On Dec 28, 2007, at 18:51 , Jonathan Lundell wrote:
With the pizza example surfacing again (and again and again...), it
struck me that what bothers me about this example is that, in real
life, deciding on a pizza is one of the few places where just about
everybody would use informal consensus.
(For an introduction to formal consensus: http://www.consensus.net/)
I've come over the years to the regretful conclusion that formal
consensus is not workable for most organizations, at least not unless
some fairly stringent preconditions are met (some are described by
Butler at the site above; they include fairly explicit agreement on
group goals, along with a lot of time an patience).
But for pizza decisions, consensus rules. In particular, we try to
accommodate singleton minorities with strong negative preferences
("concerns" in consensus-speak): anchovy-haters, the allergy-ridden.
It doesn't matter that sausage and pepperoni is the Condorcet or
majority winner if there's a vegetarian in the group; we'll find some
consensus choice (fresh tomatoes and pesto, anyone?), given a little
time, good will, and discussion.
(That points up another problem with the pizza example: nobody ever
seems to go to a pizza parlor with individual portions, or
heterogeneous pizzas. But that's another problem.)
I wonder if there isn't a better simple example out there in which
voting is a better strategy than the alternatives.
Yes, sometimes behaviour in the pizza examples and real life do not
match. Example environments with better match between real life and
the discussed concepts would be helpful.
The pizza examples have some properties like possibility of someone
being a vegetarian or allergic, and the possibility to stretch the
examples over meals of several days.
To generalize this, when evaliuation different election methods I
often miss clear description of 1) the purpose/intent and 2) the
environment. If nothing is stated my basic assumption is that people
refer to typical country level political elections.
"Purpose" refers to the sought after behaviour of the election
method. The pizza examples often pay special attention to voters that
get an unacceptably bad result (e.g. the vegetarian voter). Methods
that give different winning probability to the candidates
proportional to their support serve a totally different intent than
methods that aim at electing a compromise winner. Different purposes
favour different methods.
"Environment" is mostly relevant for evaluating the strategic risks.
The requirements are different for large scale public elections with
independent voter decision making, and for opinion poll like
elections where voters have no direct interest to strongly push their
own favourite alternative, and for elections of few voters with well
known opinions and strong fighting tradition.
"Environment" is also relevant when discussing the level of support
some method is expected to have. Proposals may look quite different
e.g. in countries with a two-party tradition and in countries with
multi-party tradition. One should also very carefully evaluate the
reactions of the incumbent politicians and the atmosphere among the
voters and other interest groups.
Use of descriptive real life like example environments would also
clarify the different purpose/intent and environment cases listed above.
It would be also good if people would more often indicate the
intended purpose and environment when they make comments on this list.
I'll come back if I find some example scenarios that would be more
natural and useful than the pizza examples and other regular stuff.
Juho
___________________________________________________________
Try the all-new Yahoo! Mail. "The New Version is radically easier to use" The Wall Street Journal
http://uk.docs.yahoo.com/nowyoucan.html
----
Election-Methods mailing list - see http://electorama.com/em for list info
