Further on the subject of my proposal to modify STV-PR by , after transfering supluses, eliminating the Condorcet loser among those votes and fractions of votes not tied up in quotas: I am afraid I have discovered that this method fails FBC (Favourite Betrayal Criterion).
To demonstrate the problem, I have slightly modified your example by introducing candidate X.
300 votes, 3 seats, Droop quota = 75.
74 A B C D X E 39 B A C D X E 75 C 37 D X E C B A 73 E D X C B A 2 X E D C B A
Compared to your example, all that has changed is that X has stood and taken 2 first preference votes from D and has occupied the runner-up Condorcet loser position. The result, CBD, is unchanged.
The problem is that if those 2 X supporters had reversed their top two preferences, then E would have got a quota and the result would have been CEB. By insincerely down-ranking their favourite, those two voters could have caused their second prefernce to be elected instead of their third.
I'd be surprised if this sort of problem is completely avoidable with ranked-ballot PR. Do you know how "Sequential STV" and "CPO-STV" fare in this respect?
Chris Benham
---- Election-methods mailing list - see http://electorama.com/em for list info
