On Nov 15, 2004, at 8:42 AM, Eric Gorr wrote:
At 12:13 AM -0800 11/14/04, Brian Olson wrote:Using the same engine as the Election Calculator I ran the preliminary data through a collection of election methods. It turns out that no race was close enough to wind up with different results when a different method is used. But, if you want to see Condorcet or Ranked Pairs tables, or IRV counts, or whatever, here are the results:
http://bolson.org/voting/sf2004/
Very nice find.
I had no idea they were going to release the individual ballots.
Fortunately, it is only a matter of time before there is a case of a close race where the method used (and perhaps even IRV tiebreak used) would make a difference in who wins.
I don't suppose you have the ballots in the form of 1:a>b>c ?
It would be easy enough for me to write the code to get them into that form, but I figure you've probably either done it or written the hardest part already which would be to parse the text.
Yeah, raw data, perl script and resulting partially digested '>' form of ballots at http://bolson.org/voting/sf2004.tar.bz2
(785374 bytes compressed, 30433280 expanded)
I plan to update this when final data is available.
Brian Olson http://bolson.org/
---- Election-methods mailing list - see http://electorama.com/em for list info
