Pairwise Rated Benchmarks

[Bart Ingles]

DEMOREP1 at aol.com wrote:
> 
> Mr. Ingles wrote in part--
> 
> A.  SINCERE SCENARIO:
> 
>       1.00 .90                       .10  0
> 45      A                              B  C
> 15      B   C                             A
> 40      C                              B  A
> 
> -----
> D-  How many times must I repeat myself ? -- There will be polls whatever
> method is being used.
> 
> A choice either gets majority acceptability or it does not (i.e. the same as
> in voting on legislative issues as in voting for executive/ judicial
> candidates).
> 
> Whatever method is used, minority choices (and their supporters) will attempt
> to strategize to defeat some choice based on such polls (with their counter-
> strategizing conspiratorial/ evasive/ lying voters).
> 
> A single winner election method will thus operate on the votes as cast (not
> on an infinite number of replayed strategized votes) to produce "a" winner
> (who is thus the "right"/"correct" winner).

Again, I only proposed the rated benchmark for use with hypothetical
scenarios, in order to find a way of saying "given these sincere
preferences, which candidate SHOULD win?"  Of course, in real life there
is no such thing as sincere votes, all election results are influenced
by strategy.


> As to the scenario, only C has the acceptability of a majority of the voters
> above 0.50.     C should win (without having to do the other math).

At least we agree on who should win in the above scenario, if not on
why.  I note that of the widely known methods, only IRO and Approval
(and possibly Bucklin) advocates would agree with us in this situation.

Bart