Proposa noves funcionalitats
#DecidimRoadmap Dissenyant el Decidim entre totes
Improve the score of comments
The ranking of comments uses upvotes−downvotes.
Thanks to David Ruescas (nVotes), I found a blog post from 2009 remarking that this is a wrong strategy (http://www.evanmiller.org/how-not-to-sort-by-average-rating.html). For instance, a comment with 100 upvotes and 0 downvotes and a comment with 1000 upvotes and 900 downvotes would be equally scored. One could expect this would be solved using the average rating (upvotes/(upvotes+downvotes)), however, such strategy is also inappropriate because there will be many comments with few votes and, therefore, high uncertainty.
To solve the above problem, the blog post suggests using the lower bound of the Wilson score confidence interval. Please find the formula implemented in Ruby in the blog post, as well as this detailed explanation of why Reddit is using this method https://medium.com/hacking-and-gonzo/how-reddit-ranking-algorithms-work-ef111e33d0d9
Llistat d'adhesions
Reportar contingut inapropiat
Aquest contingut no és apropiat?
3 comentaris
Conversa amb Xabier
Thanks @elaragon , this is a very valuable input. It also seems relative easy to implement. It is important to be able to explain this to participants as well. Because very complicated mathematical formulaes are hard to understand and trust on the platform is important for its success.
I totally agree, moreover, the simpler -> the easier to understand -> the "more democratic". For that reason, as shown in https://planspace.org/2014/08/17/how-to-sort-by-average-rating/, we could use Laplace smoothing instead: (upvotes+α)/(upvotes+downvotes+β) ; (for example, α=1 and β=2)
Sounds good too!
Deixa el teu comentari
Inicia la sessió amb el teu compte o registra't per afegir el teu comentari.
Carregant els comentaris ...