[EM] Re: [ER] Re: Explaining pairwise

Mike Ossipoff dfb at bbs.cruzio.com
Sun Mar 24 01:08:24 PST 1996

Kevin Hornbuckle writes:
> > And here's the complete description:
> > "Each voter ranks the candidates from most preferred to least
> > preferred.  The info provided in the voters' rankings is used to
> > determine the winners of each of the possible head-to-head pairings.
> > The winner is the candidate whose worst pairing defeat is smallest."
> There is a problem with this.
> If this were written as law, there would be a lot of spoiled ballots 
> because not every voter would rank all candidates. There is no logical 
> necessity to rank all the candidates.
> So it should say "may rank the candidates".

This brings up an important difference between Condorcet's method
& other Pairwise-Count methods: In other methods, including
the one known as "Copeland's method", voters who vote a short
ranking can thereby steal the election from the candidate who'd
have otherwise won, the candidate who'd beat each of the others
in separate 1-on-1 elections--with the result that the winner is
the favorite candidate of the voters who voted the short rankings.

That can't happen in Condorcet's method. An important difference
between Condorcet & Copeland. I mention Copeland because some
academic authors advocate it, & electoral reformers may hear
about it at some point. By the standards important to voters
& electoral reformers, Copeland comes out on the wrong end of
a comparison with Condorcet.

Mike Ossipoff

Mike Ossipoff

> .-


More information about the Election-Methods mailing list