https://gist.github.com/robla/b9e8df41d94a62d4190760eac044b3bc

I'm planning on using that as a means of manually checking the results when
I ran Brian's tool to get the pairwise count from the election.

I've tested my tool to create input for Brian's
voteutil/python/countvotes.py script. The results were very close, but not
identical; I haven't yet figured out if it's a bug in my code, or a bug in
his code, or sunspots, or what.

Rob
p.s. Brian, I would love it if you had the time to give my script a whirl,
but no worries if you don't. I'll eventually get around to filing a
bug/submitting-a-patch/whatever-is-appropriate
Post by Rob Lanphier
Post by Brian Olson
I wrote Python code to parse the 'ballot image' and 'master lookup' files.
https://github.com/brianolson/voteutil/blob/master/python/rcvToNameEq.py
I have this exporting to www-cgi formatting like
Alice=1&Bob=2&Carol=3
Cool, you're obviously a lot further along than I am. I've got my own (as
yet unpublished) script for parsing out the ballot image files, but I don't
have my script doing any tallying yet.
I downloaded your voteutil code, ran rcvToNameEq against SF's report14
results, and then tried running that through countvotes.py, getting it to
cough up a matrix. I don't want to say yet who my output shows as having
won a particular pairwise contest after parsing report14 (124004 to
105292), but it's kinda interesting (and probably a sign I need to debug
something rather than finding something truly interesting).
Rob
Greg Dennis
2018-06-11 17:21:32 UTC
Permalink
This document describes the layout:
http://www.acgov.org/rov/rcv/results/226/ballot_image_help.pdf
Post by Christopher Colosi
How is the ballot image file laid out? It seems like a horrible choice
for formatting. Is there any delimiter between ballots, or is this just
one gigantic string of numbers?
On Sun, Jun 10, 2018 at 10:42 PM robert bristow-johnson <
boy this is embarrassing. i didn't realize they were still counting and
that the lead has changed. looks like London Breed is ahead and will
likely win the STV contest.
and, from Brian's initial analysis, it doesn't look like the Condorcet
Winner will be different from the STV winner. so no Burlington 2009
situation.
L8r,
r b-j
---------------------------- Original Message ----------------------------
Subject: Re: [EM] RCV in SF Mayoral election
Date: Sun, June 10, 2018 10:23 pm
------------------------------------------------------------
--------------
Thank you Brian for doing this. Looks like Leno is the legit winner
and STV and Condorcet agree.
However, the Leno-Breed pair wise tally in the defeat matrix should be
exactly the same as the STV final round result. So something is wrong
somewhere.
But thanks for doing this. I was about to code up a MATLAB program to
parse and count this thing.
"Imagination is more important than knowledge."
-------- Original message --------
Date: 6/10/2018 7:46 AM (GMT-08:00)
Subject: Re: [EM] RCV in SF Mayoral election
Ok, a few lines of Python poking the raw data shows I must have some
bug in my Condorcet implementation. Digging into that...
On Sun, Jun 10, 2018 at 10:08 AM, Greg Dennis <
Brian, how is it possible that those differ? Since all the other
candidates are eliminated in the final round, shouldn't that necessarily be
the same as the pairwise contest between those two?
I processed the latest data (2018-06-09) and posted the results of the
SF Mayor election using a few algorithms:https://bolson.org/
~bolson/2018/SF_Mayor_20180605.html
The Condorcet win is now 97436 to 91740 for Leno over Breed.The IRV
final round is still just 94783 to 94393.
I'm using my software posted at https://github.com/brianolson/voteutil
curl -O http://www.sfelections.org/results/20180605/data/
20180609/20180609_ballotimage.txtcurl -O http://www.sfelections.org/
results/20180605/data/20180609/20180609_masterlookup.txt
(mkdir -p ~/psrc && cd ~/psrc && git clone
https://github.com/brianolson/voteutil.git && cd ~/psrc/voteutil/java &&
mvn package)python3 ~/psrc/voteutil/python/rcvToNameEq.py -m
20180609_masterlookup.txt -b 20180609_ballotimage.txt -o
20180609_%s.nameqjava -jar ~/psrc/voteutil/java/target/voteutil-1.0.0.jar
--rankings --full-html --explain -i 20180609_Mayor.nameq >/tmp/a.html
On Sat, Jun 9, 2018 at 9:33 PM, Greg Dennis <
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Actually, all the data you need is available from that page. The
"Ballot Image" file will give you the full cast vote record of every
individual ballot, and the "Master Lookup" is the legend that tells you
what each number means. If you have trouble interpreting the numbers, just
ping me!
Post by Greg Dennis
https://sfelections.sfgov.org/june-5-2018-election-results-
detailed-reports
