[EM] Condorcet Criterion for plurality.

[MIKE OSSIPOFF]

>Yes! I've thought more about it and there are additional problems.  You 
>see,
>part of my point about translating voting systems into preferential ones 
>was
>for the purposes of classifying what is a sincere vote in each system.  
>What
>about approval?  The preferential translation is: you vote with 1's and 
>then
>truncate your vote at any time.  However, you obviously don't like every
>candidate you mark as much as every other one.  As a result, you're
>expressing a false preference (under Mike's definition).

No. I haven't said that leaving a sincere preference unvoted is
voting a false preference. I've only referred to that as leaving
a sincere preference unvoted. And my definition doesn't call that
insincere in Approval. My definition says "...or leave unvoted
a sincere preference that the balloting system in use would have
allowed him to vote in addition to the preferences that he actually
did vote."

When you vote for Gore & Nader, failing to vote your preference
between Gore & Nader, you aren't insincere by my definition,
because, to vote Nader over Gore, you must not vote Gore over
Bush. The balloting system wouldn't allow you to vote your
Nader>Gore preference in addition to the Gore>Bush preference that
you actually voted.

>Cumulative is even
>more complicated.  You might not express a preference between two
>candidates, because you want to cumulate your votes on your favourite
>(failing Mike's definiton).

Exactly the same situation with single-winner Cumulative.
Not voting a preference between Gore & Bush because you want to
vote only for Nader isn't insincere by my definition, because
voting Gore over Bush isn't allowed in addition to voting Nader
over Gore. The unvoted preferences can't be voted in addition to
the voted one, and so there's no insincerity in not voting the
unvoted one.

One detail that should be specified for my definition is:
voting a preference for X over Y means voting X over Y, as I've
previously defined that.

>You might give two candidates the same number
>of points, even though you don't like them exactly the same (fails >Mike's
>def. again).

Say you vote for X & Y even though you don't like them the same,
and you don't vote for Z. You are leaving your X>Y preference
unvoted, but that isn't insincere by my definition, because
you can't vote your X>Y preference in addition to your X>Z
and Y>Z preferences.

So none of those ways of voting in those methods are insincere
by my definition. They won't be insincere in the definition that
I'll later write or adopt either. Until I write or adopt a new
definition of sincere voting, the one that I've been using serves
its practical purpose fine, in criteria. But it's becoming aparent
that what I stipulate & call "sincerity" in criteria should
probably be renamed, and that, as Craig L. & Norm say, sincerity
ideally should be defined differently from how I've defined it.
But my sincerity definition doesn't call the ways of voting that
Craig has described insincere. Where my definition differs from
what sincerity means to us is when it calls you sincere if you
refuse to express an opinion.

>Part of the problem is that truncating your vote is part of
>the voting design in Approval and (single winner) Cumulative.

That isn't a problem with my definition. My definition differs
from what we mean by sincerity, by calling people insincere when
they're forced to vote false preferences, in Plurality &
ordinary IRV (the kind that the IRVies are proposing).

Mike Ossipoff




_____________________________________________________________________________________
Get more from the Web.  FREE MSN Explorer download : http://explorer.msn.com