Party List P.S.

Mike Ositoff ntk at netcom.com
Wed Aug 5 18:26:45 PDT 1998


nn> 
> In a message dated 98-07-27 04:04:21 EDT, you write:
> 
> >Candidates can withdraw any time a result is announced, and so
> >the process of withdrawals & new count is repeated till no one
> >else wants to withdraw.
> 
> Of course, this method has at least three obvious flaws.  One, for an extended
> period of time after the election it is indeterminate who the winner is.
> 
Nonsense. We typically have to wait till the next day anyway. How
long will it take a count computer do the several counts. The
candidates could be required to remain in town if they want to
withdraw. In Cambridge, the wait has often been several days.


> Two, this method won't work when candidates are things/proposals, not people.

Wrong. The proponent of a proposal could withdraw it.


> 
> Three, a losing candidate who either withdraws or does not withdraw is now in
> a position to directly influence the outcome.  Issues of corruption and

Remember that if a candidate withdraws, his votes will go, not where
_he_ says, but where his voters have indicated that they want
them sent. If a winning candidate withdraws, then I'd agree with
your anger. I feel your pain. If a losing candidate withdraws,
then what's your problem?


> bribery aside, clearly the outcome is now a consequence not merely of the
> wishes of the voters but is also being manipulated by the candidates

Manipulating by not being a spoiler? By a loser not hoarding votes
that his voters would rather send elsewhere? :-)

> themselves.  A flaky or unpredictable result indicates a flaky election
> system.  In a good system the outcome is completely or primarily a reflection
> of the feelings of the voters WITHOUT the opportunity of direct manipulation
> of the outcome by certain individual participants.

You wouldn't be setting yourself up as the person with authority
to define what's "good", would you? Your error is that you're 
exaggerating the power of a withdrawing candidate. What purpose
remaining in the count, holding votes that his voters want to
help someone else with?

Any election result is unpredictable (and therefore "flaky"?).
What can be predicted, however, is that Condorcet winners will
have pretty much a sure win. Without any strategy or predictive
knowledge on the part of the voters.

> 
> Proponents of ranked voting (first, second, third, etc.) offer us a variety of
> "scoring systems" but seem unable to find a single scoring system that is not
> obviously flawed.  (And Arrow has essentially proven that they will never find

The question is whether you have a proposal that isn't flawed.
After years of discussion, we've minimized the flaws, and found
methods that meet the standards most important to people, based
on conversations & articles, etc.

> such an ideal scoring system based on ranked voting.)  In my book this
> indicates that a different approach is required.  
> One such approach involves discarding historical dogma such as "majority
> rule", "ranked voting" or the simplistic "one man, one vote" and instead asks:
> What is a method of voting which allows for full freedom of expression?  What
> is a method of scoring such votes that does not create instant paradoxes or
> contradictions?  And can such a voting/scoring system be devised that does not
> create incentives to lie or exaggerate?

> 
> Some people on this list believe that the above goals are impossible to
> achieve, and prefer instead to stick blindly to the old methods even though
> they do not produce satisfactory results in all cases.  Perhaps a single ideal
> method is impossible to achieve, but personally I think it can be done.
> 

"Some people on this list" have been involved with the subject for
longer than you have, apparently.

Do you realize how little sense you make when you attempt to
claim that what we do isn't valid, because _maybe_ someone _could_
find something better? Unless you yourself can either propose
something better, or quote someone who has, then what you're
saying has no validity.

It's been pointed out to you that the flexible point systems
that you used to propose, like -100 to 100 are not strategyk
free, and don't encourage sincere voting, and are strategically
the same as Approval (not a bad method, but we have better ones).

Then you talked about charging money from voters to ensure
sincerity. No, the votes that voters pay for would still
be used strategically.

I know of only 1 system that encourages sincerity: 
The Clarke tax, using real money. Charging real money froim
voters is quite out of the question. Allowing for differences
in income, savings, expenses, family, debts, gambling problems,
substance habits, etc. etc., to try to make fair equal incentive,
is completely unfeasible.

You've been ousted from each point from which you attacked
rank-balloting, and so now you resort to saying that maybe
some hypothetical unknown, unproposed method might be
better.

Mike Ossipoff


> Mike Saari
> 
> 



More information about the Election-Methods mailing list