[EM] Voting by selecting a published ordering

[Dave Ketchum]

On Mon, 3 Apr 2006 19:53:12 -0500 Paul Kislanko wrote:

>>Looking much like Condorcet.  From there we know that cycles 
>>can occur, 
>>needing more thought here.
>>
> 
> Only in the case of ties with respect to the pluralities that chose the
> rankings involved. I think with more thoughht we'll find that the
> tiebreaking method here corresponds to a cycle-breaking method in
> Condorcet-based systems.


There was nothing of such in the post that started this.  Again, think of:
      A>B
      B>C
      C>A
In the Condorcet debate we see need to look at the counts FOR ALL THREE 
cycle members (agreed IRV is less interested in finding the actual 
preferred candidate).

> 
> 
>>>If the most popular order is not complete, then refine it 
>>>
>>using the order that is second in popularity, then third in 
>>popularity, etc.
>>
>>>Note that Eppley's suggestion (in its simplest forms) 
>>>
>>requires only a standard plurality style ballot, and each 
>>voter marks only one alternative (a candidate's name or a 
>>code word for somebody else's published ordering).
>>
>>
>>Starts out looking good but, how many lists might there be 
>>with half a 
>>dozen candidates?
>>      What would this quantity do to the voting machine?
>>      How much might this confuse the voter looking for an 
>>acceptable list?
>>
> 
> I think (as I've said before) it is helpful to keep the tallying mechanism
> separate from the collecting mechanism. This is not terribly difficult to
> implement on the collection side, given the pre-loaded published list and
> the voter's opportunity to select one and modify it (in the ideal world that
> modification would instantly become a part of the published list, but
> finding out that many voters made the same modification to the same
> published one is trivial on the back-end tallying side.


If the voter can modify the list we have left the world of the Plurality 
ballot.

> 
> 
> 
>>Sounds like as much trouble as Condorcet and in counting complexity, 
>>though doable with present voting machines - provided they 
>>can tolerable 
>>the number of choices.
>>
> 
> I have always thought of a ranked ballot method as being a "choose from
> among these available orderings" sort of thing. Althoug the combinatorics
> get out of hand very quickly if you assume each ordering is equally likely,
> in practice there's a second-order divide on issues that keeps A>C>B from
> even being considered by A&B partisans OR by C partisans so that permutation
> never appears on either the pre-published lists or "write in" modified
> lists.
> 

The ranked ballot methods consistently allow voters to choose whatever 
orders please each of them.  They do not get out of hand PROVIDED the 
implementors get told to, and do, provide for this.

True that some orderings will get chosen more often than others - but 
matters little.
-- 
  davek at clarityconnect.com    people.clarityconnect.com/webpages3/davek
  Dave Ketchum   108 Halstead Ave, Owego, NY  13827-1708   607-687-5026
            Do to no one what you would not want done to you.
                  If you want peace, work for justice.