[EM] STV and weighted positional methods

[Raph Frank]

On Sun, Feb 1, 2009 at 7:56 PM, Kathy Dopp <kathy.dopp at gmail.com> wrote:
> Again you misconstrue my position. Like virtually all computer
> scientists who do not profit from selling or certifying computerized
> voting systems (and even some who do), I believe that there should be
> methods that average non-programming citizens can use to independently
> check the accuracy of vote counts.

Virtually all computer scientists?

> You on the other hand seem to desire to push for voting methods that
> would not allow ordinary citizens to check the accuracy of election
> results unless they hired a trusted computer programmer or unless a
> 100% hand count was done of every IRV/STV election and the ordinary
> citizens are allowed to and have time to observe.

Voters are always going to have to trust someone.  It's not like
everyone gets to hand count the ballots themselves.

If the ballots were published, it would be pretty easy to convert them
into a result.  Lots of programmers would probably do it for free.

> Again, I care much more about the public being able to have an
> understandable method of verifying the accuracy of election results
> than you do since you are promoting virtually the only voting method
> that makes it virtually impossible for an ordinary member of society
> to verify the accuracy of results.

PR-STV does allow that verification, you can make sure that the number
of ballots in each sub-pile is correct and that each ballot is in the
correct sub-pile.

> Yes, well the PR/STV counting method fails even more miserably at this
> goal of fair representation than does plurality, in ways that treat
> voters unequally and unfairly and do not allow a voter to even know
> how to vote to help his favorite or second favorite candidate have a
> better chance to win.

Rank the candidates in order of your choice is a perfectly reasonable strategy.

Some parties use 'vote management' which can slightly increase their
number of seats that they win.

Btw, I think you need to differentiate between PR-STV and IRV.  The
issues with IRV are considerably lessened with the switch to
multi-seat elections.