[EM] Re: Condorcet package

[Paul Kislanko]

 See below.

> -----Original Message-----
> From: election-methods-electorama.com-bounces at electorama.com 
> [mailto:election-methods-electorama.com-bounces at electorama.com
> ] On Behalf Of Ted Stern
> Sent: Thursday, February 24, 2005 1:47 PM
> To: election-methods at electorama.com
> Subject: [EM] Re: Condorcet package
> 
> On 23 Feb 2005 at 11:42 PST, Dave Ketchum wrote:
> >> On 23 Feb 2005 at 01:00 PST, Dave Ketchum wrote:
> >>>Counting votes:
> >>>     (wv) seems the appropriate choice.  If two voters rank a pair
> >>> of candidates (a=b) as equal, then (a>b) and (b>a) should each get
> >>> one count.
> 
> Hi Dave,
> 
> Another point occurred to me about counting equal ranking as 
> a vote each way.
> 
> Besides introducing yet another difference between margins 
> and winning votes
> (in favor of margins!), you also have the question of how you 
> count unranked
> candidates.
> 
> In your proposed method, you place unranked candidates at lowest rank.
> 
> Say there are 9 major candidates, ranked at 1-9 for by 99% of 
> the voters, and
> 100 fringe candidates, unranked except by 1% of the voters.
> 
> Your tally method would require extra votes in 9900 pairwise 
> matrix locations,
> each of the 100 fringe candidates getting one vote against 
> each of the other
> 99.
> 
> Do we really need to confuse the electorate with huge amounts 
> of extra work
> and detail in the pairwise matrix?

The electorate never sees the pairwise matrix. The voter only submits a
ranked ballot.

I have tried to express (ineloquently) before that we need to axiomatize
EM's by keeping vote-collection methods distinct from vote-counting methods.
Even in pure Condorcet, a voter is not presented with a pairwise-matrix,
since that's a tabulation tool.