[EM] 2-Balloting Approval

[MIKE OSSIPOFF]

> >But if someone had a better chance of defeating the front runner you
> >prefer less, wouldn't that someone be more of a front runner than the one
> >you preferred?
>

I reply:

>I just meant frontrunners in the 1st balloting. Maybe enough people
>who like the lower frontrunner better than the higher frontrunner
>(and probably prefer someone else as favorite) expected that they
>would need the lower frontrunner, which was what made him a frontrunner
>in the 1st balloting. But now the 1st balloting has not only confirmed
>that they can't do better than the lower frontrunner, but it's shown
>that they have to compromise still farther down their preference rankings,
>and vote for someone even worse than the lower frontrunner,
>someone between the 2 frontrunners, someone who didn't get many votes
>because people didn't realize that they need to compromise that far.
>
> >
> >The two front runners are A and B.  You prefer A, but B is ahead in the
> >polls.  So you vote down to C who is below A on your list of preferences
> >because you think C has a better chance than A of beating B.  Wouldn't
> >that mean that A wasn't really one of the front runners?
>
>A was a frontrunner in the initial balloting because he came in 2nd,
>because people expected him to be the candidate whom they could make beat
>B. It turned out that he wasn't quite a big enough votegetter, and that
>it looks like it's necessary to compromise farther. Hopefully we can
>make C into a frontrunner in the 2nd balloting, one who is top
>votegetter.
>
>Say in the 1st balloting A gets 300, B gets 500, C gets 100, & D
>gets 200. Your ranking is X,Y,Z,A,C,D,B. You, and many others, voted
>down to A, thinking that he was the best compromise they could get.
>But evidently he can't beat B, and can't win. But if those same people
>, based on info from the 1st balloting, vote down to D, they can make
>D beat B.
>
>I think a preliminary balloting is the answer for Approval with very
>few voters. Maximizing expectation based on probability estimates is
>quite complicated, because we no longer can safely make the approximation
>that 2-way ties are the only ones. With a preliminary balloting by
>Approval, or Plurality (if there's a 1-dimensional issue-space) or
>rankings, where everyone participates in both ballotings, we can avoid
>the need for using probabilities to try to maximize expectation.
>
>No one is elected in the 1st balloting unless they get a vote total
>at least equal to half the number of voters. If it's rankings, then
>they'd have to be ranked 1st by at least half the voters.
>
>I didn't realize before that, with very few voters, expectation
>maximization based on probability estimates is greatly complicated because
>ties are no longer just 2-way ties.
>
>Mike Ossipoff
>
>
>
>
>
> >
> >
>
>_________________________________________________________________
>Get your FREE download of MSN Explorer at http://explorer.msn.com
>

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