[EM] Condorcet cycles in Romanian 2009 election?

[Warren Smith]

>seppley: Without studying details of the three Romanian candidates and the voters'
preferences, the explanation of this majority cycle cannot be known for
sure.

--Indeed, it's very hard for me to understand it without knowing
Romanian, and even if I did, it'd still be hard.

>However, consider a case of three very similar candidates. The voters'
preferences in each of the three possible pairings would be nearly tied
(approximately 50% preferring each candidate over each other candidate).
In such a case, a cycle involving three small majorities would not be
rare. Almost an even bet?

--actually this is basically exactly what happens in the "random election model"
and no, it then is not an even bet; it is 8.78% probability you get a cycle.
Gilbaud's number.  Somewhat discussed in the paper
   http://rangevoting.org/Romania2009.html

--the paper is now updated, I got confirmation and denial of the
validity of certain poll results, and found approval- and
range-voting-style polls.

It seems Oprescu would have won with either approval or range voting.

Oprescu or Antonescu would have won with the good pairwise-table-based
Condorcet methods.

Geoana would have won with asset voting.

Basescu won with plurality and plurality+top2runoff, the official methods.

With IRV, it is unclear who would win among {A,B,G}.

It looks like 99.9% confidence there is a Condorcet cycle involving both B & G.
However, it is unclear whether there is no Condorcet winner, or whether it
is Antonescu.  That is because the A-vs-G and A-vs-O pairwise results, are the
least understood and least clear.

If any of you have some Romanian friends who can offer some insight or
point to some more poll data that maybe I missed, that'd be
appreciated.  Also, higher-accuracy
versions of the poll data I already have, would be appreciated too.

AN IMPORTANT QUESTION I DO NOT UNDERSTAND:
The Romanian voters residing in foreign countries constituted about
1.4% of the total voter pool and heavily favored B over G (whereas the
ones residing in Romania were
split 50-50 between B & G).   It would be good if we knew how these
expatriates would
have felt about all the other pairwise comparisons amongst {A,B,G,O}.
It is thought that this was because a lot of them are in Moldova and B
has loudly supported rights of Moldovan Romanians to get Romanian
citizenship.  What were the
stances of A,G, and O on this issue, I wonder?

-- 
Warren D. Smith
http://RangeVoting.org  <-- add your endorsement (by clicking
"endorse" as 1st step)
and
math.temple.edu/~wds/homepage/works.html