On Tue, Nov 19, 2002 at 05:53:20PM -0600, Steve Langasek wrote: > If your actual preference is A S D, then voting A D S is still an > insincere vote by definition. Note that in this case, the voter *really, > honestly prefers* either of the two non-default options over the default, > but the result he *most* prefers will *lose* the vote *IFF* he votes > and votes sincerely. This is a flaw, because not voting or voting > insincerely gives a better outcome from this voter's POV than voting > sincerely does.
Keep in mind that all voting systems have this flaw. It's just that in Condorcet/CSSD, it is difficult to exploit. So sometimes I say "this system allows strategic voting", when I mean "this system makes strategic voting much more effective than Condorcet/CSSD". I think this is still an important distinction. > You also talk about only allowing insincere votes to cause the default > option to win, but you can't distinguish between sincere and insincere > votes in the system: Right, again what I really mean is, I oppose a system in which insincere votes for the default are more effective than insincere votes in Condorcet/CSSD. The proposals which include early elimination of options beaten by the deafult (possibly with a supermajority requirement taken into account) have this property. Andrew
