[EM] [CES #3605] Re: Kemeny Condorcet method. Apparently not a good choice for those of us who want to know who won in our lifetimes.

[Warren Smith]

Kemeny might be ok with me in situations where we could be sure the
#candidates was always <20.

But for general purpose use, sorry, the NP-hardness combined with the
empirical algorithm experiments is a deal breaker for me.  I now
suspect if I tried to create hard elections intentionally (as opposed
to those experiments generating elections from real data or by certain
random generation methods) then I could probably create random
too-hard-to-do elections with only, say, 27 candidates... here's an
initial try, consider this pairwise matrix:
    http://www.RangeVoting.org/Tourn27.html
and replace all the +1s by random numbers in the interval
   [9000000, 10000000]
and all the -1s by ditto but negated, to get the pairwise margins matrix
for a 27-candidate election.   Good luck trying to find the Kemeny
winner or order...  can anybody do it reliably, or is this usually
beyond humankind's abilities?

D.Felsenthal in a survey of academic voting experts said Kemeny was
his top favorite out of 18 methods he ranked, while Range Voting was
his bottom of the 18.  Range voting does have the slight advantage,
though, that the election winner can actually be determined by
humanity.