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Converting instant-runoff counts into pairwise counts might be
(probably is?) possible, but I don't have time to do that analysis.
Post by Greg Dennis
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only
case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is
rare when the number of ballots exceeds a few hundred.
Post by Greg Dennis
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice
of
Post by Greg Dennis
choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots
had this issue. Right?
That's not a big number, but a fair counting method -- such as pairwise
counting -- would not have to discard any ballots.
The bigger number is "under votes" and admittedly pairwise counting
cannot compensate for a voter saying "here is the only acceptable choice"
(or two choices in this case).
It's great that these results are getting analyzed by people who do not
drink the FairVote kool-aid.
In haste,
Richard Fobes
"The VoteFair guy"
https://sfelections.sfgov.org/june-5-2018-election-results-
detailed-reports
Based on the most recent analysis of these numbers that I saw, Leno was
indeed the Condorcet winner, and if Breed were to beat Leno in the final
round, she would then necessarily be the Condorcet winner. The
probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case so
far.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice of
choosing the mayor. however, i think the news media made it clear
that the race was really gonna be between Leno, Breed, and Kim, so
these fringe voters might have a chance to insincerely mark either
Leno or Breed as their 3rd choice and betray their *true* third
choice and, in doing so, have an effect in the final round.
ignoring the problem of only 3 ranking levels, it is not possible
that London Breed is the Condorcet Winner (a.k.a. "pairwise
champion"). it might be the case that Mark Leno or Jane Kim is the
Condorcet Winner and if the latter is the case, this is another real
indictment against STV or IRV as a method of tallying RCV. and your
reverse namesake, FairVote, is partially (or mostly) to blame.
i wonder if the City of SF has a file of all of the cast and scanned
ballots and the full ranking for each. if so, and if they release
it to the public, we can investigate if there is a Condorcet Winner
and if that CW is or is not Mark Leno. this would be interesting.
L8r,
r b-j
Post by Greg Dennis
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by Greg Dennis
- insinuating it is not fair. I was quite bothered to have a Dem
in a
Post by Greg Dennis
progressive city insinuate that first past the post is more
fair. ...
Post by Greg Dennis
This remark does not imply support for first past the post (FPTP,
a.k.a
Post by Greg Dennis
plurality counting).
There are other ways to count the preference marks on
"ranked-choice"
Post by Greg Dennis
ballots. In particular, pairwise counting could be used instead of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had
been
Post by Greg Dennis
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots.
Unmarked
Post by Greg Dennis
choices are an indication that the choices are equally disliked.
And
Post by Greg Dennis
multiple candidates being marked at the same preference level is
also no
Post by Greg Dennis
problem.
In other words, the ballots contain enough information that they
can be
Post by Greg Dennis
counted in other ways, besides instant-runoff counting. Those
alternate
Post by Greg Dennis
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the
vote
Post by Greg Dennis
currently stands, 144 votes separate the top two candidates
(<0.1%) and
Post by Greg Dennis
over 16,000 ballots have been exhausted (all 3 choices
eliminated).
Post by Greg Dennis
About 9% of voters have been removed from the pool. It is very
possible
Post by Greg Dennis
that the result would have shifted if they had the opportunity to
rank a
Post by Greg Dennis
4th candidate, and therefore, it is possible that we won’t elect
the
Post by Greg Dennis
person who truly represents the majority. Is this common? This is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being
asked if
Post by Greg Dennis
RCV is fair? She stated “This is the system we are working with.
That’s
Post by Greg Dennis
a discussion we can have at a later time. For now, we’re stuck
with it.”
Post by Greg Dennis
- insinuating it is not fair. I was quite bothered to have a Dem
in a
Post by Greg Dennis
progressive city insinuate that first past the post is more
fair. It
Post by Greg Dennis
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also
presents
Post by Greg Dennis
herself as a minority candidate and it is my understanding that
RCV
Post by Greg Dennis
gives minority candidates better chances and causes all
candidates to be
Post by Greg Dennis
more likely to campaign to minority communities. Am I mistaken?
