Dear Juho!

> How about the weighted Borda that I described earlier? Square root
> weights would give 0, 1 and 1.4 points instead of 0, 1 and 2.

I'm totally sorry -- I just have forgotten that you proposed this. It 
will probably work, too, but I'm not sure whether those scores make it 
easier or more difficult...
 
> In line with what Forest wrote I'm not advocating these methods,
> except as challenges on this list. The most obvious problem is
> clones, e.g. one candidate from the Democrats and two from the
> Republicans. From this point of view one could require that votes 55
> A>>C>B 45 B>C>>A and 55 A>>C>B 45 C>B>>A will not elect C.

Right. Neither am I advocating them. It was just an observation that 
Borda deals better in this particular situation. We should improve upon 
that!

Yours, Jobst

Attachment: pgpJcAZlw3AH1.pgp
Description: PGP signature

----
Election-Methods mailing list - see http://electorama.com/em for list info

Reply via email to