[EM] MTA vs. MCA (was "An ABE solution")

[Dave Ketchum]

On Nov 21, 2011, at 8:53 PM, fsimmons at pcc.edu wrote:

> Jameson,
>
> like you I think ratings are simpler and easier for the voter.  But  
> it doesn't matter because ordinal ballots
> can be transformed clone free and monotonically into ratings by the  
> technique I gave in my message
> entitled
> "Borda Done Right" a few months ago.

I see July 27 as when Forest described converting rankings to  
ratings.  Actually not much worth doing unless the ranking coder  
expected this conversion (the fact that the numbers convert does not  
make the associations among them fit in their new environment).

Converting the other way is easier - keep the numbers and call them  
rankings.  This keeps all that is meaningful for we look at rankings  
for A & B and care only whether we have A>B, A=B, or A<B.

Forest described the modest effort for converting rankings to  
ratings.  I claim ratings to rankings HAS to be simpler since you  
simply change the labels without doing anything to the numbers  
themselves.

Using them also fits:
      Ranking as to two values - just see to it that all A &B pairs  
properly relate as above.
      Rating demands also deciding on the magnitude of the differences  
the voters, collectively, assign.
>
> Forest
>
> ----- Original Message -----
> From: Jameson Quinn
>> Why ranked and not graded ballots?
>