I have always been a big fan of rating films that I watch. Not only is it helpful for deciding if I want to rewatch a film, but it also helps me (and the IMDB recommendation system too!) understand what I like and dislike about films.
Over the past few months though, I have become progressively more frustrated with the traditional 1-10 rating system. The problem is that it is too one-dimensional. It is not unusual for me to watch a film and think “I really enjoyed that, but I don’t think it was a particularly good film”. Likewise, I have watched films that I thought were technically brilliant, but bored me to death. The 1-10 rating system does not allow for this distinction to be made. This leads to me giving films inconsistent ratings depending on my mood at the time of rating, which makes comparing ratings between films difficult.
Instead, I’ve decided to ditch IMDB and start recording my ratings in a custom spreadsheet using a two-axis system. The first axis will represent my enjoyment/engagement in the film and the second will represent my opinion of the film’s objective quality and value as a piece of art. For example, here a four films that epitomise each of the four quadrants:
- High enjoyment, high quality: Bullet Train
- High enjoyment, low quality: The Room
- Low enjoyment, high quality: 2001: A Space Odyssey
- Low enjoyment, low quality: Now You See Me 2
I will still upload ratings to IMDB so I can take advantage of their recommendation system, but I will do this automatically using their API and submit either the average of my two scores or some weighted average (I haven’t quite decided yet).
I’m also excited to see what the correlation coefficient will be between my two scores and will update this post when I have sufficient data to calculate it.