[EM] Re: equal rankings IRV

[Bart Ingles]

Kevin Venzke wrote:
> 
> --- Bart Ingles wrote:   [...]
> > I also have serious doubts about whole-vote ER-IRV, mainly over whether
> > the Duvergerian equilibria would still be strong enough to maintain a
> > two-party system.  If so, then the differences between top-two, IRV,
> > ER-IRV(fractional), ER-IRV(whole votes), or simply disqualifying all but
> > the top-two primary winners from the general election, are largely
> > academic, at least in U.S. partisan races.
> 
> The difference between ER-IRV(whole) and those other methods is that the
> voter can just submit an approval ballot if he wants.  ER-IRV(whole) fails
> FBC but not, I think, in a very predictable way.  Other than this, there is
> no incentive to raise compromises above preferred candidates.
> 
> If this reasoning doesn't convince you, I'd like to ask what it is about
> ER-IRV(whole) that you think would still maintain a two-party system.

I'm unconvinced because I don't believe that FBC failure is particularly
rare.  It seems to me that once you have identified the "lesser-evil"
candidate best able to defeat the "greatest threat" candidate, optimal
strategy would be the rank the lesser-evil first ahead of anyone else. 

I would only include my favorite in the top rank if I believed that my
favorite was likely to beat all disliked candidates.  But in that case,
why rank anyone else first?  Maybe for insurance, I suppose, but I think
that ranking multiple candidates first would be fairly rare compared
with approval voting.

What would convince me otherwise would be a set of strategy equations
comparable to those used for calculating optimal strategy in approval
voting, or possibly simulations with ER-IRV(whole) showing that
"lesser-evil or better" strategy is as good or better than
"lesser-evil-only" in terms of social utility efficiency.

Bart