If you could sell as practicable the type of ballot needed, then I think the Condorcet completed by the utility-based
compressing of ranks method, which I proposed/posted as "Condorcet completed by limited-rank Condorcet!" on Sat.Sep.13. (I wrongly thought that it might pass Participation), would be excellent for your pupose.
If not, then maybe: Condorcet completed by Borda. Eliminate non-Smith set members. Based on the pairwise comparisons
among remaining candidates, choose the candidate with the highest votes-for-minus-votes-against score.
A conservative idea to sell to people who like IRV: Condorcet completed by IRV.
If you have any examples for my "Improved Generalised Bucklin" method (especially to do with Partcipation,
Clone Independence, Burial or Compromise) that might be interesting or testing, then I would be interested.
Chris Benham
---- Election-methods mailing list - see http://electorama.com/em for list info
