[EM] Hello (Intro); PR, Condorcet and Approval, variants...

[Venzke Kevin]

Hello.  I've been reading lots of messages on this
list for a week or so, and I decided it would be good
to join it.  I can ask questions even if my own ideas
are not generally interesting.

I've always been interested in organization of
government (esp. comparative), but I've only recently
become interested in single-member election methods. 
Up to that point, I most admired a system like
Chile's: open-list PR for two seats per district.  (I
valued balance.)  But I've become disenchanted with
party lists and proportional representation generally.
 I think parties would wield too much power, and there
would be little reason to expect that passed
legislation would be "improved" somehow with PR.  (In
other words, even if you reduce wasted votes cast by
voters, among legislators the number of wasted votes
should end up about the same.)

I'm now more taken with Approval and Condorcet after
reading about them.  I want the elected candidate to
have the broadest base of support identifiable.  I
have an issue or two with both systems, though. 
Because of voter strategy, I worry that Approval would
do little to "centralize" the elected candidates. 
With Condorcet I worry that many voters would rank
candidates into three groups: the favored candidates
highest, the poorly understood candidates second (we
have such candidates even under plurality; wouldn't
their number increase quite a bit under Condorcet?),
and finally, the disliked candidates.  I worry that,
as a consequence, an "unknown" candidate could
accidentally win, particularly if many voters lazily
place unknown candidates in the same positions.

I'm also interested to read about exotic systems. 
(I've seen some of interest here which defy my
comprehension, as I don't understand concepts such as
"eigenvectors.")  I've devised some bad ones myself. 
One of my favorites puts the incumbent on a party list
all by himself, and every challenger is "de facto" on
a second, "challenger" list.  The voter votes for only
one candidate.  The result is that the incumbent only
wins with an outright majority; if the incumbent fails
to get a majority, the "challenger" candidate with the
most votes is the winner.  It would be an entertaining
system if nothing else.  (You could complicate it
further by permitting any candidate to run instead on
the "incumbent list," with the qualification that such
a candidate would need an outright majority within the
list's votes to take it from the incumbent.  You could
use the option to avoid splitting your party's votes.)

I am putting finishing touches on a system which is an
Approval variant aiming to select the Condorcet Winner
more often.  I hope to write a document on it.  First
I have to make sure I can defend the system's results
philosophically, but I'm very close to that point.  I
also want to write a quick program or two to see if
idiotic results ever occur in random scenarios (that
is, compared with Approval's results).  I hope it will
be interesting.

Stepjak


___________________________________________________________
Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français !
Yahoo! Mail : http://fr.mail.yahoo.com

----
For more information about this list (subscribe, unsubscribe, FAQ, etc), 
please see http://www.eskimo.com/~robla/em