Reply to Hugh Tobin on Condorcet

[Hugh Tobin]

New Democracy wrote:
> 
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - April 24 1997
> Dear Hugh Tobin and List members,
> 
> Hugh Tobin wrote: >Condorcet does not require "transfer" of any votes.
> 
> Don writes: In order for a candidate to win a pairwise race that candidate
> must receive votes from the candidates that are not one of the pair and not
> in the pairwise race. 

No, the candidates receive votes from the voters, not from other
candidates.  As has been pointed out on this list, voters could be
allowed to vote in each pairwise race directly, rather than having to
rank candidates in order of preference.  In Condorcet the ranking, in
any event, is only a shorthand method of voting in those pairwise
races.  It is not a matter of transferring votes, just a rule that if A
is ranked higher than B on a ballot, then A gets one vote against B.

> Don writes:
> What does "least absolute margin of defeat" mean? How do you define a
> margin? Is a margin the difference between the two candidates in a pairing?
> Do you then total up all the margins of defeats of each candidate and the
> winner is the one with the smallest total of defeats? So many questions -
> so few answers.
> 
You are right that absolute margin is the number of votes by which A
beats B, not the percentage difference.  If A beats B by 1, B beats C by
2, and C beats A by 3 (circular tie), then Condorcet, as I read the
Essai, says B wins.  If there are several other candidates, all of whom
lose to each of A, B, and C, then in Smith//Condorcet, their margins of
loss are irrelevant, because the winner must come from the "Smith set". 
I have seen no reasonable argument against adding the Smith criterion to
Condorcet, so I think that is the version worthy of your attention.
If candidates within the Smith set have multiple losses (probably rare)
then I do think least aggregate margin(s) of defeat is a reasonable
tiebreaker, but this can be debated.  The main point is that using any
of a number of possible tiebreakers, a system that elects the Condorcet
winner when there is one is more democratic than one that fails to do
so.

> Regards,
> Don
> 
> Donald Eric Davison of New Democracy at http://www.mich.com/~donald
> 
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Regards,

Hugh Tobin