The Key to Disseminating Life as Fiction Ratings
The reason I started doing ratings was because I wanted to remember, in a quantifiable manner, how I felt about a film. Considering the amount of films I see every year, it becomes almost necessary to be able to recall the degree to which I enjoyed something. However, it is and has never been a perfect science. They're just shortcuts for me and for the reader, and shortcuts are often a risky way to get to the final destination.
Update on February 13, 2007: Having now done 120 commentaries, I've finally started understand how I view the ratings. So, here's a simple update that should be more useful than what was here before.
As close to perfect as possible. Everything clicks.
Nearly perfect, something that will stay with me for quite a while. Probably pretty original, to boot.
An impressive work and arguably the best rating I can give without the film being nearly perfect.
Enjoyed it quite a bit, but had some nagging flaws that kept it from being truly special. Definitely better than average and potentially fantastic.
The film works. I enjoyed it, but maybe it lacked the depth necessary to stand out. Or, quite often, it's used as a "disappointment" rating, whereby films receive it because I expected a lot more but I only got a moderately satisfying film (see Children of Men).
Generally considered the lowest level at which I will actually recommend a film, but even that's rare. Uusally it's a movie with a few qualities that stand above what is otherwise a mediocre production.
This is my basic "I don't really get it" or "I finished it but would never think about watching it again" rating.
Didn't enjoy it and/or like it at all. In fact, I often hate a film and still give it one-and-a half stars (see Flags of Our Fathers) simply becuase of technical reasons.
Almost as rare as a five star rating. Why? Because this essentially means the film lacks any redeemable value and/or was horrible beyond belief.
Ratings are skewed due to a survivor bias: I usually only watch films I'm interested in, so a natural distribution occurs but is skewed towards a rating of 3.0. This signifies that I will like most of the films I watch (or else why watch them?), and will undoubtedly differentiated from a mass market distribution that would be attained at a website such a rottentomatoes.com or metacritic.com. For example, with 100 films watched in 2006, the ratings breakdown has been roughly:
As as result of the above survivor bias, there is heavy frequency in the middle but with steep drop-offs at the ends. For those interested, the average of the first 100 films from 2006 is 2.85, the median is 3.00 and the standard deviation is 0.81.
Most Recent Commentaries
Information & Basic Archives
View by Rating
View by Region
Related Externalities