[EM] Demorep & Approval

[MIKE OSSIPOFF]

>Sincere
>
>51 A (100) B (99)
>  3   B (100)
>46 C (100)
>
>100
>
>B wins (54) using simple Approval even though A has a *real* first choice
>majority.

In this example, the A voters' mis-estimate would be plausible, since
there's only a small difference between A having a majority and A
having being outpolled by C. But this scenario that makes that
mis-estimate plausible is the kind of scenario where IRV fails--
a very low-favoriteness middle compromise.

So if an IRV promoter wanted to show a reasonable mis-estimate in
Approval, he'd have to use the kind of example that's been the
standard IRV badexample. For instance, if we make B just a little bit
stronger in the above example, then B is the median candidate, the
sincere CW, and the example is an IRV badexample.

Additionally, notice that Approval errs in the above example by
electing Middle when an extreme has a 1st choice majority. When everyone
has the same erroneous predictive information, that's the only way
that Approval will fail. On the other hand, when IRV fails, it fails
by jumping to an extreme. IRV's way of failing differs by being more
drastic, more likely to be disastrous, more likely to result in irreparable 
damage--when it fails in the way that it will if everyone
votes sincerely, as the IRVies claim people will vote.

Mike Ossipoff

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