[EM] Introduction (cont.)

[Buddha Buck]

Roy One <royone at yahoo.com> writes:

> Forest Simmons <fsimmons at p...> wrote:
> >>It's always nice to hear that the strategic case of Range voting
> >>optimizes some desirable quantity, since the strategic case of
> >>Range voting is equivalent to Approval.
> 
> Why? It seems to me that strategic voting would be to normalize your
> first and last choices to the extremes of the range, and distribute
> your in-between choices proportionally.

Let's say we are working on a -100 to 100 range voting scale (the
range can be picked arbitrarily, but I chose this range to mollify
DemoREP).

There are 6 candidates, whom I fairly, after great consideration of
the issues, sincerely rank as follows:  

  Albert    87, 
  Bill      52, 
  Charles   18,
  Dave     -17,
  Edward   -48,
  Frank    -94

When voting with strategy in mind, of course I want to maximize the
chances of Albert winning and/or Frank losing, so those candidates are
going to be ranked 100 and -100 respectively on my ballot.

But what about Bill through Edward?  The problem is that if you think
Albert might lose, you have to throw more support behind Bill, so that
your #1 compromise candidate has the best chance of winning.  So
Bill's ranking get's pushed higher, to increase his chances.  If you
have any doubts about Albert's electability, then you push Bill as
high as possible -- say, to 100 as well.

Basically, if you have any fears that Dave, Edward, or Frank could
win, it's in your best strategy to push Bill and Charles to max as
well as Albert.

Likewise, if you have any fears that Dave, Edward, or Frank could win,
it's in your best strategy to push them to the minimum, as well.

Adding information isn't going to help.  If you know that it's a
horserace between Bill and Charles (with the other 5 candidates having
not very good support at all), it behooves you, strategically, to max
Bill and Albert and min Charles and the rest.  

> Is it the case that the strategic case of any system that allows ties
> is equivalent to Approval?

I don't think so, but I haven't given the matter a lot of thought.

> 
> Roy
>