Are
Post by Greg Dennis
there any legitimate arguments that FPTP can be more fair?
Thoughts?
Post by Greg Dennis
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em for
list info
Post by Greg Dennis
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts
voterchoicema.org>
p :: 617.863.0746 <tel:617.863.0746>
w :: voterchoicema.org <http://voterchoicema.org/>
:: Follow us on Facebook
<https://www.facebook.com/voterchoicema> and Twitter
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
Christopher Colosi
2018-06-12 01:47:04 UTC
Permalink
Awesome, I’ll take a look at those. Thanks.
Post by Greg Dennis
http://www.acgov.org/rov/rcv/results/226/ballot_image_help.pdf
Post by Christopher Colosi
How is the ballot image file laid out? It seems like a horrible choice
for formatting. Is there any delimiter between ballots, or is this just
one gigantic string of numbers?
On Sun, Jun 10, 2018 at 10:42 PM robert bristow-johnson <
boy this is embarrassing. i didn't realize they were still counting and
that the lead has changed. looks like London Breed is ahead and will
likely win the STV contest.
and, from Brian's initial analysis, it doesn't look like the Condorcet
Winner will be different from the STV winner. so no Burlington 2009
situation.
L8r,
r b-j
---------------------------- Original Message
----------------------------
Subject: Re: [EM] RCV in SF Mayoral election
Date: Sun, June 10, 2018 10:23 pm
--------------------------------------------------------------------------
Thank you Brian for doing this. Looks like Leno is the legit winner
and STV and Condorcet agree.
However, the Leno-Breed pair wise tally in the defeat matrix should be
exactly the same as the STV final round result. So something is wrong
somewhere.
But thanks for doing this. I was about to code up a MATLAB program to
parse and count this thing.
"Imagination is more important than knowledge."
-------- Original message --------
Date: 6/10/2018 7:46 AM (GMT-08:00)
Subject: Re: [EM] RCV in SF Mayoral election
Ok, a few lines of Python poking the raw data shows I must have some
bug in my Condorcet implementation. Digging into that...
On Sun, Jun 10, 2018 at 10:08 AM, Greg Dennis <
Brian, how is it possible that those differ? Since all the other
candidates are eliminated in the final round, shouldn't that necessarily be
the same as the pairwise contest between those two?
I processed the latest data (2018-06-09) and posted the results of the
https://bolson.org/~bolson/2018/SF_Mayor_20180605.html
The Condorcet win is now 97436 to 91740 for Leno over Breed.The IRV
final round is still just 94783 to 94393.
I'm using my software posted at https://github.com/brianolson/voteutil
curl -O
http://www.sfelections.org/results/20180605/data/20180609/20180609_ballotimage.txtcurl
-O
http://www.sfelections.org/results/20180605/data/20180609/20180609_masterlookup.txt
(mkdir -p ~/psrc && cd ~/psrc && git clone
https://github.com/brianolson/voteutil.git && cd ~/psrc/voteutil/java
&& mvn package)python3 ~/psrc/voteutil/python/rcvToNameEq.py -m
20180609_masterlookup.txt -b 20180609_ballotimage.txt -o
20180609_%s.nameqjava -jar ~/psrc/voteutil/java/target/voteutil-1.0.0.jar
--rankings --full-html --explain -i 20180609_Mayor.nameq >/tmp/a.html
On Sat, Jun 9, 2018 at 9:33 PM, Greg Dennis <
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Actually, all the data you need is available from that page. The
"Ballot Image" file will give you the full cast vote record of every
individual ballot, and the "Master Lookup" is the legend that tells you
what each number means. If you have trouble interpreting the numbers, just
ping me!
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Converting instant-runoff counts into pairwise counts might be
(probably is?) possible, but I don't have time to do that analysis.
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only
case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is
rare when the number of ballots exceeds a few hundred.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice
of
choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots
had this issue. Right?
That's not a big number, but a fair counting method -- such as
pairwise counting -- would not have to discard any ballots.
The bigger number is "under votes" and admittedly pairwise counting
cannot compensate for a voter saying "here is the only acceptable choice"
(or two choices in this case).
It's great that these results are getting analyzed by people who do
not drink the FairVote kool-aid.
In haste,
Richard Fobes
"The VoteFair guy"
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
Based on the most recent analysis of these numbers that I saw, Leno was
indeed the Condorcet winner, and if Breed were to beat Leno in the
