How to Find Good Movies to Watch (Without Scrolling for an Hour)
You cannot find a good film by browsing, because browsing is designed to keep you browsing. The method that works is simple. Follow four or five people whose taste you trust instead of an algorithm, keep a list of no more than ten films you have already decided to watch, and apply one constraint when you sit down, such as made before 1999, or under a hundred minutes, or free tonight. Then take the top of the list and start it. The choice should take two minutes, not forty.
Follow people, not algorithms
A recommendation engine optimises for what keeps you on the service. A critic optimises for what is good. Those are not the same target and the difference compounds over a year.
Pick four or five people and stick with them. Roger Ebert's Great Movies archive is free and still the best starting point in English. Letterboxd lists from people whose four star reviews you tend to agree with will beat any home screen.
Keep the list short and closed
A watchlist of two hundred films is not a list, it is a museum of your own indecision. Cap it at ten. When you add an eleventh, remove one.
The point of a short list is that the decision has already been made. You are not choosing at 9pm on a Tuesday, you are just starting the thing at the top.
Use a constraint to kill the paralysis
Constraints make choosing easy. Made before you were born. Under ninety minutes. Free on Tubi tonight. Directed by someone you have never seen a film by.
Any constraint works because the aim is not to find the optimal film. It is to start a good one. There is no optimal film and looking for it is the whole trap.
Make the last one point at the next one
The best source of your next film is the one you just watched. Look up who directed it and what else they made, or who wrote it, or which film everyone says inspired it.
This is how taste actually develops, in chains rather than in random draws. Two years of chains and you will have seen things no algorithm would have surfaced.
Give it a deadline
Nothing gets chosen without a deadline. A film club, a standing Friday night, a friend who is watching the same thing this week. It sounds small and it changes everything, because now not choosing has a cost.
- Browsing is designed to continue, not to conclude. Do not use it to choose.
- Follow four or five people with taste instead of a recommendation engine.
- Cap your watchlist at ten films and apply one constraint at the moment of choosing.
- Let each film point to the next one, and give yourself a deadline so choosing has a cost.
The whole idea
Watch one movie this week. Talk about it Friday.
We pick one film. You watch it alone, on your own time. Friday at 7:30pm PT you get ten minutes on Zoom with one other person who watched it too. No club, no homework, no small talk.
See this week's pick $5Common questions
- Why can I never find anything to watch?
- Because you are choosing at the moment you want to watch, from an interface built to keep you choosing. Move the decision earlier, cap the list at ten, and add a constraint.
- Where can I find good movie recommendations?
- Roger Ebert's Great Movies archive, Letterboxd lists from reviewers you agree with, and The Criterion Channel's curated collections. Four or five trusted sources beat any algorithm.
- How long should choosing a movie take?
- Two minutes. If it regularly takes longer, the problem is the size of the list, not the size of the catalogue.