IRV and competing methods. (fwd)

[MIKE OSSIPOFF]

Forest--

I hope you'll forward this to Xander, and to anyone else making
his claims. I'd Cc it to him, but I don't know if that would be
proper, replying directly.


>Xander wrote (in part):
>
>
> > I see the potential failings of IRV, but it does allow us to show our
> > support and greatly reduces the odds, if not eliminate them altogether,
> > of
> > spoiling.  My question about IRV is whether the failings you point out
> > are
> > as glaring in practice as they are in theory.  I don't know enough about
> > how IRV has worked in the real world to even venture a guess, but it
> > seems
> > to me that the kind of examples that reveal it's faults are based on
> > some
> > unlikely assumptions, such as that all first choice Dems would rank R
> > second.

IRV's promoters certainly make that claim, will tell you that
all the failure examples are unlikely. But it's been reported that
it's common for small party members in Australia's IRV elections to
dump their own candidate in order to vote one of the big-2 in 1st place.
They say they do that in order to not waste their vote. What does it
take to show the IRV promoters that IRV has a problem? We demonstrate
how the problems will inevitably happen. Voters in the country where
IRV is used for important political office dump their favorite, just
as we've been telling you people will often need to do.

But let's talk about how common or rare IRV's problems are.

IRV promoters often boast that if there is a majority faction that has 2
candidates who would split the vote, IRV won't let that happen, because
when one of them gets eliminated, his votes will go to the other
candidate of that faction, and so that majority will still carry the
election and elect someone from their faction.

Sure, but every example has at least 2 sides. What about the people
not in that faction?  You see, you can't be sure that you're in
such a "mutual majority". What if you're one of the other voters?

One thing for sure, your candidate can't win. He doesn't have a majority
because that other faction has a majority. And they'll transfer to
eachother. Your candidate can't possibly win. And what if you have
a significant preference for one of that faction's candidates over the
other? Should you vote sincerely? Of course not. It would serve no
purpose. Your favorite can't win, but you could help save the majority
faction candidate whom you prefer beat the majority candidate whom
you don't like as much.

The only strategy that makes any sense is for you to vote that
preferred majority-faction candidate in 1st place, dumping your favorite.

You might say "So what if I dump my favorite--he can't win."
Yes, wouldn't it be nice if we always knew those things. Strategy
would be a lot more reliable and simplet then, wouldn't it.

But you don't know that. You may have notices how cautious and
overcompromising most progressive voters are. If avoiding the "greater
evil" is the important thing to you, and you feel that the better
majority-faction candidate is going to get eliminated, and the worse
one will win, then it's to your advantage to vote the lesser-evil
majority-faction candidate in 1st place, dumping your favorite.

And then it might turn out that you misjudged, and your favorite had
a win, and it wasn't the scenario that I've just described.

IRV made you strategically dump your favorite by voting someone else
over him. That will never happen with Approval, and it isn't a problem
with Condorcet, where it could only happen under rare circumstances
and not in an important way.

Like Approval, Condorcet also makes some guarantees that no other
method can make. They're described at http://www.electionmethods.org
at the "Technical Evaluation" menu link. IRV fails every one of the
criteria that are described there, criteria intended to measure for
how well a method avoids the lesser-of-2-evils problem and protects
majority rule.

Likewise, when the voter median candidate has the most support, and
support tapers gradually away from that voter-median position,
we can expect eliminations to start at the extremes, and send transfers
inward, until some of those slightly weaker nonmedian candidates are
built up by those transfers enough to eliminate the median candidate(s).
This configuration where the voter-median candidate has the most support,
and support tapers gradually away from him, is really what we'd expect
to be typical. In other words, based on this and the previous example,
IRV's problems will be typical, not rare.

Mike Ossipoff



_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com