[EM] Doubts on Approval Strategy

[Richard Moore]

Rob Lanphier wrote:
> Let's assume a simple left-right spectrum, with candidates A, B, C, and D
> (far left to far right).  Let's also say that candidates B wins in a world
> where there's accurate data, and everyone votes according to expected
> strategy (where "B" and "C" are the only candidates who get or come close
> to getting majority approval). However, let's say that A has pretty strong
> support, with about 30% of the electorate wanting A as a first choice.
> 
> Now we've got an interesting problem.  Prior to the election, "A" 
> supporters may not want to tip their hands that they'd settle for "B", in
> hopes that "A" can be considered the "front-runner" candidate that "B"
> voters should also approve.  So, when asked by the pollsters, they "lie",
> saying that they only approve "A". 

I've thought about poll manipulation strategies before. I still don't see
a way polls can be exploited with any reliability.

Say my favorite is A, my compromise is B, and my most despised candidate
is C.

You might be able to manipulate how I vote on candidate B by either
artificially inflating or deflating the support for either A or C
in the polls. Obviously your strategy won't be able to change how I
vote on A or C.

But if your strategy can change my vote for B, won't it have the opposite
effect on a voter who favors C and hates A? I might already have been
voting your way on B, and your strategy may have cost you that other
voter's vote. So I don't think poll manipulation can be done reliably.

It will be even less reliable if there are more than three candidates.
Maybe my preferences are A>B>C>D>E, and you want to persuade me to vote
for C by making it look like E is a strong candidate. But if I've already
decided I'll vote for C, maybe this will cause me to vote for D as well.
It's not unlikely, especially in a multi-dimensional issue space, that
your favorite and your most despised candidate appear next to each other
in my ranking. So maybe you despise D. This provides another way for poll
manipulation to backfire.

Seems to me, anyone wanting to do this sort of manipulation to increase
the probability of a particular outcome would have to have extremely
accurate and detailed polling information to begin with.

They'd also have to communicate the polling strategy to other members
of their faction, without leaking the existence of the strategy to the
general public (otherwise everyone would lose trust in the polls and
default to zero-info or history-based strategies).

As for subconscious poll manipulation, sure that will happen (as I'm
sure it does today), but I think it will only produce many tiny random
fluctuations that tend to cancel each other out.

  -- Richard

----
For more information about this list (subscribe, unsubscribe, FAQ, etc), 
please see http://www.eskimo.com/~robla/em