[EM] RE : Re: Ranked Preferences, Range

[Kevin Venzke]

Hello,

--- Chris Benham <chrisjbenham at optusnet.com.au> a écrit :
> Range has no real "majority-rule" related guarantees, except a very 
> unimpressive weakened
> version of  Minimal Defense that means that for a majority to block the 
> election of X they
> must both give X a minimum rating and some other candidate a maximum
> rating.

That property is WDSC of course.

--- Abd ul-Rahman Lomax <abd at lomaxdesign.com> a écrit :
> This has been stated so many times by Warren, but obviously it bears 
> repeating. The best vote in Range is a sincere one. Sure, partisans 
> may distort their votes, turning them into black and white. But by so 
> doing they are essentially eliminating their own true preference 
> strengths from the election. They are harming themselves, not the 
> society, except to the extent that the society is harmed when 
> individuals are not satisfied. They are risking their own 
> dissatisfaction, if it turns out that, by voting black and white, 
> they effectively abstained from voting in the top two pair-off.

But while a voter using the whole scale may be able to participate 
in any possible top two, they will be significantly overpowered by
"black and white" voters who chose that top two as one to participate
in.

I would be surprised that Warren has said the above. This doesn't
seem to understand why partisans would be distorting their votes.
Range is a method that allows the voter to weigh their influence by
placing more distance between some candidates and less distance
between others. A partisan, casting a black and white vote, would be
attempting to weigh his influence in order to get the best outcome
from his own perspective. This means that their effort is by design
*not* to harm themselves. While they may be "eliminating their own
true preference strengths," the voter's goal in an election is not
to have one's true preference strength submitted, but to obtain the
best result possible from one's own perspective.

I think it should be noted that you do not even need to have 
information about other voters' preferences in order to devise a
"black and white" strategic vote.


It has seemed to me that Range could be a good method if the voters
not only care about each other's interests (which, alone, would not
be enough to require such a method), but also *do not know* what the
other voters want.

For instance, if you know many people want pie and only pie, although
you'd rather have a hamburger (with pie as a second choice), and
you *do care* that these other people want pie and not a burger,
there is nothing stopping you under any election method from listing
"pie" as your first preference. Nowhere is it written that when ranking,
you can only look out for yourself.

If however you don't know what's best for other people, even though
you do care about that, it would be useful to be able to vote in a
way that seems inefficient for you personally. Maybe for you a burger
is only worth an 8/10 and the worst choice, sushi, still scores 5/10.
It would be useful to be able to cast a vote like that, and let
other voters fill in the gaps with their own information.

Ratings in competitions illustrate this. Individual judges aren't 
trying to make a certain alternative into the winner. They value the
insight of the other judges and want to come to a result jointly. This
way there is no need to be sure to use the entire scoring range 
available, because they are not competing with each other.

Now if these judges were discussing the options prior to 
voting, and shared their sentiments, there would be less need for 
them to use Range. They could use Plurality, and each simply vote for 
the option they know to have general high favor, even if according to 
a majority vote this option would lose.

I also feel that if some portion of these judges were routinely giving
out only high and low scores, in order to increase their influence
in choosing between certain options, then these judges are utterly
preventing the method from working correctly. Their votes serve to
drown out the judges who care about the collective consensus result.
The only way to restore balance (without changing voting methods) is
for everyone to start maximizing the expression of their own 
preferences (plus whatever they can guess about what the consensus
result might be).


So it seems to me that there are a couple of necessary conditions
for Range to be a good method for decision-making:
1. that the vast majority of voters care about reaching a consensus
result, and will not throw this priority out when they realize that
voters can and will undermine this result in their own favor
AND
2. that voters lack some other way of learning what the other voters
want.

#2 seems particularly unlikely to me for almost any group. It makes
sense on a TV show pressed for time. But in a group of friends you
can find out that many people don't want a burger without handing out
ballots. For public elections there are all kinds of media to choose
from to get your information. It is often downright easy to plot
candidates on a 2D spectrum. If you can't guess which candidate would
be the consensus, and don't care to offer an opinion of your own on
the subject, how valuable can your vote be?

Kevin Venzke


	

	
		
___________________________________________________________________________ 
Découvrez une nouvelle façon d'obtenir des réponses à toutes vos questions ! 
Profitez des connaissances, des opinions et des expériences des internautes sur Yahoo! Questions/Réponses 
http://fr.answers.yahoo.com