[EM] Best Ever Single Winner Method?

[Chris Benham]

Forest,

Why, in this 3-slot method, do you want the winner to be either the Top 
Ratings winner or the Approval winner?

What is wrong with the Condorcet winner?

Given that  you have chosen that the winner be one of  two candidates, 
how do you justify not simply picking
the one that pairwise beats the other?

I am allergic to needlessly giving losing candidates any more power to 
influence the result than any other voter.

In my book all acceptable 3-slot methods should meet either FBC or 
(3-slot) Condorcet. (We know that meeting both
isn't possible).  Your suggested method meets neither.

49: A
48: B
03: C>B

In this example A is the Top-Ratings winner and B is the Approval winner.

In your "asset voting" runoff A and B both vote for themselves, but say 
C is a supporter of plurality voting and so thinks that
A is the legitimate winner and besides, A has promised him a really good 
job if he wins office (which A "might"not have
done if C hadn't been a candidate in this election) and so C votes for A 
and A wins.

B is of course the Condorcet winner and in this example all acceptable 
methods will elect B.  (Otherwise what was the point
of replacing plurality voting?).

Acceptable 3-slot methods?  For  FBC compliance I like IBIFA:

> *If any candidate X is rated Top on more ballots than any non-X 
> candidate is approved on ballots that don't top-rate X, then
>   the X with the highest Top-ratings score wins.
>
>  Otherwise the most approved candidate wins.*

Also good for FBC compliance is "Improved Condorcet", Top-Ratings (aka ICT).

For compliance with 3-slot Condorcet I suggest Smith//Approval or 
Smith//Top  or  (at a pinch, for greater simplicity)
Condorcet//Approval.

Chris Benham


On 16/05/2019 5:51 am, Forest Simmons wrote:
> This is a three slot method: voters can mark candidates "preferred," 
> "acceptable," or blank (no mark).
>
> There is a simple, low cost, (but not instant) runoff between the the 
> candidate with the greatest number of preferred ratings and the 
> candidate with the greatest approval (preferred plus acceptable ratings).
>
> The runoff is by candidate proxy, i.e. by asset voting.  A candidate's 
> asset total is the number of ballots (or fractions thereof) on which 
> she is marked "preferred."
>
> So if you mark three candidates as preferred, each one of them gets a 
> third of an asset from your ballot.
>
> In general three slot methods get off to a great start, but get bogged 
> down in deciding what to do when it is not clear whether the approval 
> winner or the plurality winner should be the method winner.
>
> And every time we propose asset voting we get bogged down in rules to 
> constrain the candidates to some reasonable way of using their assets 
> to settle on a winner.
>
> This hybrid method avoids both problems in one fell swoop!
>
> ----
> Election-Methods mailing list - see https://electorama.com/em for list info


---
This email has been checked for viruses by AVG.
https://www.avg.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20190518/da6a8579/attachment.html>