[EM] National Popular Vote & Condorcet

[Paul Kislanko]

Yes, as usual I wasn't very clear. The way that SHOULD have been worded is:

One can use plurality to count any kind of ranked ballots, but only
plurality to count plurality ballots.

One can use approval to count any kind of ranked ballots that allow equal
rankings, but only approval to count approval ballots. 

My point being that if we used the most general kind of ballot, the
COLLECTION process is independent of the COUNTING method. Which, in a way
avoids the whole strategy discussion. If I'm asked to fill out a ranked
ballot without knowing how it will be counted, I can't "strategically" vote
"insincerely."

It's more a technique for formalizing analysis, not a recommendation.  

-----Original Message-----
From: Jonathan Lundell [mailto:jlundell at pobox.com] 
Sent: Wednesday, July 01, 2009 9:23 AM
To: Paul Kislanko
Cc: 'Dave Ketchum'; 'EM'
Subject: Re: [EM] National Popular Vote & Condorcet

On Jun 30, 2009, at 10:44 PM, Paul Kislanko wrote:

> One can infer a plurality ballot from any kind of ranked ballot, but  
> not the
> other way around.
>
> One can infer an approval ballot from any kind of ranked ballot that  
> allows
> equal ranks, but not the other way around.

Except for strategic considerations. There are surely many cases in  
which my plurality vote is not the same as my first-ranked vote under,  
say, IRV or a Condorcet method.