Re: GRs, irrelevant amendments, and insincere voting

[Manoj Srivastava <srivasta@debian.org>]

On Tue, 4 Nov 2003 15:33:29 -0500, Branden Robinson <branden@debian.org> said: 

> On Sun, Nov 02, 2003 at 02:46:24AM -0600, Manoj Srivastava wrote:
>> On Sun, 2 Nov 2003 03:09:47 -0500, Branden Robinson
>> <branden@debian.org> said:
>> > Both Mr. DeRobertis and I interpreted the text quoted above as a
>> > personal attack.  How else is one to interpret "you are really
>> > contributing to infighting and intractability" and "you do not
>> > know how to be a team player"?  Are they in some way
>> > complimentary, or do they somehow provide insight into our
>> > Standard Resolution Procedure?
>>
>> If you are voting insencerely, and delibrately using majority to
>> defeat options that are not preferred, you are short sighted, not a
>> team player, and contributing to infighting and intractability.

> Not necessarily.  A person who ranks their preferences insincerely
> might simply feel they're using the system the way it was designed
> to be used.

	I challenge the use of the term insincerely here.

> In the recent "Disambiguation of 4.1.5" vote, for instance, while I
> haven't look at the tally sheet yet to see if anyone actually did, I
> would have to wonder if anyone who ranked "further discussion" above
> any of the other options was voting sincerely.

	Why? There are people who thought that the voting method was
 not discussed enough, despite having being through a similarly long
 period of discussion. Indeed, there is one ballot where the Further
 discussion was the only ranked option; so one person wanted more
 discussion rahter than choose _any_ option.

> I say this because the issue had dragged on for three years, we had
> a healthy discussion period, and I don't recall that any other
> interpretations or clarifications were raised.  (I.e., there were no
> proposed amendments that didn't acquire sufficient seconds to appear
> on the ballot.)

	There was very contentious discussion 3 years ago; I admit that
 this time around there was hardly any discussion at all.

> Instead, I find it more likely that people who ranked further
> discussion above any of the other options did so simply to
> disadvantage options they disfavored, and that they didn't actually
> prefer, say, another three years of discussion and delay.

	I could have said the same thing about voting methods. And
 does any discussion whatsoever requires a mandatory period of three
 years? (I would not like to talk about voting methods for another 3
 years, myself).

> Since the ballot encouraged people to do that very thing[1], one
> could easily argue that people who vote that way are doing exactly
> what the system intends them to do.

	The ballot did not encourage any particular vote as far as I
 can tell. It told people how to achieve certain results if that is
 what they wanted to do sincerely.

> I question the prudence of this design characteristic.  I think it
> promotes insincere balloting.  If you feel that this promotes

	Again that word. It allows people to voice strong objections
 to certain ballot options -- that does not enocurage people to not
 work towards a consensus, does not encourage them to not be team
 players, or any such thing. 

> short-sighted voting, discourages efforts to reach consensus, and
> contributes to infighting and intractability, then perhaps you
> should join me in subjecting the dynamics of "the default option" to
> closer scrutiny.

	My, my. Just the ability to voice a strong objection engenders
 such negative characteristics in people?  Without this ability we
 would suddenly change the mentality of project members?  We change
 people from team players to bad guys just by giving them a means to
 voice objections?

	Or, conversly, the way to get team players is to remove the
 ability of people to disagree? How very Stalinistic of you, Mr
 Robinson.

>> Are you asserting people voting insincerely and/or adding false
>> amendments are team players working for the good of all?

	I say that having a mechanism to voice dissent strongly, and
 having majority requirements that make us pay attention to dissent
 from a minority does not change peoples natures.

	Having a consensus merely by making it hard for people ti
 dissent is not something I desire, no.

> I'm not asserting that because I'm not sure that's the effect.  I do
> believe that given the current language we use on the ballot, that
> they can easily rank preferences insincerely and not feel that they
> are doing anything wrong or counterproductive.

	It is not insincere if they are genuinely expressing strong
 objections to a ballot option.

> Even in the absence of such ballot language, they might be right to
> feel justified in ranking their preferences insincerely.  I suspect
> the "default option" has an impact Condorcet's Method/Cloneproof SSD
> that has not been fully thought out.  I think the issue merits
> further exploration.

	You call it insincere, I call it expressing strong
 objections. Perhaps one should explicate this point on the ballot the
 next time.  But I do not think that this is a misfeature; indeed, I
 think this is a strong feature of the vote system.

	I really like the fact that we work with rough consensus.

> If it's acceptable to rank one's preferences insincerely, then
> what's wrong with working to get options on the ballot that one can
> use to arrange one's preferences insincerely and strategically?

	The goal is to come to a rough consensus. One should be able
 to express ones disapproval of options, but that is not the end goal,
 which you seem to be missing. The end goal is to WORK TOGETHER to
 make Debian the dest OS possible.


>> Colour me confused.

> I don't profess to have a full understanding of the dynamics of the
> default option.  I also don't profess that the insincere ranking of
> preferences is a phenomenon that is necessarily pejorative; maybe
> it's the way our system *should* work.  But if so, I think we should
> be a lot more up-front about that fact.

	I am trying to do so, as is my alter ego, Raul.