final
round, she would then necessarily be the Condorcet winner. The
probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case
so far.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice
of
choosing the mayor. however, i think the news media made it clear
that the race was really gonna be between Leno, Breed, and Kim, so
these fringe voters might have a chance to insincerely mark either
Leno or Breed as their 3rd choice and betray their *true* third
choice and, in doing so, have an effect in the final round.
ignoring the problem of only 3 ranking levels, it is not possible
that London Breed is the Condorcet Winner (a.k.a. "pairwise
champion"). it might be the case that Mark Leno or Jane Kim is the
Condorcet Winner and if the latter is the case, this is another
real
indictment against STV or IRV as a method of tallying RCV. and
your
reverse namesake, FairVote, is partially (or mostly) to blame.
i wonder if the City of SF has a file of all of the cast and
scanned
ballots and the full ranking for each. if so, and if they release
it to the public, we can investigate if there is a Condorcet Winner
and if that CW is or is not Mark Leno. this would be interesting.
L8r,
r b-j
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
- insinuating it is not fair. I was quite bothered to have a
Dem
in a
progressive city insinuate that first past the post is more
fair. ...
This remark does not imply support for first past the post (FPTP,
a.k.a
plurality counting).
There are other ways to count the preference marks on
"ranked-choice"
ballots. In particular, pairwise counting could be used instead
of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they had
been
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots.
Unmarked
choices are an indication that the choices are equally disliked.
And
multiple candidates being marked at the same preference level is
also no
problem.
In other words, the ballots contain enough information that they
can be
counted in other ways, besides instant-runoff counting. Those
alternate
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As the
vote
currently stands, 144 votes separate the top two candidates
(<0.1%) and
over 16,000 ballots have been exhausted (all 3 choices
eliminated).
About 9% of voters have been removed from the pool. It is very
possible
that the result would have shifted if they had the opportunity
to
rank a
4th candidate, and therefore, it is possible that we won’t
elect the
person who truly represents the majority. Is this common? This
is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being
asked if
RCV is fair? She stated “This is the system we are working with.
That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
- insinuating it is not fair. I was quite bothered to have a
Dem in a
progressive city insinuate that first past the post is more
fair. It
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also
presents
herself as a minority candidate and it is my understanding that
RCV
gives minority candidates better chances and causes all
candidates to be
more likely to campaign to minority communities. Am I mistaken?
Are
there any legitimate arguments that FPTP can be more fair?
Thoughts?
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em
for
list info
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts
p :: 617.863.0746 <tel:617.863.0746>
w :: voterchoicema.org <http://voterchoicema.org/>
:: Follow us on Facebook
<https://www.facebook.com/voterchoicema> and Twitter
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
----
Election-Methods mailing list - see http://electorama.com/em for list
info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
Christopher Colosi
2018-06-13 16:18:52 UTC
Permalink
Does the master ballot image separate day-of voting from mail-in ballots?
And does it further separate mail-in by arrival prior to election, drop off
at ballot locations, and arrival after election? It would be interesting
to see how much the vote varied by these groupings. Doesn’t look like that
info is in there unless it is the Tally Type field. I guess by comparing
the final report on the day of to the ultimate final report, we could split
early mail-ins and day-of from drop offs and late arrivals. Might be able
to further split early mail-ins from day-of if the first report is prior to
tallying of any day-of votes. Has anyone done this?
Post by Christopher Colosi
Awesome, I’ll take a look at those. Thanks.
On Mon, Jun 11, 2018 at 10:21 AM Greg Dennis <
Post by Greg Dennis
http://www.acgov.org/rov/rcv/results/226/ballot_image_help.pdf
Post by Christopher Colosi
How is the ballot image file laid out? It seems like a horrible choice
for formatting. Is there any delimiter between ballots, or is this just
one gigantic string of numbers?
