[EM] Condorcet-Approval hybrid method

[Dave Ketchum]

Why bother?

How often do you get a cycle in a real election?  You need at least three 
candidates close to a tie, with each beating the next in line (meaning 
each of these candidates are linked in order to another of the leaders).
     While we must provide for cycles because they would CERTAINLY happen 
if we did not make provision, I claim there are limits to how fancy we 
need to get for something so rare.

Yet I do promote permitting and counting equal ranking, though not very 
loud as to the counting.  The doing is desirable because voters do often 
see candidates as equal and it does not cost that much to let them say it.

Why limit the voter to three candidates?  Agreed the voter can, almost 
always, say everything important with 3, but permitting more is not that 
hard to do, and simplifies the rules a bit.

Mixing in Approval complicates strategy considerations even though, as 
noted above, it rarely affects results.

Providing for an "approval cutoff point" complicates the rules a bit more.

Why bother?

DWK

On Fri, 4 Mar 2005 01:54:50 +0100 (CET) Kevin Venzke wrote:

> Russ,
> 
>  --- Russ Paielli <6049awj02 at sneakemail.com> a écrit : 
> 
>>Why not combine Condorcet and Approval into a hybrid election method? 
>>The voter ranks the candidates, but truncates at his approval cutoff 
>>point. If a Condorcet winner exists he wins, otherwise the winner is 
>>determined from the Smith set using Approval rules, where ranked 
>>candidates and considered approved and unranked candidates are 
>>considered unapproved.
>>
>>I would tend to oppose allowing equal rankings (except for implied equal 
>>rankings for the unranked candidates, of course).
>>
> 
> I don't know why you want to disallow equal rankings. Just to make the
> count somewhat easier? If you use Approval, you're not using any defeat
> strength measure, so the WV vs. Margins question doesn't exist.
> 
> 
>>How would this method stack up? What would the voting strategy look 
>>like? And has it been proposed before? If so, when and where? Thanks.
>>
> 
> I think it has been proposed many, many times. I don't know when, but I would
> try a search for "Condorcet-Approval hybrid."
> 
> I like the method, though, especially with a three-slot ballot. Would-be 
> offensive strategists cannot bury a candidate without giving much approval, 
> possibly enough to win, to the minor candidates under which they bury their 
> opponent.
> 
> But if you feel Condorcet already has too much truncation incentive, then
> using Approval as a cycle-breaker will just make things worse.
> 
> Kevin Venzke

-- 
  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.