>> Would you make up your mind if you are promoting or condeming these
>> practices?

> Sure, after I've got enough data to do so.  I have a predilection
> towards suspicion of insincere ranking of preferences as a bad
> thing, as it appears you may as well, but I'm not going to let that
> take the place of a sober judgement grounded on real analysis of our
> voting system and Standard Resolution Procedure.

	Actually, I draw a distinction between insincere voting and
 sincerely expressing a strong objection.

>> No. I object to your blindly labelling people who genuinely prefer
>> further discussion to some option as insincere voters.

> I do not think I am doing so.  Obviously, people who *genuinely*
> prefer "further discussion" to other options and rank their
> preferences accordingly are not voting insincerely.  But that's a
> tautology.

	Ah. Now, how does one know that _any_ of these votes was not a
 sincere expression of the desires of the voter?

> The more interesting question to me is, can we reasonably have
> anything more than sunny-minded hope that people *aren't* ranking
> their preferences insincerely?  A think a useful perspective to
> adopt is an economic one: conceive of each voter as a
> self-interested actor who will fill out the ballot to maximize his
> personal benefit, and then consider how the system is likely to be
> used given that premise.

	I would rather not be a part of a project that works like
 that. If I thought that was a reasonable model of Debian, I would
 resign from the project.

> I don't assert that this "economic" perspective is truly descriptive
> of people's motivations, but it doesn't have to be.  For me, its
> utility comes in helping us to uncover what may be flaws in our SRP
> or voting system -- areas where our system is susceptible to
> "attack" by "bad actors".

	Or voting mechanism is a tool that allows us to make
 decisions, and in my mind it the fact that it allows us to
 objectively define how rough a rough consensus needs to be is a great
 idea. But it still is a tool of a project that has historically
 worked by rough consensus; and is not meant to be a used by people
 not of common cause.

	Indeed, any tool that is designed to work in a hostile
 conditions like you envisage would probably be too cumbersome for us
 to use; and I would rather not go there.

> I personally don't think it promotes an atmosphere of mistrust or
> suspicion in our project to try to understand what might be
> weaknesses in our collective decision-making procedures, but your
> mileage may vary.

	We have had three years of discussion on this already, I am
 perhaps a bit tired.

> I could be wrong about that, of course, which is why I'd like to see
> some more analysis.

	May I suggest you look at the voting methods ctte mail logs
 and see how most of this was looked at over and over again there?

>> If not, how can you allow people who prefer further discussion from
>> expressing their preference, in your zeal to stamp out so called
>> insincere voting?

> Since your premise is false (see above), this (implicit) conclusion
> does not follow.

	I am not sure I agree.

>> I am saying that people who vote in a manner that does not
>> represent their true preferences merely to defeat some option may
>> be "short-sighted", are "infighting",and displaying
>> "intractability", and not being a "team player".

> Do we have any way of knowing whether or not this has happened?  Do

	Do we have a crystal ball?


> we have reason to suspect it hasn't?  Does our system operate in
> such a way that the insincere ranking of preferences "rewards" the
> voter more than a sincere ranking?  If so, what is the magnitude of
> the impact of those rewards?  If insincere ranking of preferences is
> a strategy that works well for those who exercise, is it
> unreasonable to assume that we will see more of it as our voters
> become more accustomed to the system?

	These are social issues that the voting mechanism niehter
 hinders nor promotes.


>> You have objections to that? Tough.

> Not necessarily, but I am not sure that reproaching people in strong
> language is a sufficient disincentive to the phenomenon, especially
> when they always have plausible deniability.  As you said above:

>> No. I object to your blindly labelling people who genuinely prefer
>> further discussion to some option as insincere voters.

> Do either of us have a good way to perceive the genuine intentions
> of the voters without reference to the ranking of their preferences
> on the ballot?

	No, we do not.


> If not, does the presumption of sincere ranking of preferences
> provide a sufficiently foundation upon which we can ground a voting
> system that is intended to produce results that are satisfactory to
> the electorate?

	The voting method does not change the nature of the project,
 or the people, or the way we make decisions.  If we have social
 issues that prevent us reaching decisions, or coming to an agreement
 with whatever prepoenderance of near unanimity required, the voting
 method shall not help us out there.

>> > Let it be noted that you're the person introducing the term
>> > "dumb" into this conversation.  I personally am not sure its're
>> > warranted, due to its crudity and high emotive content.
>>
>> Calling attempts to make your less preferred option fail majority
>> by mistating your preferences a smart policy is wrong; I do think
>> it is a dumb idea in the long run.

> Identifying activities as "wrong" and "dumb" doesn't stop them from
> happening.  Might it be wise to avail ourselves of procedural
> disincentives to "wrong" and "dumb" behavior as well?


	No. The procedures can not, and should not try, to make us
 better than we are. At best, procedires should be transparent, and
 allow us to be ourselves, and express our true desires, as far as
 possible. 

	manoj

-- 
The more cordial the buyer's secretary, the greater the odds that the
competition already has the order.
Manoj Srivastava   <srivasta@debian.org>  <http://www.debian.org/%7Esrivasta/>
1024R/C7261095 print CB D9 F4 12 68 07 E4 05  CC 2D 27 12 1D F5 E8 6E
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C