On Sun, Jun 10, 2018 at 10:42 PM robert bristow-johnson <
boy this is embarrassing. i didn't realize they were still counting
and that the lead has changed. looks like London Breed is ahead and will
likely win the STV contest.
and, from Brian's initial analysis, it doesn't look like the Condorcet
Winner will be different from the STV winner. so no Burlington 2009
situation.
L8r,
r b-j
---------------------------- Original Message
----------------------------
Subject: Re: [EM] RCV in SF Mayoral election
Date: Sun, June 10, 2018 10:23 pm
--------------------------------------------------------------------------
Thank you Brian for doing this. Looks like Leno is the legit winner
and STV and Condorcet agree.
However, the Leno-Breed pair wise tally in the defeat matrix should
be exactly the same as the STV final round result. So something is wrong
somewhere.
But thanks for doing this. I was about to code up a MATLAB program
to parse and count this thing.
"Imagination is more important than knowledge."
-------- Original message --------
Date: 6/10/2018 7:46 AM (GMT-08:00)
Subject: Re: [EM] RCV in SF Mayoral election
Ok, a few lines of Python poking the raw data shows I must have some
bug in my Condorcet implementation. Digging into that...
On Sun, Jun 10, 2018 at 10:08 AM, Greg Dennis <
Brian, how is it possible that those differ? Since all the other
candidates are eliminated in the final round, shouldn't that necessarily be
the same as the pairwise contest between those two?
I processed the latest data (2018-06-09) and posted the results of
https://bolson.org/~bolson/2018/SF_Mayor_20180605.html
The Condorcet win is now 97436 to 91740 for Leno over Breed.The IRV
final round is still just 94783 to 94393.
I'm using my software posted at
https://github.com/brianolson/voteutil
commands (needs maven installed for compiling Java, and needs
curl -O
http://www.sfelections.org/results/20180605/data/20180609/20180609_ballotimage.txtcurl
-O
http://www.sfelections.org/results/20180605/data/20180609/20180609_masterlookup.txt
(mkdir -p ~/psrc && cd ~/psrc && git clone
https://github.com/brianolson/voteutil.git && cd ~/psrc/voteutil/java
&& mvn package)python3 ~/psrc/voteutil/python/rcvToNameEq.py -m
20180609_masterlookup.txt -b 20180609_ballotimage.txt -o
20180609_%s.nameqjava -jar ~/psrc/voteutil/java/target/voteutil-1.0.0.jar
--rankings --full-html --explain -i 20180609_Mayor.nameq >/tmp/a.html
On Sat, Jun 9, 2018 at 9:33 PM, Greg Dennis <
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Actually, all the data you need is available from that page. The
"Ballot Image" file will give you the full cast vote record of every
individual ballot, and the "Master Lookup" is the legend that tells you
what each number means. If you have trouble interpreting the numbers, just
ping me!
On Sat, Jun 9, 2018 at 4:36 PM, VoteFair <
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
I quickly looked at the vote data and saw that lots of ballots are
categorized as "Exhausted by Over Votes" and "Under Votes," but there is no
data indicating exactly how those ballots were marked, so we lack enough
information to be sure of final results.
Converting instant-runoff counts into pairwise counts might be
(probably is?) possible, but I don't have time to do that analysis.
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only
case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner --
is rare when the number of ballots exceeds a few hundred.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was
effectively
"disenfranchised" by being unable to weigh in on the final
choice of
choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so
ballots had this issue. Right?
That's not a big number, but a fair counting method -- such as
pairwise counting -- would not have to discard any ballots.
The bigger number is "under votes" and admittedly pairwise counting
cannot compensate for a voter saying "here is the only acceptable choice"
(or two choices in this case).
It's great that these results are getting analyzed by people who do
not drink the FairVote kool-aid.
In haste,
Richard Fobes
"The VoteFair guy"
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
Based on the most recent analysis of these numbers that I saw, Leno
was
indeed the Condorcet winner, and if Breed were to beat Leno in the
final
round, she would then necessarily be the Condorcet winner. The
probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only case
so far.
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
the limitation to only three levels of ranking is a problem. if
someone ranked all three levels and none of the candidates ranked
were either London Breed nor Mark Leno, that voter was effectively
"disenfranchised" by being unable to weigh in on the final choice
of
