On Jan 10, 2010, at 3:42 PM, Juho wrote:
On Jan 10, 2010, at 10:23 PM, robert bristow-johnson wrote:
On Jan 10, 2010, at 2:40 PM, Jonathan Lundell wrote:
This is a point that bears repeating, since it doesn't seem to
sink in. It's much to easy to casually assume that ballots cast
under one system (in this case IRV) can be recounted under some
other rule with the assumption that voters would have cast the
same (or at least equivalent) ballots under that other rule.
but i think that it is reasonable to "casually" assume that the
ranked-order ballots marked for IRV would, for the most part, be
the same if the election were to be decided by Condorcet rules.
that, plus the Freedom of Information laws, allows us to say what
would have happened with Condorcet rules. the tabulation rules
are different, but the ballots are the same.
I believe the voters would have voted in the same way in Condorcet.
That is because they did not understand how IRV works nor how
Condorcet works (or at least they didn't understand what strategic
opportunities there are). They were told that in IRV you just rank
the candidates sincerely, and that's what they probably did. I have
not heard of any (IRV specific) strategic advices that would have
been given to the voters before the election.
probably not in 2009.
So I believe the votes were quite sincere rankings. In the next IRV
election things might be a bit different.
i think so. just like the "Libruls" learned that their vote for
Ralph Nader (particularly those in Florida) ended up electing the
Abomination in 2000 (which cured them of any idealistic notion of
voicing their vote for 3rd party candidates), the GOP Prog-haters in
Burlington have learned that they were indispensable to electing the
candidate they hated (by their sincere vote for the candidate they
liked). IRV literally transferred the burden of having to think
strategically from the liberal majority to the conservative minority
in Burlington.
--
r b-j [email protected]
"Imagination is more important than knowledge."
----
Election-Methods mailing list - see http://electorama.com/em for list info
