[EM] Runoff terminology --> Seeded Condorcet

[Forest Simmons]

On Wed, 14 Mar 2001, LAYTON Craig wrote:

> Forest wrote:
> 
> >Would "instant voteoff" be too suggestive of elimination (even though it
> >naturally includes full Condorcet through the round robin playoff analogy
> >mentioned above)?
> 
> That sounds a bit Survivor.
> 
> >By the way, I'm no sports fan, but I understand that the playoff
> >tournaments of the major leagues are "seeded" in such a way that the
> >highest ranking teams do not get pitted against one another in their first
> >games; otherwise some of the best teams could be eliminated too early, and
> >the final games would not be as interesting on average.
> 
> It's a sensible rationale.  This is off the topic, so skip if you're not
> interested - a football competition here uses a very refined finals system.
> The top 8 teams (in terms of points) make it to the playoff.  1 vs 4, 2 vs
> 3, 5 vs 7, 6 vs 8.  The losers of 5 vs 7 and 6 vs 8 are eliminated.  The
> winners play the losers of 1 vs 4 (contest A) and 2 vs 3 (contest B).  The
> losers of the match between the 5 vs 7 & 6 vs 8 winners and the contest A &
> B losers are eliminated.  The A & B winners play one another.  The winner of
> this contest makes it to the final, and the loser plays the winner of the
> final two uneliminated teams.  The winner of this match makes it to the
> final.
> 
> That's probably not very comprehensible, but I just thought that their
> system of ensuring that the top two teams make it to the final is pretty
> neat.  I'm often amazed that sports organisations have a better grasp on
> fair resolution of contests than political parties and other such groups
> that should know better.
> 
> <snip explaination of seeded Condorcet>
> 
> >In other words, this version of the seeding idea is for those that
> >appreciate Approval, but prefer Condorcet.
> 
> Yep, I can see the advantages of this.  It would also be very easy to count,
> in comparison with Condorcet completion methods, making it proposable for
> small committees or anyone without access to appropriate tallying
> technology.
> 
> I wonder if there could be some kind of safeguard from electing a bad
> Condorcet candidate.  Perhaps the bottom candidate (least approved) should
> be eliminated.  There is a good argument that she shouldn't win, even if she
> is the Condorcet winner.
> 
> 

I think that all candidates with less than 50% approval should be
eliminated, except perhaps when that would eliminate all of them.

Forest