Constantly updated film recommendations, based on your tastes.
How does it work?
You login with your Trakt account.
Your Trakt account will magically receive recommendation lists. They're under Profile > Lists in Trakt.
The lists will update regularly based on movies you rate in Trakt.
You choose the types of recommendations you want to receive.
How does it really work though?
Basically, it finds people with similarly weird tastes to you, and recommends films that they like more than average people (it's a collaborative filter).
Lists are ordered by confidence. The first film is the strongest recommendation.
It looks for "diamonds in the rough": lesser-known films you may have not heard of that you may like. It recommends these more than well-loved blockbusters that you probably know about.
This evolved from the board game recommender I wrote for Reddit users, which I've operated for free since 2014 from my home computer.
I've tried more "modern" AI approaches. They suck at recommending under-the-radar weird stuff, which is the whole point of this thing.
It doesn't use an off-the-shelf engine. It was written from scratch in Java. It uses 30 million film ratings to compile its recommendations for you.
How many lists can I create?
There's a maximum of 10 recommendation lists per new user, to keep costs low. If you politely DM me on Reddit, I'll uncap it for you.
What does "couchmoney" mean?
Finding great, under-the-radar films is like finding money in your couch.
Why does it take so long for the recommendations to show up?
Because great recommendations take time and a lot of CPU cycles. It may take several hours for your lists to first appear but they normally appear in a few minutes. Updates happen as quickly as the server can handle them.
Your lists get updated after you mark new movies as watched or rated in Trakt.
How good is it?
You tell me!
It's given me many gems that I love but never would have found. Your mileage may vary.
I don't have a Trakt account. Can I use this?
No. For now, you need a Trakt account with some rated movies in it. Setting up a Trakt account is fast and free.
As soon as you create a Trakt account and rate a few films, you'll be good to go.
Why can I only choose from 18 movie languages?
Those are the languages with enough data to make acceptable recommendations for, but I'm working on that. Even now, most of the languages other than English are lacking comprehensive data.
If you have access to any sources of film ratings from non-English-language speakers, please let me know!
You recommended a film I hate!
That's a statement, not a question. But yeah that can happen. Give the film a low rating in Trakt and the engine will learn from its terrible mistake.
The recommendations don't have a lot of foreign films I like.
That's also a statement. Maybe I should change this section to Frequently Stated Statements.
It will recommend non-English-language films, but the data sources are quite skewed to US & English-language tastes; I'm working on that.
Does the world need another source of film recommendations?
I built this during the COVID-19 crisis because I was fed up with the recommendations I was getting from my existing services. But if you have a recommender you like better, use that!
Will it really never recommend The Shawshank Redemption?
Correct. This differentiates it from other recommenders, as you can see from this chart:
Recommendation engines that recommend popular, non-controversial films are boring. Hopefully couchmoney finds you some unpopular, controversial films that you love.
Do I have to like The Shawshank Redemption to use couchmoney?
No, but if you don't, your unusual tastes are welcome here.
Why are Karl Malden and Marlon Brando in the title bar?
Because they are two of the greatest actors of the 20th century. Also because One-Eyed Jacks is in the public domain.
Is One-Eyed Jacks a good film?
Can I contact you?
Yes. But I might not reply.