[EM] Approval & CR are 2nd best; Approval STV

[Forest Simmons]

On Wed, 31 Mar 2004, [iso-8859-1] Kevin Venzke wrote:

>
> A method I would rank between Condorcet and Approval is AER, or "Approval STV."
> It's IRV, but the elimination order is based on approval.  Because the approval
> counts don't change, and the elimination order is thus fixed from the beginning,
> it's monotonic.  It's intuitive, and gives similar results to WV methods.


What if you recalculated the approval order after each elimination by
making a minimal adjustment to the approval cutoff on each ballot so that
no ballot that still had any candidates ranked would either approve all or
disapprove all candidates?

Would the method still be monotonic?

Forest