[EM] IRV inconsistency

Forest Simmons fsimmons at pcc.edu
Wed May 16 11:49:06 PDT 2001



On Wed, 16 May 2001, Markus Schulze wrote:

> Dear Forest,
> 
> here is an example that shows that the Condorcet criterion
> and the consistency criterion are incompatible.
> 
>    40 voters vote A > B > C.
>    35 voters vote B > C > A.
>    25 voters vote C > A > B.
> 
> Case 1: Suppose that candidate A is elected. Then when you add
> 51 A > C > B voters and 49 C > B > A voters, the winner is
> changed to candidate C.
> 
> Case 2: Suppose that candidate B is elected. Then when you add
> 51 B > A > C voters and 49 A > C > B voters, the winner is
> changed to candidate A.
> 
> Case 3: Suppose that candidate C is elected. Then when you add
> 51 C > B > A voters and 49 B > A > C voters, the winner is
> changed to candidate B.
> 

This confirms what I said, that a Condorcet failure of consistency will
only be in the case when one or more of the two precincts fails to have a
Condorcet winner.

> ******
> 
> You wrote (15 May 2001):
> > As Richard says, IRV ignores and trashes valuable information
> > willy nilly and still pretends to come up with a top notch
> > winner. All of IRV's shortcomings can be traced back to this
> > wasting of information.
> 
> It is not feasible to say that the one election method uses
> more information than the other. All election methods use the
> same information. They only interpret it differently. Election
> methods only differ in how important they consider which part
> of this information.
> 
> ******
> 
> You wrote (15 May 2001):
> > This is analogous to the fact that Condorcet methods violate
> > the Reverse Symmetry Criterion.
> 
> The Condorcet criterion and the Reverse Symmetry Criterion
> are not incompatible. There are good Condorcet methods that
> meet Reversal Symmetry. Please read Blake Cretney's website
> (http://www.fortunecity.com/meltingpot/harrow/124).
> 
> Markus Schulze
> 
> 

Thanks for correcting me on the RSC compatibility. Nevertheless, my main
point stands; any Condorcet method that fails the RSC does so only in
cases where there is no Condorcet Winner.

Forest



More information about the Election-Methods mailing list