At 04:24 PM 7/22/2007, Steve Eppley wrote:
>I think Warren Schudy could have written a stronger negative comment
>about Range Voting. Comparing it to Approval in his paper, he said it
>offers "little or no gain" (see below). That suggests outcomes with
>Range Voting would tend to be at least as good as with Approval.
>Outcomes with Range Voting could be much worse.
Yes. But they can also be much better. Which effect prevails?
> What happens if many
>altruistic voters tend to try to vote sincerely and selfish voters tend
>to use the optimal strategy of extremizing to the limits of the range? Ugh.
But this *assumes* that optimal strategy is voting the extremes!
Further, most of us assume that the vast majority of voters will max
rate at least one candidate and min rate at least one. This is,
already, "extremizing to the limits of the range." Further, it is
clear that voting "altruistically" in Range, unless one really
understands the consequences, is feeding Range distorted information.
Voters should actually vote their own preferences, not what they
think are the preferences of others, unless they *know* these and
prefer to accept those.
What we are discussing is intermediate ratings for others than the
favorite and least favorite. This is where the controversy is. Not
about the max and min candidates.
The "selfish" thing is a red herring. If voters vote as sincerely as
possible, all of them, we actually do maximize overall utility. If
some don't, for some reason, there is loss of overall utility, but it
does not follow that the rational counterstrategy is to also vote in that way.
So far, the simulations are indicating that intermediate ratings
remain optimal even in the presence of a largest minority of voters
(less than half, but only 1 less) voting "Approval style."
Unless we know what those votes are, in which case only one
candidate, typically, must be extreme rated in response, and the
others are moot, so might as well vote sincerely!