[EM] Kristofer, April 3, '12, Approval vs Condorcet

[Juho Laatu]

On 17.5.2012, at 4.39, Dave Ketchum wrote:

> Oops - took so long stripping Mike O's zillion words that I forgot to respond.
> 
> On May 16, 2012, at 10:05 AM, Dave Ketchum wrote:
>> On May 15, 2012, at 2:55 PM, Juho Laatu wrote:
>>> On 15.5.2012, at 11.11, Michael Ossipoff wrote:
>>>> 
>>>> Juho and Kristofer:
>>>> 
>>>> Just a few preliminary words before I continue my reply to Kristofer that I
>>>> interrupted a few hours ago:
>>>> 
>>>> We all agree that Approval would be much easier to propose and enact than
>>>> would Condorcet. Therefore, we must also agree that, given the same level of
>>>> effort, the expected time needed to enact Approval is quite a bit less than
>>>> the expected time needed to enact Condorcet.
>>>> 
>>>> Now, given that, there are two reasons why you could say that we should try
>>>> for Condorcet instead of Approval:
>>> 
>>> I'm still not quite certain what elections this proposal refers to. If it refers to use of different single-winner methods in single-winner districts of a multi-winner election to elect members to some representative body, then I'm not ready to recommend elther of those changes before I understand what the goals are.
>>>> 
>>>> On another subject:
>>> 
>>>> But if you want to suggest that others shouldn't propose Approval, then you
>>>> need to give a good reason.
>>> 
>>> Approval may be an easy and acceptaböe first step. My opinion is that you should plan also next steps, in case someone wants to cancel the reform, drive it further, or if the strategic vulnerabilities of Approval pop up in some election (like the Condorcet criterion problem popped up in Burlington, althogh that was maybe not even noticed by all).
> 
> I know of no useful reason for rejecting Approval's replacement of Plurality - it's permission to approve of more than one as equally desired while rejecting less than Plurality and the increasing in complexity is trivial.

Use of Approval could be a good move although Approval itself is not problem free. And the political system might get some hickups. But there may well be more benefits than problems. And this is a practical place to start the reform.

> 
> But stepping from Plurality or Approval to Condorcet is also doable and worthy.

Condorcet methods are good single-winner methods. Using them to elect representative bodies from single-member districs is an integersting non-classical approach. Approval and Condorcet are in the same category in the sense that they favour centrists and they move away from the alternating two-party power. Use of Condorcet avoids some of the problems of Approval.

Juho


> 
> This is a bigger change because it allows voting for unequally desired candidates with unequal ranks, thus directing those preferred to be given preference in winning.  This preference allows voters to include less-liked candidates while directing counters to consider better-liked candidates as preferred.
> 
> Note:  Burlington, as displaying IRV's weakness, is not truly Condorcet, for it has restrictions on ranking and its counters must make decisions without considering all the content of the votes.
> 
> DWK
>> 
>>> Juho
>>>> Now, to resume my Kristofer reply:
>>>> Mike Ossipoff
> 
> 
>