Report from the Single-Winner Committee (draft) (fwd)

[Mike Ossipoff]

Mike Ossipoff writes:

[When I initially sent this, using the "reply" option, it got
sent only to Steve, not to EM. So I'm forwarding this, and one
other letter that happend to, to EM. In the future I'll
use the "group-reply" option when I want my replies to go to
the whole EM list (as is usually the case), and it might
be necessary for you to do the same]

> From dfb Sat Aug 17 01:36:40 1996
> Subject: Re: Report from the Single-Winner Committee (draft)
> To: Steve Eppley <seppley at alumni.caltech.edu>
> Date: Sat, 17 Aug 96 1:36:38 PDT
> From: Mike Ossipoff <dfb at bbs.cruzio.com>
> Cc: dfb at bbs.cruzio.com
> In-Reply-To: <199608161940.MAA04327 at alumni.caltech.edu>; from "Steve Eppley" at Aug 16, 96 12:45 pm
> X-Mailer: ELM [version 2.3 PL0]
> Message-ID:  <9608170136.aa14293 at cruzio.com>
> 
> Alright. I move that we use that report as-is, with the necessary
> method definitions added. It's a complete description of the
> election.
> 
> But I also move that we only define a few of the most advocated
> methods: Condorcet, Smith//Condorcet, Instant-Runoff (including
> its Instant-Runoff-1 version), & Approval. I don't believe there's
> any other method with more than 1 advocate on these lists.
> 
> A suggestion: When someone includes NOTB in their ranking, aren't
> they voting that they prefer not having any winner, over the
> alternatives they ranked below it or didn't rank? So couldn't
> it be said that NOTB is used by voters to mean "no winner"? Because
> if I rank NOTB over Regular Champion & Instant Runoff, I'm saying
> that I'd rather have no winner than have one of those as winner,
> and isn't that what it means to say that I disapprove of those
> 2 alternatives enough to vote that they're unacceptable? So a
> compact way to define the alternative called NOTB would be to
> say that, in the election voting & count, NOTB represents the
> alternative of not having any winner of the election. That gives
> a strong meaning to the disapproval definition.
> 
> The reason why I'd include Instant-Runoff-1, but not other
> not-widely-advocated method modifications is that Instant-Runoff's
> usual definition doesn't say what to do about equal ranking, so
> that neither of those 2 versions is
> explicitly favored by traditional rules & definitions (though
> the Instant-Runoff-1 version has been criticized by some 
> electoral reform organization leaders).
> 
> Definitions:
> 
> Of the several methods that I suggest defining, the only
> ones not defined in Steve's draft are Approval &
> Instant-Runoff, & Instant-Runoff-1. Here are some suggested
> definitions for those methods:
> 
> Approval: A non-ranked method the same as Plurality except that
> a voter is permitted to vote for any number of alternatives
> instead of being limited to 1 alternative. Each alternative
> that someone votes for receives 1 whole vote from that voter.
> As in 1-vote Plurality, the alternative with the most votes
> wins.
> 
> Instant-Runoff [I'd use this brief definition that's
> different from the wordier one usually used by Instant-Runoff
> advocates]:
> 
> Repeatedly, eliminate from the rankings the alternative occupying
> highest position in fewest rankings. If at any time an alternative
> occupies highest position in the most rankings, that alternative wins.
> 
> Instant-Runoff with equal rankings permitted:
> 
> [This version is the only one that's awkward to define this
> way, but even this version is likely more briefly defined
> this way]
> 
> An alternative possesses a "point" for each ranking that it
> alone occupies highest position in, and 1/N points for each
> ranking where it's one of N alternatives sharing highest position.
> 
> Repeatedly, eliminate the alternative with fewest points. If
> at any time an alternative has a point total numerically equal
> to at least half the number of voters, then it wins.
> 
> 
> Instant-Runoff-1:
> 
> Repeatedly, eliminate from the rankings the alternative occupying
> or sharing highest position in fewest rankings. If at any time
> 1 or more alternatives occupy or share highest position in a majority
> of the rankings, then the alternative which at that time occupies or
> shares highest position in the most rankings wins.
> 
> ***
> 
> I re-emphasize that, to avoid confusing or discouraging ER
> members, we should (I claim) define only the methods that
> I listed, the ones advocated by significant numbers of people
> on these lists. That, of course, is part of my motion, which
> I stated at the beginning of this message.
> 
> By the way, I no longer question the desirability of including
> the pairwise preference matrix in the posting to ER. It makes
> complete election information available to ER, and is necessary
> to show the count for the pairwise methods.
> 
> ***
> 
> Mike
> 
> 
> 
>  
> 
> 
> 
> 
> -- 
> .-
> 


--