Range Voting

[kevin_standlee]

The folks advocating Range Voting contacted WSFS (actually, the WSFS webmaster, sfrose) lobbying WSFS to change its voting system from the Instant Runoff Voting system we currently use for site selection and the Hugo Awards. Sharon told them how our rules work and suggested that if they want to change them, they come to WSFS business meetings and propose and debate the changes there, like all other rule changes. The advocate's response, in my opinion, amounted to, "Our proposal is so obviously Right that we shouldn't have to do all that hard, expensive work. You should change your rules because we tell you to do so."

I often tell people who come to me with rules-change proposals, "If you think it's worthwhile, come and submit it yourself. I'll help you with all of the technicalities to the best of my ability, but you have to make your own case, lobby people yourself, and get the votes by convincing people." Most of the time, this discourages them -- democracy is hard work! But sometimes we get people who are willing to work and debate, and sometimes we even get workable changes and improvements.

WSFS rules are intentionally designed to be resistant to change; however, they can be changed if people work hard enough at it. But it's not enough to just lobby a Board of Directors or subvert the Chairman; you have to convince the members.

Comments

Re: Disaster?

[thebrokenladder.livejournal.com]

With two candidates, strategic range voting is the same as single preference: everybody is going to vote 0 and 10 for the two candidates (assuming a 0-10 range).

This is called Approval Voting (http://rangevoting.org/rangeVapp.html), and it's a great voting method - vastly better than IRV. But since a lot of people vote honestly with Range Voting, it produces greater satisfaction and better representation of minor parties.

Also consider that with multi-winner Range Voting (called 'Reweighted Range Voting (http://RangeVoting.org/RRV.html)') there is a huge incentive to be honest with your scores, because in each round, your ballot is weighted based on the total of the scores you have given to already-elected candidates. Using a full range for single-winner elections as well, achieves continuity if nothing else.

With multiple candidates, the "ideal way" to vote strategically depends, as I've said before, on your estimate of how everyone else will vote...It's easy to show that voting 10 for your favorites, and 0 for the rest, is optimal for some value of .

Not necessarily, but usually.

But you don't know which value.

Which is a good thing - it prevents the problematic "bullet vote" strategy. But even if all voters knew the exact preferences of all other voters, and bullet-voted, this would just be to a Condorcet method. Range Voting at its worst is as good or better than Condorcet at its best.

So the simulations used a strategy that depended on voters knowing in advance the approximate results; what happens when those assumptions are erroneous?

Then Range Voting does even better, because the zero-info strategy is to guess your expected satisfaction with the outcome, and minimize your score for everyone who would make you less satisfied than that, and maximize your score for the rest. This removes the strategy to polarize your scores for two front-runners, both of whom you like. This is effectively honest Approval Voting.

The simulation results also didn't describe the distribution of preferences; if they were independent random variables, they don't resemble real preferences well enough to be useful.

At least two of the methods are described here (http://RangeVoting.org/vsr.html):

Here's the paper (http://math.temple.edu/~wds/homepage/rangevote.pdf). Utilities are generally based on a random Gaussian distribution of voters and candidates within n-dimensional issue space. A 2-issue space would be like a Nolan chart (http://is3.okcupid.com/graphics/politics/chart_political.gif). Each voter's utility is a function of his ideological distance from the candidates.

Ironically, the more issues we add, the closer the utilities comes toward matching a random distribution! Even more unintuitive, the results obtained from a simple 2-issue utility generator are negligibly different than those obtained by a straight random distribution. So it turns out, your intuition that random utilities would not sufficiently model real preferences turns out not to be correct. In other news, massive objects bend light, contrary to intuition.

As for bad examples, I can construct some (based on voters guessing wrong about other voters) where the winner is not the majority or Condorcet winner, nor produces the greatest social utility. In fact, he can be in the bottom half of the pack.

Expected value is value times probability, which is why Smith's calculations used hundreds of millions of simulated elections. For every situation where Range Voting performs poorly, there are more scenarios where IRV performs more poorly.

I also thought I also might mention this shout out (http://davidbrin.blogspot.com/2007/02/dark-light-scenarios-swinging-from.html) to Range Voting recently, from Hugo Award recipient David Brin.