James G-A,
You wrote (Thur.May 26):
Yes, CDTT seems like one interesting way to bring IRV toward
Smith-efficiency. CWO-IRV is another, in my opinion.
I'm strongly opposed to CWO (Candidate Withdrawal Option). If such a
thing were proposed in Australia or (I am sure) the UK, it would be
strongly opposed by people
mindful of the principles of constitutional democracy (and many
others) as an attempt by "politicians"/political parties to usurp the
sovereignty of the voters.
CDTT,IRV looks like my July 26 UMID,IRV proposal.
Based on various precedents, when I refer to the name of some compound
of two methods (or a "social choice function" like CDTT and a method),
separating the two with a comma means to use the second to order the
candidates and the first to pick the highest-ordered that is a winner of
the first method.
Separating them with a double forward-slash on the other hand means to
use the first to eliminate some candidates, and then proceed with the
second with
those eliminated candidates dropped from the ballots to pick the winner
from those remaining.
So in my book (while "UMID" is the same as CDTT) your proposal was
UMID//IRV, not "UMID,IRV".
As I explained in an earlier message, this difference causes UMID//IRV
(or CDTT//IRV) to fail Mono-add-Plump and Mono-append.
Chris Benham
----
Election-methods mailing list - see http://electorama.com/em for list info
