<html>
David Gamble wrote:<br><br>
<blockquote type=cite class=cite cite><font size=2>27 A1>A2>B
<br>
26 A2>A1>B<br>
7 B>A1>A2<br>
6 B>C<br>
34 C>B<br><br>
Any good ( in my opinion) method should elect A1, A2 and C who all
have a Droop quota of first preference votes.<br><br>
[.snip] <br><br>
C wins the final seat against A2.</font></blockquote><br>
This sort of issue crops up in PR versions of both approval and COndorcet
that ate conducted in a SEQUENTIAL fashion. That is, methods that
pick one candidate, then re-weight, then pick another, et cetera.
In order to avoid these problems, you have to devise a way to compare
entire slates of candidates to one another.<br><br>
Forest chewed through all of this last November, and I believe the last
iteration was this:<br><br>
<a href="http://groups.yahoo.com/group/election-methods-list/message/10308" eudora="autourl">http://groups.yahoo.com/group/election-methods-list/message/10308</a><br><br>
-Adam</html>