[EM] In defense of the Electoral College (was Re: Making a Bad Thing Worse)

[Raph Frank]

On Fri, Nov 7, 2008 at 8:58 AM, Kristofer Munsterhjelm <
km-elmet at broadpark.no> wrote:
> If some
> use Condorcet, those have an advantage, and if some want to use cardinal
> weighted pairwise, they can do so.

You can also create a pairwise matrix from a range/score voting election.

Basically, each candidate's total score sum is divided by the max score and
all entries for that candidate in the table are set equal to that value.

For example,

Condorcet voting
A ballot of A>B>C is converted to

   A  B  C
A  X  1  1
B  0  X  1
C  0  0  X

Range/Score voting:
A ballot of A(99) B(49.5) C(0) is converted to

   A   B   C
A  X   1   1
B  0.5 X   0.5
C  0   0   X

e.g. the ballot is renormalised so that max value = 1 vote, A(1) B(0.5) C(0)

Approval voting
A ballot of A + B is converted to

   A   B   C
A  X   1   1
B  1   X   1
C  0   0   X

Plurality voting
A ballot of A is converted to

   A   B   C
A  X   1   1
B  0   X   0
C  0   0   X

Since each method is summable, this means that all the matrices can all just
be added together.

If all the states use one type of ballot, then the result is equivalent to
just using that type of election nationwide.

Methods like IRV would be harder to fit into the system.  It would require
counting all the ballots (probably at the State level) and then generating a
psuedo-matrix.  Alternatively, maybe the matrices could be filled in as best
as possible based on the announced results.  It would give a mix of
condorcet and plurality voting.

First choices would effectively work like plurality and then the extra info
from later rounds would help fill in the remainder.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20081107/4117702b/attachment-0003.htm>