[gleenn]

We actually switched to Wilson score. Doing it later has some weird effects, when you've already have a lot of people typically voting on the first definition, and then suddenly the order gets switched because something has a higher ratio giving it higher confidence. We're honestly not sure it's done anything that great for UD, sometimes simple is just better.

[thanatropism]

You might want to weigh less controversial (as in abs(upvotes - downvotes) higher. This would be somewhat like Effect Size in science.

The Bayesian approach would be to assume the true vote distribution is binomial and use a beta prior (possibly with Jeffrey's degenerate bimodal prior). Then as the total number of votes increases the posterior distribution tightens. Ranking score is prob(score>0).