[EM] Re: Clarifications/commentary on solutions to Gerrymandering.

[Adam Tarr]

Since I transitioned into talking about PR at the end there, let me
make a few more comments on that.

1)  In my opinion, PR methods that require actual ranking of
candidates more or less requires a user-friendly touch-screen
interface that allows the voter to do things like drag an entire party
list into their ballot, and freely move candidates from one slot to
another.  Things like cumulative voting, open list PR, and
proportional approval voting are less dependent on this.

2)  My favorite effort so far at PR Condorcet is this:

http://lists.electorama.com/pipermail/election-methods-electorama.com/2002-November/008855.html

obviously very difficult to completely understand, but in fact less
difficult to vote in than STV (due to allowed equal rankings, of
course).

3)  I had a possible insight about how to expand PAV to range voting,
to get "proportional range voting".

Just take the standard PAV definition:

http://groups.yahoo.com/group/election-methods-list/message/6367

But we replace the key part of the definition here:

"Suppose that ballot B approves exactly k of the members of the coalition
C. Then this ballot contributes support for this coalition in the amount
1 + 1/2 +...+ 1/k ."

With this:

"The ballot B gives ratings for each of the n candidates in a
coalition.  WLOG we order those ratings in descending order as
R_1...R_n.  Then this ballot contributes support for this coalition in
the amount
R_1 + (R_2)/2 +...+ (R_n)/n ."

If the voter votes as if it were an approval ballot, then of course
this is just PAV.  If they vote differently, then it acts differently.
 Would this be a great PR method?  Frankly, I have no idea, but it
seems worth mentioning.