It’s Sunday morning. I’m out of bed but slightly hungover and heading for the couch. I need something gentle to watch. My wife and I like to watch made-for-TV whodunnits and courtroom dramas on mornings like this, but with several streaming services to browse, it can take half an hour or more to unearth a good prospect.
We finally agree on one and settle down, steaming mugs of tea in hand, but it soon dawns on my wife that we’ve seen it. I’m not convinced and argue that we should give it another 10 minutes. Sometimes the reason you don’t remember a movie is that it’s a major turkey; sometimes it’s following such a predictable formula that you merely think you’ve seen it before. “She did it but dressed up as the guy,” my wife implores. Damn, she’s right. We have seen it. Back to the endless search.
Streaming services track everything I watch, so why can’t I filter out what I’ve seen? I’d love a way to mark things as watched that I may have seen years ago or on another service too. While we’re at it, why not let me filter out musicals, The Big Bang Theory, and anything with James Corden? Most streaming services have been slow to add these quality-of-life updates, but it can save precious time on those rare worry-free weekends.
We undoubtedly watch too much TV, especially over the nearly two years of intermittent lockdowns. But I know there are unfamiliar gems on these streaming services; it’s just tough to get at them. The likes of Netflix, Amazon Prime Video, and Disney+ all offer up row upon row of movies and shows divided into categories and genres, but I increasingly get the feeling that each row has the same titles, just jumbled into a different order (looking at you, Prime Video).
Deciding what to recommend has been a problem for streaming services from day one. There are objective rows, such as the top 10 most popular in your country or what’s trending, but how do these platforms decide what you might want to watch next? Netflix has a thumbs up or thumbs down system, but it’s not entirely clear what it does. So I asked.
If you rate anything thumbs up or down, Netflix assumes you have watched it on the service or elsewhere, a Netflix spokesperson told me. A thumbs up should result in related content suggestions, while a thumbs down delivers fewer similar shows or movies. So far, so simple.
Related content depends on a group of people at Netflix who tag all the shows and movies on the platform. These people attach descriptors like “horror, mystery, understated, ominous, and scary” to releases like Midnight Mass, for example. That way, the system can cross-reference tags to suggest similar shows and movies. Or not, if you didn’t like the show.
Take a look at someone else’s profile, and you will likely see a huge difference in the types of shows and movies Netflix recommends. Unfortunately, just because you rate something, it doesn’t necessarily disappear, which can make the rating system feel like it isn’t doing much.