django-ratings
django-ratings copied to clipboard
dcramer
explaining weight
this is not an issue - I´m just not sure about the implementation of "weight" (I probably just don´t understand it). with IMDB (for example), the weight refers to the total votes for an object. with django-rating, the weight is a static value. do you have any additional information about this? thanks in advance.
I suppose it may be specific to my use cases, but it's designed so that if you are using some kind vote system where once it hits like a "+4", that the +4 can't be reached by simply 4 up votes. Basically, it adds N votes on to the end of it, that are all neutral.
I´m sorry, but IMHO this is far to simple ... http://www.imdb.com/help/show_leaf?votes http://en.wikipedia.org/wiki/Weighted_voting http://www.ctl.ua.edu/math103/POWER/wtvoting.htm
here´s the IMDB voting formula:
weighted rating (WR) = (v ÷ (v+m)) × R + (m ÷ (v+m)) × C
where: R = average for the movie (mean) = (Rating) v = number of votes for the movie = (votes) m = minimum votes required to be listed in the Top 250 (currently 1300) C = the mean vote across the whole report (currently 6.8)
Very interesting. I've been looking into different ways to allow anonymous + registered votes, but have anonymous be much less meaningful.
might come down to the usecase ... here´s mine: we´re having a movie-site where users can rate movies (and other stuff). let´s say "fight club" gets 10 votes with 5 stars for each vote (which is the maximum score). on the other hand, "buffalo 66" gets 300 votes with an average of 4.56 stars. question is - which movie is rated better? I´d say "buffalo 66", but that depends on a) the average rating over all movies and b) all votes given to movies.
When I get time (unless someone beats me to it), I will be integrating (most likely) IMDBs algorithm for weighting, and it will replace my current half-ass solution :)
Hi, really like this project and they way you have implemented it! Have you read this article on ratings? http://www.evanmiller.org/how-not-to-sort-by-average-rating.html
It is based on: Score = Lower bound of Wilson score confidence interval for a Bernoulli parameter
