[EM] Problems with finding the probable best governor

[Bart Ingles]

MIKE OSSIPOFF wrote:
> 
> One of my answers to the disliked-middle argument is that if he
> isn't really middle, people will know it. For instance, I wouldn't
> have ranked him, or Bush, in a rank-balloting election. At least one
> faction will know that he isn't a middle compromise.
> 
> Or, if he _is_ middle, then, for Democrats & progressives, he'd
> be somewhat better than Bush, and for Republicans, he'd be somewhat
> better than Clinton. So if that compromise wins, nothing wrong with
> that, right? If he were really between Clinton & Bush.

That kind of compromise wouldn't satisfy me.  But then I still suspect
that a utility-maximizing voter would have incentive to truncate, so it
may not be a major concern.


> Or he could be middle, but have some sort of nonpositional disutility.
> Maybe he could be perceived as insincere (and therefore maybe not
> really middle). Or unstable, or crooked, etc., or just not liked.
> But, again, that would keep him from getting those majorities over
> Bush & Clinton. If he isn't really better, in spite of being middle.
> If he isn't better, they won't rank him over.

Nonpositional disutility seems as good a term as any -- why should every
attribute belong to a bi-polar scale?  What about things like
intelligence, honesty, competence, experience, ability to communicate,
etc.?  It seems to me that some attributes wouldn't have adherents at
either end of the scale, but instead would be seen as positive (or
negative) by all voters.

>From an individual voter's perspective, it would make no difference
whether he was compromising by choosing someone farther away on a
bipolar or monopolar scale.


> When a good voting system (Condorcet or Approval) is used, there
> will be plenty of candidates, plenty of selection & choices. And
> the voter-median point will be a popular & crowded position. How
> likely is it that the only candidate there will be someone unknown
> and even more insincere than the other middle candidates?
>
> In any case, he wouldn't remain unknown for long. There'd be plenty
> of writers & commentators who'd dig up information. The campaign
> organizations of his rivals would publicly expose his real nature.
> 
> But the possibility that Condorcet could elect a low-utility middle
> candidate could be tested by Norm's simulation, by printing-out
> , for each voting system, not only its average SU over the run,
> but also it's maximum & minimum utilities. But I'd expect that
> a median candidate really will have better SU, in those spatial
> simulations that don't include nonpositional disutility. And if
> you added that in--if, for some middle candidates, we added a certain
> constant disutility to each voter's rating of that candidate--I'd
> expect it to become unlikely for that candidate to win. But it could
> be investigated in the simulations.

That's something that I would like to try, if I ever get time to learn
Python.  It would be interesting to see, both with and without a
nonpositional utility component.

Bart