>>>>> "Barak" == Barak A Pearlmutter <barak@pearlmutter.net> writes:
Barak> On Mon, 19 Apr 2021 at 16:35, Sam Hartman <hartmans@debian.org> wrote:
>> I think we need voting reform around how the amendment process
>> works and managing discussion time ... ... Preferences can be
>> of different strengths. .... Which is to say that the gaps
>> between preferences might be relatively weak.
Barak> Sam, you make an excellent point about gaps between options,
Barak> and that a ranking does not show the strength of
Barak> preferences. Like, I might prefer ALPHA >>> BETA > GAMMA
Barak> while you prefer ALPHA > BETA >>> GAMMA.
We agree so far.
>So if it's down to
Barak> ALPHA vs BETA, my vote should shift things more than yours,
Barak> while if it's down to BETA vs GAMMA, your vote should shift
Barak> things more than mine. And
That's a big jump, and I don't think I agree.
At least not when you phrase it that way.
Why should my preference matter less just because it's weaker? It's
still my preference and I'm attached to it very much:-)
You then later talked about a voting system in which we somehow assigned
numerical scores to the result. I'll admit that as a theoretical
exercise I'd love to explore something like that.
I think it would be years before it could be debugged, and reviewed, and
all the weaknesses explored enough that I'd want to consider it for
Debian.
I was actually trying to say something different.
I think we're debating about what properties our voting system should
have now.
I think we've left the math behind a while ago, and are debating what's
desirable.
My claim is that our voting system seems to do the following:
1) Ignoring super majorities, if there is a winner of the pairwise
elections, we choose that as the winner of the election. I think no one
has disputed this as a desirable property. People have argued about
whether they'd be willing to give up this property to get something
else, but I think at least in this discussion this has not been
controversial as something we desire.
2) We let voters indicate whether they consider an option acceptable.
That is we let them decide whether they would prefer that option be
selected or whether they would prefer the decision making process
continue. We never select an option if most voters would prefer to
continue the decision making process to selecting that option.
3) If there are options that a sufficient number of voters (often a
simple majority) prefer to continuing the decision making process, we
will pick one of those options. There are several points in the process
where the desire to pick an option if there is one that defeats FD is
strongly encoded.
4) No really, we're quite serious about wanting to be done if there is
something that a majority of the voters consider acceptable. So much so
that there are situations where we'll pick a less preferred option just
to be done because the more preferred option requires a supermajority
it didn't meet. As an example, we might pick a simple statement over a
constitutional amendment even if more people prefer the constitutional
amendment. This is only interesting if a majority of voters consider
both the constitutional amendment and the simple statement acceptable.
The simple statement gets picked rather than the constitutional
amendment if it was not preferred by a sufficient super majority of
voters.
And yes, I have high confidence that the above were intentional
decisions. We may disagree; we may change our minds. But this has all
been debated time and time again, and for the most part on these aspects
of the voting system people were aware.
And certainly by the time we considered revising the voting system (I
think that was around 2003) we were very aware of these issues.
My take away is that the voting system is designed with an implication
that there is a huge preference gap between acceptable and unacceptable
options, and that by the time the GR procedure is called into play, it's
better to have a decision if that is at all possible.
That certainly mirrors my experience as a voter.
I generally find I am able to find a line of acceptability on most of
our ballots.
And I find that above that line I really would be able to accept any
outcome.
Yes, I want my vote to be counted.
And if there is a pairwise winner, I want that.
But if there is a cycle, well, okay, pick something.
We have a strong history of being able to get "no statement" options on
the ballot when we need them.
People aren't afraid to vote for them.
This is at least the second election where such an option won.
Attachment:
signature.asc
Description: PGP signature