DeletedUser
Guest
I think it'd be best to use positive reputation only, but I'm more a supporter of a rating system using Bayesian averages. It's a rating system that uses popularity, voting averages, and the actual votes to give a rating to something. It works great for user voted lists.
Here's the formula for Bayesian averages:
(v ÷ (v+m)) × R + (m ÷ (v+m)) × C
R = average rating
v = number of votes
m = minimum votes
C = mean vote
Here's the formula for Bayesian averages:
(v ÷ (v+m)) × R + (m ÷ (v+m)) × C
R = average rating
v = number of votes
m = minimum votes
C = mean vote