choosing the mayor. however, i think the news media made it clear
that the race was really gonna be between Leno, Breed, and Kim, so
these fringe voters might have a chance to insincerely mark either
Leno or Breed as their 3rd choice and betray their *true* third
choice and, in doing so, have an effect in the final round.
ignoring the problem of only 3 ranking levels, it is not possible
that London Breed is the Condorcet Winner (a.k.a. "pairwise
champion"). it might be the case that Mark Leno or Jane Kim is
the
Condorcet Winner and if the latter is the case, this is another
real
indictment against STV or IRV as a method of tallying RCV. and
your
reverse namesake, FairVote, is partially (or mostly) to blame.
i wonder if the City of SF has a file of all of the cast and
scanned
ballots and the full ranking for each. if so, and if they release
it to the public, we can investigate if there is a Condorcet
Winner
and if that CW is or is not Mark Leno. this would be interesting.
L8r,
r b-j
... She stated “This is the system we are working with. That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
- insinuating it is not fair. I was quite bothered to have a
Dem
in a
progressive city insinuate that first past the post is more
fair. ...
This remark does not imply support for first past the post
(FPTP,
a.k.a
plurality counting).
There are other ways to count the preference marks on
"ranked-choice"
ballots. In particular, pairwise counting could be used instead
of
instant-runoff counting, and that is fairer than FPTP.
1. May not elect majority candidate
...
Is this common? This is
probably an abnormally close race. Thoughts?
I doubt the voters would regard this as a close race if they
had been
able to fully rank all the choices. The 3-choice limitation is
simplistic, and complicates the counting.
Pairwise counting does not result in any exhausted ballots.
Unmarked
choices are an indication that the choices are equally
disliked. And
multiple candidates being marked at the same preference level is
also no
problem.
In other words, the ballots contain enough information that they
can be
counted in other ways, besides instant-runoff counting. Those
alternate
counting methods could reveal a clearer outcome.
In haste,
Richard Fobes
Curious to hear people’s thoughts on some issues.
1. May not elect majority candidate
In SF, we restrict to 3 choices to simplify the process. As
the vote
currently stands, 144 votes separate the top two candidates
(<0.1%) and
over 16,000 ballots have been exhausted (all 3 choices
eliminated).
About 9% of voters have been removed from the pool. It is very
possible
that the result would have shifted if they had the opportunity
to
rank a
4th candidate, and therefore, it is possible that we won’t
elect the
person who truly represents the majority. Is this common? This
is
probably an abnormally close race. Thoughts?
2. What are your thoughts on London Breed’s response to being
asked if
RCV is fair? She stated “This is the system we are working
with.
That’s
a discussion we can have at a later time. For now, we’re stuck
with it.”
- insinuating it is not fair. I was quite bothered to have a
Dem in a
progressive city insinuate that first past the post is more
fair. It
also felt divisive. If Leno wins, will her supporters feel that
democracy prevailed, or that the election was stolen? She also
presents
herself as a minority candidate and it is my understanding
that RCV
gives minority candidates better chances and causes all
candidates to be
more likely to campaign to minority communities. Am I
mistaken? Are
there any legitimate arguments that FPTP can be more fair?
Thoughts?
Regards,
—Chris
----
Election-Methods mailing list - see http://electorama.com/em
for
list info
----
Election-Methods mailing list - see http://electorama.com/em
for
list info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
*Greg Dennis, Ph.D. :: Policy Director*
Voter Choice Massachusetts
p :: 617.863.0746 <tel:617.863.0746>
w :: voterchoicema.org <http://voterchoicema.org/>
:: Follow us on Facebook
<https://www.facebook.com/voterchoicema> and Twitter
----
Election-Methods mailing list - see http://electorama.com/em for
list info
----
Election-Methods mailing list - see http://electorama.com/em for
list info
----
Election-Methods mailing list - see http://electorama.com/em for
list info
----
Election-Methods mailing list - see http://electorama.com/em for
list info
----
Election-Methods mailing list - see http://electorama.com/em for
list info
--
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
----
Election-Methods mailing list - see http://electorama.com/em for list info
robert bristow-johnson
2018-06-11 02:32:43 UTC
Permalink
Remember to strip of the city council race on the bottom (with 00021 as a Contest_ID).  


