[EM] IRV non-monotonicity probability
warren.wds at gmail.com
Tue Mar 17 17:51:20 PDT 2009
now summarizes knowledge on this - updated & corrected.
Summary at end.
I think the Quas-Smith theorem is rather devastating theoretically speaking -
you get 100% limit probability (if large # candidates) of having
nonmonotonicity existing in your election in a huge number of ways
I mean, that's about as bad as one could imagine.
[Rather annoyingly, though, neither Quas nor I have
proven this for the "1D political spectrum" model. It seems clearly true
since Quas's computer has produced convincing numerical evidence for it,
but no proof. Proving this would be the icing on the cake.]
Warren D. Smith
http://RangeVoting.org <-- add your endorsement (by clicking
"endorse" as 1st step)
More information about the Election-Methods