I accidentally watched Enola Holmes. It was bad. It’s a movie for children, more specifically female children. Not that there’s anything wrong with that, just that I’m not a female child, nor do I have one.
It was not immediately apparent, as with, say, Frozen. You see the word “Frozen”, you know it’s not for you. You’re too old.
Now, I want a better way to filter what I watch.
I’m not looking for a recommendation engine, rather I’d like to have a reason not to watch something. Especially if it’s something aimed at women (Outlander) or children (Enola Holmes).
I started with Rotten Tomatoes.
It’s a weird place to start, because Rotten Tomatoes is what got me into this mess in the first place. Certified fresh, 90% and not “rotten” for Enola Holmes. Apparently, these 2 metrics in isolation are not enough.
So my first reaction was to maybe search for some curator or critic who marked this movie as rotten, and try to go from there. But this critic would need to watch basically all 1 million movies and TV series in existence, for this “non-recommendation” approach to be useful. Need another solution.
Next stop was IMDB.
I was pleasantly surprised on 2 counts.
First, user reviews on IMDB have an upvote feature, which means Enola Holmes gets a well-deserving “Perfect for a 15 yro girl” rating.
Another important metric, that apparently only exists on IMDB, is user ratings by age group and gender. It also only exists on desktop version, not the app.
Below are IMDB user ratings for Outlander, comparing rating by demographic. (https://www.imdb.com/title/tt3006802/ratings?demo=males_aged_18_29)
If discrepancy of 0.4 between 18-29 boys and 30-44 men is not enough, check the difference of +0.9 between males and females of all ages.
The process I’m using right now involves IMDB app + website and Rotten Tomatoes site. You search for the movie in IMDB app, then:
1. check out the average ratings,
2. then open reviews to see the most upvoted ones,
3. finally check the ratings by demographic in desktop version of IMDB.
Finally, I might also check out Rotten Tomatoes.
Verifying the process
Ok, let’s check out 2 movies and 2 TV series to check if this stuff works. I don’t have exact criteria for red flags, everything was added manually.
Rotten Tomatoes Critics 90%
Rotten Tomatoes Audience 65%
🚩Rotten Tomatoes (Audience-Critics) difference = -35%
🚩Male / female difference = -0.5
🚩Top comment = “Perfect for a 15yo girl”
All [18-29] vs [30-44] = -0.1
Rotten Tomatoes Critics 89%
Rotten Tomatoes Audience 90%
Rotten Tomatoes (Audience-Critics) difference = +1%
🚩IMDB (Male - female) difference = -0.9
Top comment = “Outstanding!”
🚩IMDB (Male [18-29] - [30-44]) = 0.4
Giri / Haji
Rotten Tomatoes Critics 100%
Rotten Tomatoes Audience 88%
Rotten Tomatoes (Audience-Critics) difference = -12%
IMDB (Male - female) difference = -0.2
IMDB Top comments by Helpfulness = “Just Perfect!”, “Gets better”
IMDB (Male [18-29] - [30-44]) = 0.2
Attack on titan
Rotten Tomatoes Critics 94%
Rotten Tomatoes Audience 93%
Rotten Tomatoes (Audience-Critics) difference = -1%
IMDB (All Ages Male - female) difference = -0.1
IMDB Top comments by Helpfulness = “Masterpiece. I’ve never witnessed this level of care for every detail”
🚩IMDB (Male [18-29] - Male [30-44]) = 0.3
How to get this info, fast
Going through this process for multiple movies is time-consuming. So how to speed it up?
I didn’t find such a way, crucial info from IMDB ratings by age and gender is not easily available through public APIs or ready-made applications.
Logical solution of course if to drop everything and build this solution myself.
Some math. There are 560k feature films and 196k TV series in IMDB. So we can round that up to about 1MM movies.
For this solution to be in any way useful, it needs to already have 1MM movies in database + more than 1,000 new movies and god knows how many TV series are released every year.
For now, I’ll just spend 2 more minutes to scroll both IMDB and Rotten Tomatoes.