[EM] The Green scenario, and IRV in the Green scenario, is a new topic here. Hence these additional comments. Clarification of position and why.

[Kristofer Munsterhjelm]

On 02/04/2013 02:40 PM, Peter Zbornik wrote:
> Being a green party member (although a Czech one and not US), I would
> advocate only the top-two-run-off
> variant of IRV, i.e. elimination of the candidates and transfer of
> votes until two remain, no quota for election (or quota=100%) except
> for the case where one candidate has more than 50% of first
> preferences.
>
> The top two candidates would meet in a second round in IRV.
> A candidate would be elected if he/she would get more than 50% of the votes.
>
> Empty votes would count as  valid votes in both first and second round.
>
> If no candidate would be elected in second round new elections would take place.
>
> The advantages of the proposed election system are
> 1) the voters are given a chance to concentrate only on two candidates
> in the second round, and are thus allowed to change their preferences.
> 2) blank votes together with IRV might make the candidates less
> polarized, as, given a large number of blank votes, the candidate with
> the highest number of votes in the second round would have to rely on
> the second preferences of the voters for the opposing candidate in
> order to get 50%+ votes.

Perhaps this method would work for runoffs if you can get a more 
sophisticated base method through, say for internal elections:

- Run a single-winner election using your method of choice. Call the 
winner w_1.
- Use a proportional ranking method to determine the second runoff 
candidate w_2 so that the virtual council {w_1, w_2} represents as much 
as possible of the population.
- Have a runoff between w_1 and w_2.

If w_1 is a strong winner, he'll win in the runoff. If he's a weak 
winner (e.g. the "bland politician being everybody's second choice" 
scenario), w_2 wins.

In IRV, this would be like running two-member STV where the IRV winner 
is barred from being disqualified.

There could be a problem, though, in a society that has a bland centrist 
politician and strong left- and right-wing candidates. Since the runoff 
can only hold two candidates, either the left-wing or the right-wing 
candidate would be disqualified; and if the bland politician is 
sufficiently bland, then the wing candidate would pretty much win by 
default. IRV "solves" this by not letting center-squeezed candidates win 
in the first place. Another option is to have multiple candidates in the 
runoff, but then the simplicity and strategy resistance properties of 
the second round go away.