[EM] Fwd: New Criterion

Michael Ossipoff email9648742 at gmail.com
Sat May 24 14:31:45 PDT 2014

When I replied to this, I forgot to add EM to the "To:" field. So I'm now
posting this reply:

On Tue, May 20, 2014 at 9:15 PM, Forest Simmons <fsimmons at pcc.edu> wrote:

> Mike,
> in one of your articles you mentioned that voters do not tend to
> strategize in polls.

Yes, voting in the polls at Condornet Internet Voting Service (CIVS) seems
to be sincere, because top cycles for 1st place in the output ranking are
rare. I don't know if I've ever noticed one. That at least shows that
successful strategy, in thre form ofchicken-dilemma defection, or of
offensive truncation or burial, isn't happening.

CIVS has a number of political polls, about 2012 presidential candidates,
U.S. political parties, and categories of political parties.

CIVS is at http://www.cs.cornell.edu/andru/civs.html

> So from the polls we can estimate the MAM winner, and then publish the
> Nash equilibrium votes necessary under Benham to achieve that winner.
> There is no guarantee that the voters will take the advice or that the
> popularity of the candidates won't change between the polls and election
> day, but it might be better than nothing.

Yes, if people vote as sincerely in big national pre-election polls as they
do at CIVS, then that poll, followed by the publication of the MAM winner
and the strategy that would elect hir at Nash equilibrium in Benham, would
mean that, even though MAM itself doesn't meet CD, and MAM's SFC and SDSC
are incompatible with CD, it would still be easy to elect an ideal
majoritarian winner (MAM's winner) in a method (Benham) that doesn't have
chicken dilemma.

It shows that, even when giving up SDSC and SFC in order to get CD, there's
still a way to make it easy to get what was given up.

> Even if IRV were the adopted method, you could do the same thing as far as
> possible.  Then also publish the Benham equilibrium position so that the
> voters could see how much less they would have to distort their true
> preferences to achieve the same result under Benham.
And maybe also announce  the voted CW, pointing out that Benham would have
electe hir, to show how Benham would avoid drastic strategy need for voters
who aren't in a mutual majority (MM). Though I think IRV is fine for the
Green scenario, due to its compliance with MMC and CD (plus the bonus of
LNHa and LNHe), Benham is better if you're concerned that you might not be
in a MM, or if you want to avoid vulnerability to replacement of the voting
system by a dis-satisfied majority.

Michael Ossipoff
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20140524/4bba7ef4/attachment-0002.htm>

More information about the Election-Methods mailing list