I'm working on a website where people can vote for their favorite songs and I'll have a rankings page. I'm want to use a Bayesian Estimate formula to calculate a weighted ranking that takes into account the number of votes a song has. I want to use the formula that IMDB shows at the bottom of this page. My question is in regards to the value of C, which it says is the "mean vote across the report".
Does this mean it's the average of all the votes together, regardless of who the vote was for? Or is it the average of each individual average?
I.e., if I have songs A and B. A has votes (10, 5, 3) and B has votes (7, 2). Is C = (10 + 5 + 3 + 7 + 2)/5. Or is C = [(10 + 5 + 3)/3 + (7 + 2)/2]/2?
I can't tell which one it means, and I haven't been able to find anything on Google.
Thanks for your help!
Does this mean it's the average of all the votes together, regardless of who the vote was for? Or is it the average of each individual average?
I.e., if I have songs A and B. A has votes (10, 5, 3) and B has votes (7, 2). Is C = (10 + 5 + 3 + 7 + 2)/5. Or is C = [(10 + 5 + 3)/3 + (7 + 2)/2]/2?
I can't tell which one it means, and I haven't been able to find anything on Google.
Thanks for your help!
Facebook
Twitter