At 07:52 PM 6/14/2013, Jameson Quinn wrote:
So. Abd and I now agree that a Bucklin system
which uses just the above-median votes to break
ties is probably the best first step towards median voting.
Let's stop saying it that way. It's true, but
it's confusing. First of all, you are using "tie"
in an nonstandard way. You mean that the median
vote is the same. But when we say that an
election is tied, we mean something *very different*.
The "tie" here is that both candidates have a
majority approval. But the votes are not tied (unless they are, a rare thing.)
I'd like to get the details worked out, so we
can stop using different terms ("Bucklin",
"MJ", "GMJ") and settle on a single
clearly-defined proposal. I'd love to hear what
others feel about these issues (though this
isn't really the place for debating whether
some other class of voting system, such as
Score or Condorcet or whatever, is better or
worse than Bucklin/Median systems in general).
We should know. Remember I declared a reform
path. I suspect that it ends with a hybrid system
that uses a Range ballot. This is the real issue:
outcomes can be improved by using a system that
allows greater information input from the voters.
That's practically a tautology. However, there is
a cost to that. At some point, the incremental
value is less than the incremental cost. Hence a
particular reform path will terminate, probalby,
short of a naive estimation of perfection,
because a perfect system will *also be economical.*
I don't see going above Range 10. However, this
is important. A rigid reform process must decide
on the ultimate goal before there is adequate
information to do so. That's not necessary. We
can have some idea of the ultimate goal, we do
*not* need to know details, we merely need to
know enough to take the next step.
*Ultimately,* I don't see single-winner elections
at all! So how much should be invested in making
single-winner elections perfect? What I see as of
long-term value is a recognition of the usage of
Range methods for *polling* participants, making
repeated ballot *more efficient*, settling
faster. We may ultimately see, as experience
justifies it, a creeping up of the "quota." I.e,
from the Droop quota, toward the Hare. With
binary elections, that means *toward unanimity.*
In small groups, this is possible, and this was a
common experience in small groups in the 20th
century. Now, how can we scale that up, *without*
the massive inefficiency that often accompanied the seeking of full consensus?
Now, to the point here:
1. How to best express the system? Two equivalent definitions:
* Top-down: "Count the votes at the highest
grade for each candidate. If any one candidate
has a majority, they win. If not, add in lower
grades, one at a time, until some candidate or
candidates get a majority. If two candidates
would reach a majority at the same grade level,
then whichever has the most votes above that
level wins. If there are no votes above that
level, the highest votes at or above that level wins."
First of all, "grade," sorry. It grates. It will
mislead, possibly. These are *actually* category
ranks, i.e, ranks with equal ranking allowed (and
empty ranks being meaningful). Because it is
"grading on the curve," it is actually
accomplished by ranking, as a starting point,
with, then, adjustments to more accurately show
preference strength. The concept of "grade"
confuses that, even though "grading on the curve"
would, in fact, do the same thing. Bucklin used
"ranks." Bucklin is Ranked Approval, that's been
a common name for it. The ranks are "preference
categories." To fit with voting systems
tradition, I'd call them ranks. "Score" takes us back toward "Grade."
The name of "Range" referred to the entire range of possible fractional votes.
The original Bucklin ballot used "First Choice,"
"Second Choice," "Third Choice." And that's what
the votes are: choices. "Choice Approval," what do you think?
The name "Definite Majority Choice" has already
been used.
http://wiki.electorama.com/wiki/Definite_Majority_Choice
says that it is also called "Ranked Approval
Voting." Of course, we are talking about a ranked
Approval method. It's really, as I've pointed out, a Range method.
The final rule discourages multiple approvals in
the first rank without prohibiting them. They
make sense wherever the voter has little
preference between favorite candidates. I would
word it this way, as to what I think today:
Count the votes at 1st Choice for each candidate.
If a single candidate has a majority, this
canditate wins. If not, add in lower choices, one
at a time, until a candidate or candidates gains
a majority. If two or more candidates reach a
majority at a stage, then whichever candidate has
the most votes above that stage wins. If this is
1st Choice, or if all the choices have been
amalgamated, and no candidate has a majority,
then the candidate with the most votes wins.
The last rule is necessary if the method is to complete with a single ballot.
* Bottom-up: "Count the votes at the lowest
grade against each candidate. If any candidates
have a majority against, eliminate them from
consideration. Continue adding in the
next-lowest grade, until there is just one or
zero candidates left. If there's one left, they
win. Otherwise, if the last few candidates are
eliminated together, choose whichever of that
group was eliminated by the smallest majority against."
Right away, what is a 'vote against." I don't
know what "the lowest grade against the
candidate" means. While this may be equivalent
mathematically -- I have not checked that -- it
is not equivalent as to easy access to the
concept. The first description is immediately
clear, and is consistent with how voting is
normally done. It is sequential approval voting,
no eliminations, repeated ballot, with voter-controlled additional approvals.
The difference between this and regular Bucklin
is only in how multiple approvals are handled.
The two methods will *usually* produce the same
result. Awarding the win to the most votes
*including the multiple approvals* fits with
voting tradition on multiple conflicting ballot
questions. That tradition, however, did not
contemplate many-candidate situations, and this
is a characteristic of sequential approval. In
pure repeated ballot, no eliminations, people
will not rush to add in approvals in the presence
of strong preference. So when this happens in a
3-rank ballot, that there are multiple approvals,
and even if the lower-vote candidate might be a
fair winner, we can suspect that the voters who
first preferred the other candidate would not
have added the extra approval if they had known.
