[EM] no opinion option

James Green-Armytage jarmyta at antioch-college.edu
Sat Jun 12 01:44:01 PDT 2004


Jobst Heitzig <heitzig-j at web.de> writes:
>Back from Persia I still wonder why you still don't consider ballots
>that allow voters to specify their preferences more accurately... Almost
>all the methods in the "Ranked choice Voting" category can easily be
>applied (or easily modified to be applicable) to ballots on which the
>voter specifies for each pair of options whether (a) she prefers the
>first, (b) she prefers the second, (c) she considers both equivalent, or
>(d) she is undecided about this pair or wants to abstain from the
>decision over this pair -- in other words: they can specify any binary
>relation on the set of options (including linear orderings aka
>"rankings", of course). Otherwise you simply cannot express that you
>prefer A to B and B to C but don't want to evaluate D at all because you
>would have to put D either above or below A,B,C, or between two of them,
>or equivalent (aka tied) to one of them.

	Hi. On his Condorcet Internet Voting Service (CIVS) web page, Andrew
Myers uses a kind of ballot where you can give any candidate a rating of
"no opinion". The effect of this is that your vote will not count in any
of the pairwise comparisons involving that candidate. I think that this is
equivalent to what you're talking about, without the hassle of having to
write your preference between each individual pair. (Consider that if
there were ten candidates, you would have to vote between 45 separate
pairs, and so on.)
	I don't like the idea of using this sort of ballot for public elections,
because I want it to count heavily against candidates if few people have
heard of them at all. However, there are some uses for such a ballot. On
Andrew's page there is a vote for the best movie of all time, and it seems
only fair that you should vote no opinion on a movie if you haven't seen
it.
	The address of the page is
http://www5.cs.cornell.edu/~andru/civs/public_elections.html
	It's well worth checking out.

best,
James





More information about the Election-Methods mailing list