71 Messages
•
1.8K Points
anomalous "B positive" series rating versus episode ratings
The new series "B positive" has an unusual series rating. The series rating is (5.1) an outlier versus the 3 episodes with ratings (6.0, 7.1, 7.7). The series rating itself has an anomalous distribution (137 out of 603 votes are 1-star). IMDb has safeguards against ballot-stuffing, but these 137 1-star votes are skewing the series rating. I would like to think that I could convince 137 of my friends to do anything (if I were so inclined). :-)
bderoes
Champion
•
5.1K Messages
•
118.7K Points
5 years ago
https://www.imdb.com/title/tt11839500/ratings
Seeing the difference between series and episodes here really lends credence to the request (posted somewhere here on sprinklr) that a series rating should be constructed from the episode ratings.
Seeing that arithmetic mean of the 21 Top-1000 voters is 3.8 (7 gave it a 1) makes me wonder if the right voters are included in the Top-1000 (which I believe should be expanded to 10,000). Do these voters have normally-distributed ratings patterns? Or do they usually vote at the extremes? Have they voted with consistent frequency over a long period of time? Or have they voted for a _lot_ of titles in a brief period of time?
I know IMDb is not going to answer those questions. They keep their algorithms private to avoid giving the ill-intentioned the ability to circumvent them. But I wonder if they've considered the Top-1000 membership well enough. I see a _lot_ of low ratings among that group for most of the titles where I look at the ratings demographics.
0