W3C home > Mailing lists > Public > public-w3process@w3.org > August 2014

Re:voting simple illustration

From: <chaals@yandex-team.ru>
Date: Mon, 25 Aug 2014 21:13:08 +0200
To: David Singer <singer@apple.com>
Cc: public-w3process@w3.org, Brian Kardell <bkardell@gmail.com>
Message-Id: <622591408993988@webcorp02f.yandex-team.ru>

> Yes, we are in a similar boat. It is indeed a nice example.

No, it is an unfortunate consequence of the years-long reluctance to consider the question, or to try and understand the problem in theoretical terms instead of being faced with a bad decision to make.

> In the voting scheme proposed, you would rank A,B,
> <no other candidate>, as I understand it.

Only if you really think C and D are so bad they *shouldn't get elected at all*. Otherwise you just rank them, or leave the rest of your vote blank.

> Then if others rate
> B,A,<no other candidate>, your vote rolls over to B who gets elected,
> but you don’t roll to C or D.

It's equally possible that you have a preference between C and D - and that you consider them qualified even though you have a ranking in preference (i.e. you would be happier to see them win than suggest there needed to be a new call for candidates).

> As I said before, I think that having the <no other candidate> be implicit in the voting instructions (“rank the candidates you would be willing to see elected, in your preferred order”) is simpler on small-brain voters…

It means you can't distinguish between  "I don't really care after this set of preferences", and "I would prefer the election to be declared invalid than have these candidates win". Since that's hopefully a minority case, it makes a bad default.



> On Aug 21, 2014, at 9:56 , Brian Kardell <bkardell@gmail.com> wrote:
>> All,
>> It seems to me that a lot of illustrations and Chaals' descriptions
>> have been fairly complicated about strategic voting and many
>> candidates, but it strikes me that the current TAG election is
>> actually an excellent simple illustration.
>> We have 4 candidates for a single seat. Of course, they are all
>> qualified, but we're voting here so lets be honest, you're going to to
>> pick someone you think would be best at this stage of things. In my
>> mind, I'm having trouble picking between two candidates which I both
>> see as having more pros than the other two, and I slightly prefer one
>> to the other in a single regard. The point is, I'll be considerably
>> happier if either or my preferences win, and considerably less so if
>> the others win. The strategic voting question and rationale for
>> preferential expression is all right there.
>> If I think that A is slightly better than B for some subtle reason,
>> but that both A and B are better than C or D, there is no way to
>> express that. If B is better known, they have a better chance of
>> winning. Do I throw in with them because I think that they have the
>> best mix of "can actually win and is the best candidate?" Do I talk
>> to my friends about "A or B"? This means that they'll split a vote
>> whereas very possibly C or D won't, and C will walk away with it...
>> Even though very possibly most people agree with the idea that both A
>> and B are preferable to C.
>> At the end of the day, this is kind of silly because the above isn't
>> that complicated - if I could say "I slightly prefer A to B, but I
>> significantly prefer A and B to C and D" then we'd yield the result
>> that the most people are the most happy with.
>> --
>> Brian Kardell :: @briankardell :: hitchjs.com
> David Singer
> Manager, Software Standards, Apple Inc.
Received on Monday, 25 August 2014 19:13:39 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:35:11 UTC