--r b-j                     ***@audioimagination.com
"Imagination is more important than knowledge."




-------- Original message --------
From: Brian Olson <***@bolson.org>
Date: 6/10/2018 7:46 AM (GMT-08:00)
To: EM <election-***@lists.electorama.com>
Subject: Re: [EM] RCV in SF Mayoral election

Ok, a few lines of Python poking the raw data shows I must have some bug in my Condorcet implementation. Digging into that...
On Sun, Jun 10, 2018 at 10:08 AM, Greg Dennis <***@voterchoicema.org> wrote:
Brian, how is it possible that those differ? Since all the other candidates are eliminated in the final round, shouldn't that necessarily be the same as the pairwise contest between those two?

On Sun, Jun 10, 2018, 9:46 AM Brian Olson <***@bolson.org> wrote:
I processed the latest data (2018-06-09) and posted the results of the SF Mayor election using a few algorithms:https://bolson.org/~bolson/2018/SF_Mayor_20180605.html

The Condorcet win is now 97436 to 91740 for Leno over Breed.The IRV final round is still just 94783 to 94393.

I'm using my software posted at https://github.com/brianolson/voteutil
commands (needs maven installed for compiling Java, and needs Python3):
curl -O http://www.sfelections.org/results/20180605/data/20180609/20180609_ballotimage.txtcurl -O http://www.sfelections.org/results/20180605/data/20180609/20180609_masterlookup.txt
(mkdir -p ~/psrc && cd ~/psrc && git clone https://github.com/brianolson/voteutil.git && cd ~/psrc/voteutil/java && mvn package)python3 ~/psrc/voteutil/python/rcvToNameEq.py -m 20180609_masterlookup.txt -b 20180609_ballotimage.txt -o 20180609_%s.nameqjava -jar ~/psrc/voteutil/java/target/voteutil-1.0.0.jar --rankings --full-html --explain -i 20180609_Mayor.nameq >/tmp/a.html

On Sat, Jun 9, 2018 at 9:33 PM, Greg Dennis <***@voterchoicema.org> wrote:
I quickly looked at the vote data and saw that lots of ballots are categorized as "Exhausted by Over Votes" and "Under Votes," but there is no data indicating exactly how those ballots were marked, so we lack enough information to be sure of final results.
Actually, all the data you need is available from that page. The "Ballot Image" file will give you the full cast vote record of every individual ballot, and the "Master Lookup" is the legend that tells you what each number means. If you have trouble interpreting the numbers, just ping me!
Post by Greg Dennis
https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports
I quickly looked at the vote data and saw that lots of ballots are categorized as "Exhausted by Over Votes" and "Under Votes," but there is no data indicating exactly how those ballots were marked, so we lack enough information to be sure of final results.



Converting instant-runoff counts into pairwise counts might be (probably is?) possible, but I don't have time to do that analysis.
Post by Greg Dennis
The probability of IRV not elected the Condorcet winner appears to be
exceedingly low in practice. We're up to about ~200 IRV elections
conducted nationwide since 2004 and Burlington 2009 is the only
case so far.
Yes, circular ambiguity -- in which there is no Condorcet winner -- is rare when the number of ballots exceeds a few hundred.
Post by Greg Dennis
On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson
     the limitation to only three levels of ranking is a problem.  if
     someone ranked all three levels and none of the candidates ranked
     were either London Breed nor Mark Leno, that voter was effectively
     "disenfranchised" by being unable to weigh in on the final choice of
     choosing the mayor.
Based on a very quick guesstimate it looks like about 30 or so ballots had this issue.  Right?



That's not a big number, but a fair counting method -- such as pairwise counting -- would not have to discard any ballots.



The bigger number is "under votes" and admittedly pairwise counting cannot compensate for a voter saying "here is the only acceptable choice" (or two choices in this case).



It's great that these results are getting analyzed by people who do not drink the FairVote kool-aid.



In haste,

Richard Fobes

"The VoteFair guy"





On 6/9/2018 6:25 AM, Greg Dennis wrote:


San Francisco always make the cast vote record public:

https://sfelections.sfgov.org/june-5-2018-election-results-detailed-reports



Based on the most recent analysis of these numbers that I saw, Leno was

indeed the Condorcet winner, and if Breed were to beat Leno in the final

round, she would then necessarily be the Condorcet winner. The

probability of IRV not elected the Condorcet winner appears to be

exceedingly low in practice. We're up to about ~200 IRV elections

conducted nationwide since 2004 and Burlington 2009 is the only case so far.



On Sat, Jun 9, 2018 at 2:07 AM, robert bristow-johnson

<***@audioimagination.com <mailto:***@audioimagination.com>> wrote:





    Richard, a few points:



    the limitation to only three levels of ranking is a problem.  if

    someone ranked all three levels and none of the candidates ranked

    were either London Breed nor Mark Leno, that voter was effectively

    "disenfranchised" by being unable to weigh in on the final choice of

    choosing the mayor.  however, i think the news media made it clear

    that the race was really gonna be between Leno, Breed, and Kim, so

    these fringe voters might have a chance to insincerely mark either

    Leno or Breed as their 3rd choice and betray their *true* third

    choice and, in doing so, have an effect in the final round.



    ignoring the problem of only 3 ranking levels, it is not possible

    that London Breed is the Condorcet Winner (a.k.a. "pairwise

    champion").  it might be the case that Mark Leno or Jane Kim is the

    Condorcet Winner and if the latter is the case, this is another real

    indictment against STV or IRV as a method of tallying RCV.  and your

    reverse namesake, FairVote, is partially (or mostly) to blame.



    i wonder if the City of SF has a file of all of the cast and scanned

    ballots and the full ranking for each.  if so, and if they release

    it to the public, we can investigate if there is a Condorcet Winner

    and if that CW is or is not Mark Leno.  this would be interesting.



