[EM] Definite Approval/Disapproval

[Forest Simmons]

On paper Approval has the best standard criteria compliances of any method.
On top of that it has all around simplicity going for it. Yet nobody likes
it, and the most common complaint from experts and lay citizens alike is
that the definition of "approval" is so vague that it could drive an
indecisive person to distraction: There is no clear guideline for
partitioning the candidates into two distinct categories with a crisp
boundary between them.

In close second is the related complaint about lack of expressive power ...
in particular the inability to distinguish favorite, compromise, and
anti-favorite with three separate levels of ballot support.

"Definite Dis/Approval" (DD/A) addresses head on the basis for these
complaints with judgment style ballot instructions... "mark as definitely
approved (DA) only the candidates that you are absolutely sure that you
want to support" ... "mark as definitely disapproved (DD) only the
candidates that you very strongly feel to be unsuitable for the position."
Otherwise, mark the remaining candidates as either ... "somewhere in the
middle (Mid) between strongly suitable and strongly unsuitable" or "No
basis (NB) for an opinion." A blank (i.e.abstemtion/undecided) is counted
with the NB's.

What if I'm not sure? Then most definitely it would be dishonest to mark DA
or DD. If you cannot decide between NB and Mid, then leave it blank ... the
ultimate expression of "Undecided".

Now suppose that for each candidate k, you have the total counts DA(k),
DD(k), Mid(k), and NB(k), and no other information.

1. How would you use those totals to decide the single winner?

2. How would you construct a finish order if need be?

3. How would you resolve ties?

-Forest
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20220430/acf1546d/attachment.html>