reversed rankings

Mike Ositoff ntk at
Fri Oct 30 19:04:45 PST 1998

> On Wed, 28 Oct 1998 18:28:42   Mike Ositoff wrote:
> >
> >I'd like to answer the reversed-rankings nonsense, though I
> >already have.
> >
> >Blake claimed that VA will pick the same candidate if you
> >reverse all the rankings. Or at least implied that it will
> >do so often under ordinary plausible conditions. Nonsense.
> >
> I never made any statement about which methods pass and which
> fail this criterion.  In fact, Condorcet (EM) and IRO fail,
> Schulze and Approval pass.  Margins or VA is irrelevant.

Ok; I thought you'd used reversed rankings to argue that
Margins was better than VA.
> >With a 1-dimensional policy space, one of the twe most extreme
> >candidates will win with a 1st choice majority if rankings
> >are reversed, since everyone will have an extreme as last choice.
> >
> >Markus said that Schulze(VA) can't pick the same candidate
> >when rankings are reversed. Do you disagree, Blake?
> No.
> >
> >Blake says that if that happens, it means that VA says that
> >the best candidate is also the worst. Nonsense again. MEthods
> >don't pick the best or worst. Each voter has his best &
> >worst, and the method picks based on certain more meaningful
> >& concrete & verifiable criteria than picking the best.
> Well, of course we differ on this point.  I think that the
> point of election reform should be to get better government.
> Do you have a different goal?  If so, how do you justify it.

Conceivably, if voters had bad enough judgement, better voting
systems wouldn't produce better govt. Of course it's difficult
to claim that even with the worst electorate, a worse voting
system could produce better government. IRO's nonmonotonicity
isn't that dependable.

So if we get better govt, that comes as a result of better
voting systems, though it isn't guaranteed to.

So yes, I have a different goal than better govt, because
a voting system can't decide what's better, only what is
more democratic, or more consistent with some criteria
that we consider important. Sure, since bad voting systems
can't be expected to do anything reliable toward better govt,
and since better voting systems just might, then indirectly
better govt is my goal also.

> I think you're right that whether the method picks the best
> guess for best candidate is not verifiable.  It is, however,
> falsifiable.  Clearly, if the method would pick the same
> candidate for best and worst, it is wrong in one direction.

The fact that some methods pick the same candidate with reversed
rankings doesn't mean that they're picking him as best & worst.
To believe that, you have to believe that methods pick the best
if rankings aren't reversed and that they pick the worst if
rankings are reversed--and that's what I'm saying methods can't
be assumed to do. You're imputing a power of governmental judgement
to voting systems that they don't have. Only people know good
govt from bad.

> ---
> Blake
> -----== Sent via Deja News, The Discussion Network ==-----
>  Easy access to 50,000+ discussion forums

More information about the Election-Methods mailing list