Woes with the Elections Reform List

Hugh Tobin htobin at redstone.net
Wed Apr 23 20:37:49 PDT 1997

New Democracy wrote:
> [snip] 

>      My post to the ER list are not being accepted [snip] 

I hope that before you repost this one you will correct your description
and critique of Condorcet.  The following is not correct.

> Condorcet - If there is no majority candidate then the candidates are
> divided into every possible combination of two - this is called pairing. If
> we had four candidate we will have six sets of pairs. Each pair is worked
> like a run-off. The other candidates are dropped and their votes are
> transferred to the two candidates of the pair. The candidate that wins most
> of the pair run-offs is the winner of the election. It is possible that no
> candidate has a majority of the pair wins.

Condorcet does not require "transfer" of any votes. The Condorcet winner
is the one that wins each of his or her pairwise races (not a
"majority").  If nobody wins on that basis, there is a tiebreak.  The
tiebreak methods I have seen do not depend on the number of pairwise
wins, and would not be decisive if they did.  Least total against in
worst loss has been popular on this list; I believe Condorcet intended
least absolute margin(s) of defeat.

> In these three proposed methods the lower preferences of a voter receive
> support while his first preference is still a contender. Savvy voters will
> see this and will refuse to make more than one selection. 
> [snip]
> Donald Eric Davison of New Democracy at http://www.mich.com/~donald

As to Condorcet this is a misconception, as was extensively explained on
this list some months ago.  It would not be rational strategy in
Condorcet to truncate rather than voting for one's true second choice;
indeed in a circular tie this tactic could lead to the defeat of the
first choice in the tiebreak. (Of course, it also could lead to one's
least favorite candidate winning outright.) In limited circumstances a
"saavy" voter might choose to misrepresent his or her second choice,
despite the risk involved, but not to omit a second choice. See prior
exchanges with Steve Eppley, who argued that his preferred Condorcet
tiebreaker is superior due to its "truncation-resistance" feature, but
(unless I missed it) did not produce a plausible scenario, in Condorcet
with the margins of defeat tiebreaker, where a tactically minded voter
would truncate rather than reverse order.

Failure to comment on other parts your posting does not indicate

-- Hugh Tobin

More information about the Election-Methods mailing list