On May 27, 2008, at 18:52 , Dave Ketchum wrote:
> In summary, yes, that is what the rules could look like. I'm very
> flexible to what kind of set of rules each user would adopt. The
rules
> also could be much simpler than including all the listed
possibilities.
> My intention is just to show various paths that could be used to
make
> the basic random vote method more applicable to the needs.
This puzzles. You need ONE set of rules for all to understand,
with a few details such as number of nominators for a nominee
tailored to getting reasonable quantities of candidates.
Ok, the example below gave one set of rules for one need. Generally I
just identified a list of tricks that can be used in a random vote
based methods to make them usable in various situations.
>> At the moment, my grasp of your suggestion does not allow a firm
>> opinion. Can we flesh out parts of it with greater detail?
>
>
> I presented the proposal as a family of methods that might use
> different rules in different ways. In order to go to greater detail
> (maybe to lesser amount of details too) one could take some example
> situation and example method. We could for example see what kind of
> rules could be used in electing ten people of a city to act as
trusted
> citizens monitoring the criminal interrogations of the police.
>
> There is probably no reason to require any specific skills => normal
> people will do. Maybe all volunteers can be expected to have good
> intentions => no need to control the candidates from this point
of view
> either. Maybe we could require some width of support => let's
say three
> support votes needed. We could allow voters to list e.g. three
> candidates. After collecting the ballots (and counting the
number of
> support votes for each candidate) we would pick random ballots and
> elect the first candidate (who has not been elected yet) with at
least
> three support votes overall from each ballot. If we don't know if
> someone has volunteered we could call him and check (and move to
the
> next candidate or ballot if the answer is negative). If all
citizens
> can be uniquely identified with good enough probability (in unclear
> cases the previous ten elected citizens may interpret the intended
> meaning of the vote) there may be no need for a formal
nomination process.
Good intentions? Desirable, but attempting non-destructive control
could, itself, be destructive.
My assumption here was that these positions were light weight enough
to allow some fellow citizens to make the decisions according to
their best understanding, and that would probably not lead to any
major conflicts of interest. It is for example not very likely that
any of the decision makes would know any of the to be elected
candidates here.
Random ballots? I admit to choking at the thought:
If the voters identify a winner, that should end it.
If the leading candidates are near a tie then it matters
little which wins, but I would go for chance only on a true tie.
>
> This method is quite simple and straight forward and might work well
> enough for this simple task. Just one example among many.
For most elections I am for Condorcet, which permits:
Bullet voting, suitable when a voter does not care beyond
naming a first choice.
Ranking all liked candidates above those liked less.
Ranking all candidates, suitable for ranking hated enemies at
the end.
Random ballot based methods were addressed to offer solutions to
(what I thought to be) the requirements of Mr Gohlke. From this
perspective random ballots can be used to open up the possibility to
elect also some regular citizens in addition to (or instead of) the
party controlled candidates.
Juho
___________________________________________________________
All New Yahoo! Mail Tired of [EMAIL PROTECTED]@! come-ons? Let our SpamGuard protect you. http://uk.docs.yahoo.com/nowyoucan.html
----
Election-Methods mailing list - see http://electorama.com/em for list info