This could cut both ways, and if we really wanted
to be certain, we'd hold a runoff. I see
selecting the most-approved at the higher level
as being safer, somewhat, if we *must* decide
now. There is a more complex option that
considers the votes as range, and we might look at that.
I.e., if there is a multiple majority at a level,
then the votes are valued at range value -- (0-4)
or (0,2,3,4) --, are added, and the candidate
with the highest total wins. That, then, does
give precedence to the higher ranks, but also
allows consideration of the lower.
I am not strongly attached to either approach,
between the two pure approval approaches, i.e,
highest votes at the majority-finding level, or
highest votes at the level above.
But I don't want to see this described as bottom-up. It's totally confusing.
2. How many rating/grade/rank levels should be
used, and how should they be labeled? I'd
suggest the following 5, along the lines of something Abd proposed:
A: Unequivocal support
B: Probable support (unless there's a candidate with majority "A" support)
C: Neutral (support or oppose, depending on other candidates' results)
D: Probable opposition (unless all other
candidates have majority "F" opposition)
F: Unequivocal opposition
I'd prefer to give them the well-known Grade
Point equivalents. That is, 4,3,2,1,0. I don't
like the names given, what does "probable" mean?
And I don't see we are quite ready to agree on
the rating of 1. We need to be very clear about
that rating. I consider it an unapproved rank.
The rating of 2 *is* an approved rank. Calling it
"neutral" is somewhat misleading. It is an
*allowance,* sometimes called in group process a
"stand-aside." You did reflect that in "depending
on other candidate's results," but it is *not* an
opposition. The explanations are *way* too complicated.
Is a vote a support or not? I don't see how that
would depend on what others have voted. Please,
don't explain it to me. I could figure out an
explanation. These descriptions should be
self-contained and self-evident. If they need
extra explanation, they aren't there yet.
(I've relabeled the categories to help clarify
their strategic meaning; for instance, I changed "strong" to "unequivocal")
Full. This is the definite meaning: this
candidate is the favorite, or close enough to also get full support.
I would also be open to having blank votes count
as "E" rather than "F", but I think that's
probably an unnecessary complication to begin with.
If we are not going to use F, we lose the grade
analogy, but, as you know "*I don't want that.*
This is tossing weights in a series of buckets,
until they lift the "Majority bucket" on the
other side of the balance. If it's most votes
with a multiple majority, then whichever bucket
weighs the most. If it's most votes at the stage
before, then we have to have colored weights and
pull out the last added. I am *not* sure that this is best.
What are we trying to avoid? We are trying to
avoid completing with a multiple majority. But
why? To avoid accusations of violating the majority criterion?
I actually think that strategic questions like
this should first be well-defined as questions,
i.e, what are our various choices here? Then we
need to do *actual research* as to how people
will understand these definitions and procedures.
Focus groups. It's about time that we start
approaching these matters professionally, or with professional quality.
3. What should we call this system? Abd seemed
happy with "Instant Runoff Approval Voting". I'd
be fine with that too, but before we settle on
that, we should look at the downsides:
It's probably a good name.
* FairVote has been moving away from
"Instant Runoff / IRV" and towards "ranked
choice/ RCV" in recent years. I don't know all
of their reasons, but I suspect it is partially
to do with the legalism of ballot initiative
language. That is, IRV is technically neither
instant nor a runoff, though it is certainly close on both counts.
They went to Ranked Choice Voting, as I recall,
in San Francisco, because of a limit to three
choices on the ballot. Btu there may have been
other reasons. I don't think the "legal" argument
about names is valid. They could call a snail,
instant, and if the voters pass it, a snail is
instant. "Bad name" is not a valid legal argument.
* It could lead to confusion between IRAV
and IRV. That has its upsides piggybacking on
FairVote's existing publicity but at also its
downsides as we know, IRV is actually a pretty flawed sysstem.
It was missing the A. We are supplying that.
Count All the Votes. (And I'm not totally content
that what we are describing is an optimal second
step. I am not actually thrilled with only
counting the ranks to a certain level. Ahem. Range.)
So I think we should have a poll with various
options (using the system itself to rate the
options, of course). I'll start out with some proposals and my votes:
-IRAV: B
-Descending Approval Threshold (DAT) Voting: A
-Descending Approval Threshold Adjudgment (DATA voting): B
-Majority Approval Threshold (MAT; note that the
M could also be backronymmed to "Median"): A
-Bucklin: F (not that we shouldn't say that this
system is a Bucklin system, just that that shouldn't be our only name for it)
-Bucklin-ER or ER-Bucklin: D (has already been
used for other systems, not a descriptive name)
-Graded Approval Threshold (GAT): C (Not bad, but not great)
-Majority Assignment of Grades (MAG): C (ditto)
-Graded Majority Approval (GMA): B (this one seems simple and descriptive)
IRAV: A
Bucklin: B
Note that all of the above names could, in
principle, apply to almost any Bucklin system;
but whichever one we pick, we'll arbitrarily
define it as being this system in particular.
Abd and anyone else who has an opinion: please vote among the above options.
Of the others: Bucklin-ER is also accurate. This
doesn't describe the amalganation tweak. C because it's geek-speak.
DAT is accurate, on the theory of it. B-. You
meant "Adjustment"? Too many words. D
Default F
----
Election-Methods mailing list - see http://electorama.com/em for list info
