[EM] Re: Use a "turkey" filter

[Stephane Rouillon]

It seems I do not believe voters behavior
obeys to your description:

Even in the case described below,
I do not think anybody can predict if
party A's voters will follow party A's recommendation.
If party B is party A's ennemy, but for me they are ranked
A then B then others, I would not obey to politicians
and would be my own judge.

I think a great majority of people are not party members of
any kind, they have their own preferences and would
apply them no matter what their favorite party official preference...

Condorcet is the best according to me and what you call turkeys
are just concensual winners. Some "middle" runners are just
consensual loosers, it happens too.  And every candidate has
the opportunity to make his(her) subject the main one, hopefully
it would be a subject that puts him(her) as a consensual candidate...

Stephane Rouillon

Anthony Duff a écrit :

>  >so you can be sure that if  the
> > method used was
> > Condorcet (especially with truncation allowed)  they
> > would not be
> > advising  voters to elect so-called  "turkeys"
>
> How can you be so sure?  Is the following hypothetical
> so unrealistic?:
>
> In a highly polarised campaign, a strong candidate
> emphasizes the negatives of the incumbent and repeats
> the catchy message "Put [the incumbent] last".  The
> incumbent replies with similar negativity, "[the
> challenger] is dangerous, put him last"
>
> If two large minorities put each other’s favourite
> last then it becomes very likely that a minor
> candidate will be a CW.  And seeing as the campaign
> was so polarised, that minor candidate may have been
> subjected to very little scrutiny.
>