[EM] Empirical data on cycles

Adam Tarr ahtarr at gmail.com
Sat Sep 3 21:03:10 PDT 2005


Thanks for the info Andrew. Some commentary below.

On 9/2/05, Andrew Myers <andru at cs.cornell.edu> wrote:

I thought the folks on this list would find it interesting to see
> some actual empirical data on how often cycles happen. I have data on
> 99 CIVS elections that have been run in which more than 10 voters
> participated (max was 1749) and in which there were at least three 
> candidates
> (max was 72). These 99 elections break down as follows:
> 
> had a Condorcet winner: 85
> no Condorcet winner, but a unique unbeaten candidate: 7
> multiple unbeaten candidates in real ties: 3
> real cycles requiring completion: 4
> 
> These results suggest to me that the concern about cycles arising in 
> Condorcet
> methods is a bit excessive.


It suggests to me that _natural_ cycles are very rare. This does not 
automatically mean that cycles can never be a problem. The important thing 
is to pick a Condorcet method where, when a Condorcet winner exists in 
sincere preference, it is extremely rare than any faction has a tactic where 
they can cause a favorable cycle. (I am referring, of course, to winning 
votes.)

This is anecdotal, but I looked at the four elections in which cycles 
> occurred,
> and my impression was that these were usually elections with a lot of
> candidates and a poorly informed electorate that couldn't effectively 
> judge
> between them.


I think that's probably true, but the possibility does exist for natural 
cycles of sincere and informed preference. It would be interesting to see 
if, with Condorcet voting, more candidates started popping up whose views 
fit less comfortably in the traditional left/right political spectrum.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20050903/eee6d469/attachment-0003.htm>


More information about the Election-Methods mailing list