[EM] A few Voter's Choice comments

[MIKE OSSIPOFF]

As I mentioned in a reply, Approval would be better as the "base method"
of Voter's Choice than Plurality would. I suggested Plurality because
it's official in the U.S., and because it has no advocates here, meaning
that it can't be said to favor anyone.

But Approval is better, because if your designated method missed the
sincere CW by voting too far down your preference ordering, then
Approval minimizes the mistake, by voting for your more-liked
alternatives too.

The way Approval would be used is: You automatically give an Approval
vote to the altenative chosen by your designated method, and to every
alternative that you've ranked higher (or rated higher, if you
didn't vote a ranking).

Also, anyone who prefers to should be able to simply indicate that
they're casting an Approval ballot that says where they want their
Approval votes to go. In other words, Voter's Choice is optional.
A voter can opt to not use it for the placement of his Approval votes.

Likewise, if in a Public election, Voter's Choice were being used with
Plurality as the base method, a voter would have the option of just
casting an ordinary Plurality ballot and specifying exactly where his
Plurality vote will go. That way, Voter's Choice is optional, and no
one can have anything to complain about.

Tom suggested Runoff Approval. Well, the rankings and Approval ballots
(actual & inferred) give sufficient information to do a Runoff Approval
count, without a 2nd balloting. So Tom could designate Runoff
Approval. Likewise, there's sufficient information for an Approval-Seeded SP 
count.

Though I consider ordinary Approval better for public elections, for
a number of reasons, here Runoff Approval probably does a better job
of picking a sincere CW, picking the best compromise you can get.

What method does the very best job? Condorcet. No matter what method
you prefer, it would be the best idea to designate a Condorcet
version (PC, Cloneproof SSD, BeatpathWinner, etc.) as your designated
method. Of course if everyone did, it would just be a Condorcet
election.

Voter's Choice is probably the best way to vote when a definite choice
is needed, and people don't agree on the method. For instance, in the
vote on which polling topic to use, a definite choice is needed.
It has the advantage of demonstrating every method that someone
designates.

I haven't heard any alternatives, except for Don's suggestion of
5 compulsory ballots, repeated up to 5 times.

In the subsequent actual poll, a definite choice isn't really being
made, since no one's sending out for pizza or buying ice cream, and
so what's of interest is to compare the methods' results, and what
it's like to use the methods--those things are really essential for
a mailing list that studies voting systems, and so demonstration polls
are very valuable for such a list. Even though a definite choice isn't
needed in the actual poll on pizza toppings or ice cream flavors, etc.,
I'll still do a Voter's Choice count, just out of curiosity. If people
don't designate a method in that poll, I'll use their designations from
the initial poll on pollng topic. But it would be better to designate
a method in the actual poll too, to find out what would have been
chosen had it been necessary to make a definite choice, without an
agreed-upon voting system.

Mike Ossipoff

_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com