[EM] Re: Issues, Condorcet, and IRV (was: IRV vs. plurality)

[Dave Ketchum]

Alex Small wrote in part:
 > Adam Tarr said:
 >
 >>>So Condorcet gives 2 of the 3 parties an incentive to "go 2D" in this
 >>>example, while IRV only gives 1 party an incentive to "go 2D" in this
 >>>model.
 >>>
 >>Doesn't that constitute a strong argument for Condorcet, given your
 >>views?
 >>
 >
 > Possibly.  The analysis I presented suggests that IRV will keep us in the
 > 1D rut, and Condorcet _might_ get us out of the 1D rut, since IRV can
 > embrace competition in a 1D issue space, while Condorcet DEMANDS 2D issue
 > space for meaningful competition.
 >
 > If Condorcet still fails to enlarge our issue space we have a choice:
 > Permanent centrist monopoly via Condorcet, or continued competition
 > between left and right via IRV, with the need to edge out the center
 > moderating them at least a little.  They have to go close enough to the
 > center to assure themselves more first-place votes.
 >
 > I suspect that I'd still take the centrist monopoly, at least for a while.
 >  If the centrist monopoly grew stale and corrupt then that would introduce
 > a second axis to supplement the left-right division:  fresh vs. stale.
 >
 >
 >>Empirically, we know that politics tend to remain 1-D, with two dominant
 >> parties, in IRV.  Maybe Condorcet would be different.  My intuition,
 >>and  your example. both suggest that Condorcet would lead to parties
 >>offering  different permutations of "left" and "right" viewpoints.
--------------------------------


I quote the above as an example of describing IRV and Condorcet as 
different methods with characteristics that are cause for choosing between 
them.

Makes NO SENSE to me:
      Ballots and basic voter instructions are identical (true that for 
IRV there can be plots to take advantage of spoilers).
      Except for cases involving IRV spoilers, winners are IDENTICAL - 
repeat IDENTICAL!!!
      Condorcet backers emphasize the IRV spoilers because those are the 
major reason for choosing Condorcet over IRV (Condorcet claims simpler 
counting of votes as an additional advantage).

So, HOW OFTEN do the IRV spoilers affect results, and is the spoiler 
effect desirable to promote?  After attending to the candidates with few 
votes, which both attend to together (although via different methods):
      One or two candidates left - identical since spoilers cannot happen.
      More than three left - possible, but I will do three as an example:
           A - voted by some.
           B - liked by those who dislike A.
           BC - these B backers see C as second choice (not usually 
mentioned in this example but, when CB exists BC almost certainly will 
also exist and I want to mention BC in the discussion).
           CB - these B backers like C even better - likely B and C share 
basic positions with a minor variation that has caused the two candidates 
to exist.

      A < B+BC+CB - or A would win.
      A > B+BC - or A would lose and leave it to B and C.
      A > CB - ditto.
      CB > B+BC - or C would lose and B would win.
      A > BC+CB - or C would win when B lost.

Thus we have:
      A > BC+CB > CB > B+BC - as the case for spoiler.

Note the boundary effect as additional CB votes move the results into the 
spoiler area in IRV:
      Less CB votes - B wins.
      More CB votes - >>>>A<<<< wins!!! - not what the C backers desire!!!

-- 
davek at clarityconnect.com  http://www.clarityconnect.com/webpages3/davek
  Dave Ketchum   108 Halstead Ave, Owego, NY  13827-1708   607-687-5026
            Do to no one what you would not want done to you.
                  If you want peace, work for justice.