[EM] [ESF #1101] Why Not Condorcet?

[robert bristow-johnson]

On May 13, 2010, at 12:40 AM, Dave Ketchum wrote:

> On May 12, 2010, at 9:58 PM, Dr. Carl S. Milsted, Jr. wrote:
>
>> Condorcet is not simpler! Describing the results of a Condorcet  
>> election requires an NxN matrix. Makes my eyes glaze and I have a  
>> math degree. Most citizens do not. The potential for cycles is also  
>> frightening. Bush/Gore was bad enough.

i still have no idea who decided to arrange the Condorcet pairs in  
that NxN matrix they way they do and why they continue to do it.  it  
should be arranged as a triangle as follows (at least if there is a CW).

this is the same data of the 5x5 matrix taken from actual ballot data,  
the 2009 Burlington VT mayoral race (decided by IRV, but we had access  
to the individual ballot data and could run a hypothetical Condorcet  
election on it).  the number in <> brackets is the <defeat strength>  
for each pairwise defeat.  Dr. Milsted, if your eyes glaze over on  
that and cannot see exactly what the deal is, you might have to see an  
opthamologist:


   M 4064
   K 3477
    < 587>

   M 4597     K 4313
   W 3664     W 4061
    < 933>     < 252>

   M 4570     K 3944     W 3971
   S 2997     S 3576     S 3793
    <1573>     < 368>     < 178>

   M 6263     K 5515     W 5270     S 5570
   H  591     H  844     H 1310     H  721
    <5672>     <4671>     <3960>     <4849>

Condorcet, without a cycle, is simple to understand.  It is directly  
comparable to a 2-person race where no one disagrees how the ballots  
should be counted.  Ranked Pairs is not a particularly complicated way  
to resolve a cycle and cycles would be very rare.

--

r b-j                  rbj at audioimagination.com

"Imagination is more important than knowledge."