The EM vote on single-winner methods

[Steve Eppley]

I received the following private message from an EM subscriber.  He 
asks some questions which others may also have, so I'm posting my 
answers in EM.

>I have been a subscriber to this interesting list for some months,
>and have tried to follow most of the postings, but I am not clear
>about this current vote.  Is it only among members of the
>"Committee", or are all subscribers eligible?  

All subscribers are eligible.  The committee is composed of anyone
who wants to be on the committee.

>If they are, then perhaps you could post the following information:
>Who is entitled to vote; what is the form of ballot; 

I recommend you use either a ranked ballot (where you sort the 
choices from most preferred to least preferred) or a rated ballot 
(where you rate each choice on a scale such as -100 to +100).  If you 
use a rated ballot your rankings can be derived from it, but both 
ballots cast so far have been ranked ballots.  (My ranked ballot also 
included the "None of the Below" choice to delimit where my approval 
transitions to disapproval.)

>who are the declared candidates (where the definitions can be
>found); 

Below is a list of method names.  Their definitions are scattered 
throughout the EM messages, so it's going to take some work to 
find and copy them into one message.  I think we'll have to postpone 
the poll closing deadline while this is done.

You are also free to "write-in" other methods not included on this 
list.  (Bruce Anderson and Demorep1 have posted definitions of other 
methods.  Bruce's favorite, Regular Champion, is included below.)

  MPV (aka Instant Runoff, aka Hare)
  Condorcet
  Smith Condorcet
  Random solution of pairwise circular tie
  Double Complement (a modification of Runoff)
  Approval
  Regular Champion (aka Copeland with a Plurality tie-breaker)
  Condorcet using rated, not ranked, ballots
  Smith Condorcet using rated, not ranked, ballots

>and how a winner will be selected (or will the votes be tabulated
>using alternative methods, and if so may one vote different rankings
>for tabulation under different methods)?

We can tally the ballots in more than one way, and if results differ
we will note this in the report.  If you want to submit different
ballots for different tally methods, go ahead.

>Though I hesitate to quibble with the decisions of those leading a
>discussion to which I have not contributed, in my opinion, if the
>vote were intended to represent the views of the subscribers
>generally, then setting a deadline with only a few days notice,
>over the Independence Day holiday (and without anyone calling
>attention to the deadline in the subject line of a posting),
>probably would not be the best approach. 

Good point.  Let's extend the deadline further.  I would suggest 
sometime on or about 7-10, except we now know that at least one voter 
doesn't have the definitions of the methods.  (I'm also concerned 
that just reading the definitions won't be enough to let you make 
an informed vote--we've been discussing pros and cons for months.
I'd be willing to forward my entire EM email folder upon request, 
as one huge message or attachment, upon request.)

In a sense, though, the deadline wasn't set on short notice.  When
the poll was opened many weeks ago, it was announced then that it
would be considered closed if ever 7 days elapsed without any
ballots received.  No one objected to that, and it's been a long time
since any ballots have been received.  The purpose of setting a
deadline isn't to exclude votes; it's to encourage people to cast
votes. 

---Steve     (Steve Eppley    seppley at alumni.caltech.edu)