    L8r,







    r b-j







    > On 6/8/2018 6:24 PM, Christopher Colosi wrote:

    > > ... She stated “This is the system we are working with. That’s

    > > a discussion we can have at a later time. For now, we’re stuck

    with it.”

    > > - insinuating it is not fair. I was quite bothered to have a Dem

    in a

    > > progressive city insinuate that first past the post is more

    fair. ...

    >

    > This remark does not imply support for first past the post (FPTP,

    a.k.a

    > plurality counting).

    >

    > There are other ways to count the preference marks on "ranked-choice"

    > ballots. In particular, pairwise counting could be used instead of

    > instant-runoff counting, and that is fairer than FPTP.

    >

    > > 1. May not elect majority candidate

    > > ...

    > > Is this common? This is

    > > probably an abnormally close race. Thoughts?

    >

    > I doubt the voters would regard this as a close race if they had been

    > able to fully rank all the choices. The 3-choice limitation is

    > simplistic, and complicates the counting.

    >

    > Pairwise counting does not result in any exhausted ballots. Unmarked

    > choices are an indication that the choices are equally disliked. And

    > multiple candidates being marked at the same preference level is

    also no

    > problem.

    >

    > In other words, the ballots contain enough information that they

    can be

    > counted in other ways, besides instant-runoff counting. Those

    alternate

    > counting methods could reveal a clearer outcome.

    >

    > In haste,

    > Richard Fobes

    >

    >

    > On 6/8/2018 6:24 PM, Christopher Colosi wrote:

    >> Curious to hear people’s thoughts on some issues.

    >>

    >> 1. May not elect majority candidate

    >> In SF, we restrict to 3 choices to simplify the process. As the vote

    >> currently stands, 144 votes separate the top two candidates

    (<0.1%) and

    >> over 16,000 ballots have been exhausted (all 3 choices eliminated).

    >> About 9% of voters have been removed from the pool. It is very

    possible

    >> that the result would have shifted if they had the opportunity to

    rank a

    >> 4th candidate, and therefore, it is possible that we won’t elect the

    >> person who truly represents the majority. Is this common? This is

    >> probably an abnormally close race. Thoughts?

    >>

    >> 2. What are your thoughts on London Breed’s response to being

    asked if

    >> RCV is fair? She stated “This is the system we are working with.

    That’s

    >> a discussion we can have at a later time. For now, we’re stuck

    with it.”

    >> - insinuating it is not fair. I was quite bothered to have a Dem in a

    >> progressive city insinuate that first past the post is more fair. It

    >> also felt divisive. If Leno wins, will her supporters feel that

    >> democracy prevailed, or that the election was stolen? She also

    presents

    >> herself as a minority candidate and it is my understanding that RCV

    >> gives minority candidates better chances and causes all

    candidates to be

    >> more likely to campaign to minority communities. Am I mistaken? Are

    >> there any legitimate arguments that FPTP can be more fair? Thoughts?

    >>

    >> Regards,

    >> —Chris

    >>

    >>

    >>

    >> ----

    >> Election-Methods mailing list - see http://electorama.com/em for

    list info

    >>

    > ----

    > Election-Methods mailing list - see http://electorama.com/em for

    list info

    >



    --



    r b-j                         ***@audioimagination.com

    <mailto:***@audioimagination.com>



    "Imagination is more important than knowledge."



















    ----

    Election-Methods mailing list - see http://electorama.com/em for

    list info









--

*Greg Dennis, Ph.D. :: Policy Director*

Voter Choice Massachusetts



e :: ***@voterchoicema.org <mailto:***@voterchoicema.org>

p :: 617.863.0746 <tel:617.863.0746>

w :: voterchoicema.org <http://voterchoicema.org/>



:: Follow us on Facebook

<https://www.facebook.com/voterchoicema> and Twitter

<https://twitter.com/voterchoicema> ::





----

Election-Methods mailing list - see http://electorama.com/em for list info




----

Election-Methods mailing list - see http://electorama.com/em for list info


----

Election-Methods mailing list - see http://electorama.com/em for list info




----

Election-Methods mailing list - see http://electorama.com/em for list info
Loading...