[EM] Mono-add-top method suggestion

[Chris Benham]

I think if there was an informed honest debate on the relative merits 
(for public political elections) of Hare versus one of the best 
Condorcet methods I think Hare's compliance with Mono-add-Top would be a 
big point for it.

MinMax Margins is a Condorcet method that meets Mono-add-Top, but pays 
too heavy a price for that to qualify as a good Condorcet method.   Hare 
meets Mono-add-Top and so does Approval.

This is my (not too fanciful thought experiment) idea for a method that 
meets Mono-add-Top and is more Condorcet efficient than Hare (but more 
complicated) using ranked ballots with explicit approval cutoffs:

*Elect whichever of the Hare winner and the MinMax Margins winner is 
more approved.*

Unfortunately as well as failing Condorcet I am sure it also fails 
Double Defeat (meaning a candidate pairwise beaten by a more approved 
candidate might win), but a candidate that is both the CW and the AW 
always wins.

There is no zero-info voter incentive to truncate or to do anything 
(aside from the approval cutoff) other than give a full sincere 
ranking.  If MMM tries to flout the Plurality criterion I can't see any 
way that the Hare winner won't be more approved.


Chris