[EM] Re: IRV in action.....Improved IRV

[Chris Benham]

James,
You wrote : "Your example..did not really present the issue I presented".
Maybe so, but the METHOD  I suggested certainly does, as I explicitly 
mentioned. To take your
first example:
10  FR > R
38  R > C > FR
  3  C
39  L  > C > FL
10  FL > L
C would be the third candidate to be eliminated. At each step we 
eliminate the Condocet loser
from the set of candidates whose tallies are below average and also not 
above 25%. Therefore
the initial set of candidates for elimination is  C , FL and FR. C 
certainly pairwise beats FL and
FR, so say FR is the Condorcet loser of  those three and  is therefore 
the first eliminated.
This gives :
48  R > C
  3  C
39  L  > C
10  FL > L
The new set of candidates for elimination is C and FL. C pairwise beats 
FL and so FL is gone.
This gives your second example:
48  R
 3   C
49  L
C ( being not above 25% and below average) is eliminated, so what is the 
problem ?
 
Chris Benham