in reply to Sorting Votes, Confidence & Deviations
The bottom line issue is that you want to show two orthogonal metrics with one number - interest (the number of votes) and rating (the pattern of votes).
So the problem becomes "How do I map two parameters to generate a single figure of merit" which is pretty much what you have provided some methods for. Bearing in mind though that you have two orthogonal parameters, a different way to show the vote result is as a table where rows represent (some quantisation of) ranking and columns represent (some quantisation of) interest. Method3 comes close to this concept.
|
|---|