Mock Election & Results

[Mike Ossipoff]

Steve Eppley writes:
> 
> Mike O wrote [in the ER list]:
> -snip-
> >At least in Approval there's never need to actually falsify a
> >preference ordering, which seems to put it ahead of most
> >rank-balloting count methods. 
> -snip-
> 
> That's not the way I view Approval.  Here's an example Approval
> vote, expressed using a restricted form of my ordering syntax:
>    A=B=C=D > NOTB > E=F=G=H
> 
> If this voter's preference ordering is actually 
>    A>B>C>D > NOTB > E>F>G>H
> then I'd say this voter has to falsify his preferences, and in a way
> which matches Mike's definition of drastic defensive strategy.  So
> what does Mike mean? 


True, Approval does require drastic defensive strategy, and that's
why I don't consider it as good as Condorcet. But, with Approval,
that drastic defensive strategy merely takes the form of
insincere equal ranking, rather than outright reversal of
preference orderings, as are sometimes required in Copeland,
Regular Champion, & MPV, & random circular tie solution.

Also, I realize that not everyone would agree with me in
comparing all of those rank-balloting count methods to Approval
based on whether they can require defensive order-reversal. But
I've talked about MPV's faults on ER, a while back, and I really
feel that it's worse than Approval in other ways too.

I'd be willing to not pursue my claim that Regular Champion,
, Copeland, & random circular-tie solution aren't as good
as Approval, because that could be a more controversial claim
than when I say that those methods aren't as good as Condorcet,
and could distract debate from the more important issue of which
method is the best one to recommend. Proponents of all those
rank-balloting methods agree that some rank-balloting method
is better than Approval, and of course I agree with that, if
the method is Condorcet.


Mike


> 
> ---Steve     (Steve Eppley    seppley at alumni.caltech.edu)
> .-
> 


--