[EM] Cooperative or confrontational? Under approval & range voting.

[Warren Smith]

>I have never seen persuasive arguments about the incentives for
>candidates to be "cooperative" or "confrontational" under various
>untried methods.
>Kevin Venzke

--I agree.   OK, so let's start moving a tiny amount in that direction.
To simplify things let's talk about approval voting (the simplest system).

Suppose a candidate A by being "cooperative" with B can get
a fraction F (0<F<1) of the approve-B disapprove-A voters to also approve A.

Suppose a candidate A by being "confrontational" with B can get
a fraction F (0<F<1) of the approve-B approve-A voters to disapprove B.

And we'll assume there is no way for A to cause a polarity-switch
(B-and-not-A switching to A-and-not-B voting).

This model is entirely arbitrary and I see no reason to believe it,
but what the hell.

OK, then it would follow immediately which is the better strategy for
A to pursue
(confront or cooperate) and we find that it can be either one, it
depends on the present vote-counts:  If there are more B-but-not-A
voters than A-and-B voters, then
cooperate.

That was assuming A or B is going to win.  If however A thinks somebody else
(C, D, etc) also might win, then A is not getting much mileage by
being confrontational with B. A gets more mileage by cooperativity
with B because more A-approvals helps A against all rivals, but fewer
B-approvals only helps A versus B, not against C,D.

So based on this incredibly naive model, I guess I would conclude that
overall, approval voting inspires "cooperation" more than it
incentivizes "confrontation."
We can make essentially the same hokey model and get the same hokey
conclusion about range voting.

-- 
Warren D. Smith
http://RangeVoting.org  <-- add your endorsement (by clicking
"endorse" as 1st step)
and
math.temple.edu/~wds/homepage/works.html