Seems unreasonable to expect the voters to know 500 candidates. The winner would likely be someone who is well known to many of the voters, though not necessarily in a way that makes the winner qualified for the job (Schwartzenegger), or the winner might be very popular with a plurality of the voters, but disliked by everyone else (David Duke, Jean-Marie Le Pen).
I would suggest requiring candidates to collect signatures (people should be allowed to sign more than one petition, unlike Texas). This narrows down the field to those candidates who are motivated to go after the job, or who have lots of supporters who are willing to help gather signatures. Then the 10 candidates who get the most signatures would be qualified as candidates (could still have write-ins) and could campaign before the election was held. As you imply, candidates listed near the top of the ballot are more likely to win. That effect can be alleviated by "rotating" the candidate names, starting different ballots with different candidate names. Which voting method should be used? Well, I like Range Voting (or Score Voting, I prefer to call it). RangeVoting.org. I believe it chooses winners with better social utility than other practical methods. Cheers, - Jan On 7/3/06, dave smith <[EMAIL PROTECTED]> wrote: > Hi All, > > I need help in determining what voting system would be best for the > following criteria: > > ~500 candidates > ~1000 voters > 1 winner > > I was going to do a form of PV, but limit the number of ranks to 3. > The problem with this is that what if there are more candidates than > voters, and the voters decided to stop paging through candidates after > the 3rd - 4th page of candidates. > > Maybe an approval system would be better here? > > Thanks > ---- > election-methods mailing list - see http://electorama.com/em for list info > ---- election-methods mailing list - see http://electorama.com/em for list info
