Dear Debian community: My name is Clay Shentrup and I am an Ubuntu user - so I have much respect for your endeavors and your hard work on them. I write to discuss a simple improvement that could be had with your elections, merely by changing to a better and simpler voting method: Range Voting.
You currently use Condorcet voting. It is ironic that Condorcet methods are differentiated by their handling of a Condorcet cyclic ambiguity, because such ambiguities by their very nature refute the underlying axiom of Condorcet voting: that if A is preferred by a majority to B, he is a better winner. We at the center for Range Voting employ what we feel to be a more rational metric called "social utility efficiency". This is a measure of how good a voting method is at giving people what they said they wanted - a voting method with a greater SUI makes more voters more satisfied. Warren D. Smith, the Princeton math Ph.D who founded the Center for Range Voting, explains this concept in detail at the links below. He calls the term "Bayesian regret", whereas social utility efficiency is a way of turning Bayesian regret numbers (which are sort of like "cubits" in that they have arbitrary magnitude) into a pure ratio. The values are scaled such that a 100% is the candidate we'd select if we were to be able to read the voters minds and pick the one who made the most voters the most happy. A 0% social utility efficiency is, by definition, the utility that would be produced by randomly selecting a "name out of a hat". See: http://RangeVoting.org/UniqBest.html http://RangeVoting.org/BayRegDum.html http://RangeVoting.org/vsi.html We have a page specifically addressing the many advantages of Range Voting over Condorcet voting, including the fact that it is massively simpler. http://rangevoting.org/CondorcetExec.html Dr. Smith also has done some analysis of Debian's elections, which represent a wonderful dataset for scientific study. http://rangevoting.org/Debian2003.html I personally came into the Range Voting movement after independently inventing Condorcet voting last summer while living in Portland. I decided to see whether anyone else had thought of the idea, and quickly stumbled upon the Center for Range Voting web site. My first writings to Warren D. Smith were extremely skeptical, and admittedly I did not become a convert overnight. Nevertheless, after much extensive scrutiny, I came to the conclusion that Range Voting is objectively a superior voting method, that will give the Debian project more convenient elections with better outcomes. Regards, Clay Shentrup San Francisco 415.240.1973
