Exaggerated opinions

[Saari]

In a message dated 98-04-23 08:08:44 EDT, you write:

<<      I have some suggestions that will produce exaggerated opinions.
  One: Allow each voter one to ten votes per preference race. The voter is
 allowed to assign any number of votes to each choice... >>

You have missed the point of my post.  The desire is to *allow* various levels
of opinion to be expressed, but *not* be vulnerable to the problem of
exaggeration.

It is easy to create a system that allows gradations of opinion to be
expressed.  The challenge is to define a method which is invulnerable to the
tendency of some people to exaggerate their degree of feelings.  Most systems
with a fixed maximum vote produce an exaggerated effect.  A "ten" vote for
Apple pie cannot be distinguished from a separate "ten" vote for a $million
reward.

The "pay extra for extra votes" example was meant ONLY as an example to show
that it is POSSIBLE to define a system to control the "exaggeration" problem.
It was meant as an end-all solution, merely a demonstration of possibility.

Here is another method, one that does not involve money.  Require that every
vote beyond the "free" baseline level requires handwritten words of
explanation.  Again, use an exponential scale.  Thus, for instance:
single vote - free
double vote - requires at least one handwritten word of explanation
triple vote - requires at least ten handwritten words of explanation
quadruple vote - requires at least 100 handwritten words of explanation
etc.
Here the "cost" is in time instead of money.

To see how this would work in practice, imagine a true "working democracy"
where you, the voter, is constantly beseiged with new proposals to consider
and vote on.  Although you have the right and theoretical ability to
quadruple-vote on everything in sight, in reality you would not do so because
it would be a waste of your time.  For issues that you have only slight
concern about or knowledge of, you simply will not take the extra effort/cost
to vote at a level beyond your "true" feelings.

We have no language to express "true" feelings, so it is an especially
difficult challenge to find a voting system to allow expression of a full
range of feeling without the attendant problem of exaggeration.  In nearly all
systems it is impossible to determine if a "very strong" vote is a true
feeling or merely grandstanding.  I'm looking for an exception to this rule.

Mike Saari