[EM] Blake: criteria

MIKE OSSIPOFF nkklrp at hotmail.com
Thu Apr 11 23:10:11 PDT 2002


Blake said:

Someone could cause their candidate to defeat the CW by bullet voting,
but since Mike accepts this in approval, I suspect he would accept this
here as well.

I reply:

It doesn't make me reject Approval, and so of course it wouldn't
make me reject another method either. Bullet voting can't defeat a CW
in Approval unless people preferring the CW think they need
to vote for the bullet-voters' candidate. In Approval, if they
make that mis-estimate, the count results will show their CW's
support accurately enough so that they'll know that next time
they don't need to vote for anyone else. Approval will quickly
home in on the voter-median position, and stay there.

To misjudge about Favorite having a majority, vs Worst having
a higher vote total than Favorite, is a big mis-estimate.

And mis-estimates like that will tend to elect Middle, rather
than going to extremes. If it looks as if Worst will outpoll Favorite,
then Favorite voters will elect Middle. If it incorrectly looks
as if Favorite has a majority, then Favorite voters won't vote for
Middle, but Worst voters will, and Middle wins.

Blake continues:

I don't know if this method would pass his endless parade
of criteria.

I reply:

For a long time, I've been using these criteria:

FBC, SFC, GSFC, WDSC, & SDSC.

Five is hardly an endless parade.

Recently someone brought up game theory, and so I began determining
which methods always have sincere equilibria. That resulted in
a classification of voting systems. It could of course be written
as a set of criteria, though I didn't word it that way. But you
could reasonably speak of the classification as criteria.

However, all that time without proposing any new criteria other
than those 5 that I've been using, and then, after all these years,
describing my classification of methods--that hardly qualifies as
an endless parade.

What we have here is another of Blake's overdramatic exaggerations
that he likes so much.

But it's understandable for Blake to dislike my criteria and my
classification, because his Ranked-Pairs(margins) fails all of it.
I guess I should repeat that the criteria and the classification
were chosen because they're about things that I & many others
consider important.

Again, though, I'm not saying that it's wrong to propose methods
that look bad by the ways that I evaluate methods, because what's
important to Blake is different from what's important to me.

But it must be obvious to Blake now that the margins methods
aren't proposable. Opponents of pairwise-count methods, including
IRVists, criticize the pairwise-count methods for their opportunities
for offensive strategy, and the resulting strategic mess, as people
attempt strategic countermeasures.

Margins methods can't be defended against that criticism. The
critics don't know about the special guarantees of the wv methods,
and their criticisms don't apply to wv. But they apply to all other
pairwise-count methods, including margins.

It isn't just my criteria, Blake; people don't like the
strategy-riddenness of ordinary pairwise-count methods such as margins.
In recent days, it's been shown that margins supremely justifies
and confirms all the strategy concerns that people have about
pairwise-count methods, and that it certainly can't be defended
against objections based on those concerns.  The concers are justified.
The objections are right.

My criteria & my classification reflect concerns that aren't only
mine.

Blake continues:

As for me, I like the ranked ballot because it allows a limitless number
of candidates to be sorted. It seems to me that with 6 candidates,
there would be a tendency to use all grades, and therefore approve of
exactly 3, which isn't necessarily what you want.

I reply:

It isn't entirely clear if this is referring to Approval, but
if it is, the strategy of voting for the best half of the candidates
would be optimal if one's only information about the candidates is
their order of merit. That isn't considered realistic. Voters
rate the candidates' merit, rather than just ordering them.

Sure, it's nice to be able to express all of one's preferences,
but, depending on how the rankings are counted, they can cause
strategy problems instead of solving them, as in the case of
margins.

Mike Ossipoff


_________________________________________________________________
Send and receive Hotmail on your mobile device: http://mobile.msn.com



More information about the Election-Methods mailing list