[EM] Summability

[Alex Small]

Markus Schulze said:
>> I'm surprised to read this.  I thought "simple strategy" was a
>> virtue for an electoral method.  Surely runtime isn't considered a
>> serious issue for summable methods...?
>
> No! It is a desirable property that there is no simple way to
> manipulate the result of the elections.

That depends on what you think of insincere voting.

I don't see anything wrong when people vote in whatever way they think
will best advance/protect their interests.  That's sort of what a public
election is all about.  Seen in that light, manipulation should be simple
when it is possible.  There should be a simple cause-effect relationship
between what a person marks on his/her ballot and what the outcome is.  It
shouldn't be too complicated to determine whether sincere voting is the
best way to go.

At the same time, the incentives for insincere voting should be as few as
possible.  If the goal is to pick "the choice of the people" it's best
that the information used to pick this person be an accurate reflection of
the will of the people.

I'll refrain from commenting on which methods I think best satisfy these
two criteria.  I will take this opportunity to trash Borda:  If multiple
factions vote insincerely to elect whoever their favorites are, the result
can easily be the election of somebody that NONE of the voters were
strategizing to elect.

Finally, I don't like using the word "manipulate" to describe strategic
voting.  The negative connotations of "manipulate" generally involve
outside forces deceiving people to achieve a certain end.  A bogus poll
that tricks people into picking a particular strategy is manipulation. 
But looking at accurate information and deciding to support a compromise
candidate is no different from politicians forming coalitions to advance a
certain cause, or two sides compromising so an issue can be resolved.



Alex