[EM] Ranked Preferences, Range

[Abd ul-Rahman Lomax]

At 12:19 AM 10/31/2006, Dave Ketchum wrote:
>This certainly DOES NOT earn a need for special assistance to such a voter.
>
>Whatever information may be available, if the voter does not know enough
>from it to HAVE a personal preference, the obvious response is to not vote
>- leaving to others establishment of a consensus.
>
>DWK

Right. This kind of thinking is behind the proposal that blank votes 
be excluded in determining average Range votes. It does make sense, 
but, if this is done, then it is possible that a winner disapproved 
by a majority of voters can win. Then a Rube Goldberg rule, the 
so-called 25% rule, was proposed, which I think says that, to win, a 
candidate must have been rated by 25% of the voters. I'd prefer more 
direct ways of addressing the problem, possibly even involving a 
runoff. It is clear to me that someone who is rated very highly by a 
significant percentage of voters, but who is largely unrated, 
deserves a serious shot at winning. But it is also clear that such a 
person, even with 25% rating him or her, has not yet earned the 
general approval of the electorate. A runoff of some kind would 
resolve this. This Range winner would no longer be so little known.

What the threshhold should be that triggers a runoff, I'm not yet 
clear about. When there is a Condorcet winner if the Range ballots 
are analyzed as if they were ranked ballots is another situation that 
might call for a runoff. In this case, too, there would be a lack of 
explicit public approval of the Range winner in the election. Good 
chance the Range winner *is* the best, but that has not yet been demonstrated.

With the insufficiently informed voter above, I would simply leave 
the question of how to vote to the voter. There can be and should be 
all kinds of resources to make becoming informed easy. This, indeed, 
is part of the purpose of the FA/DP proposals, which are quite 
independent of the methods being used in public elections. We can 
expect that members of FA/DP political organizations who have named a 
proxy, and who have made a reasonably decent choice in doing this, 
will have good advice on how to vote. "Good" means that it is from 
someone who is more informed, and who is trusted by the voter. They 
may put as much or as little effort into confirming this as they choose....

Requiring that everyone become fully informed would be bad system design....