Order-reversal deterrence -- an old idea?

[Mike Ositoff]

On Tue, 22 Sep 1998, Hugh R. Tobin wrote:

> Given the credibility problem with order-reversal deterrence strategy by
> the "middle" candidate's supporters in Condorcet, I feel sure the
> following must have been considered:

I've talked about "credibility" at great length several
times, including a few months ago. The middle voters have
great credibility of threat because a victory by one extreme
instead of the other would bother them less than the opposite
extreme. And defenders have more credibility than attackers.

More later.

> 
> Allow a candidate to declare irrevocably in advance (by simultaneous
> secret submission), that the second and lower-place votes by voters who
> choose that candidate first would not count, either (option 1) for any
> purpose; or (option 2) in the tiebreaker in case of a circular tie.
> Under option 2 these votes still would count to determine whether there
> is a Condorcet winner and who is in the Smith set.  In either case the
> candidate's election of this option would be stated in voters' pamphlets
> and on the ballot, perhaps with the candidate's explanation for
> depriving his supporters of some of their voting power.
> 
> Does this concept in either form have a name, and has it been beaten to
> death?  I do not  want to burden readers with half-baked analysis of a
> discredited notion.  (I realize that if all candidates chose this option
> we would have plurality, or a plurality tiebreaker, but I am not sure
> how likely that is.)
> 
> -- Hugh Tobin
> 
>