This weekend, one of my favorite shows on Netflix bit the dust. Victor Fresco’s gleefully bizarre “Santa Clarita Diet,” about an undead realtor and her family murdering bad folks to keep her fed, was one of the strangest and most stylistically unique shows around. It was consistently hilarious in the way that audiences often appreciate, yet fail to properly value in a show until it’s ostensibly gone (see also: the fan outcry when the under-the-radar Brooklyn 9-9 was cancelled).

The cancellation was especially shocking, not just because the show ended on a major cliffhanger, but because nobody really knew it was in danger. We couldn’t, because Netflix famously does not share viewership numbers with anybody, even the people who actually create the shows. It’s easy to see why Netflix keeps this information close to the vest; if nobody knows how popular their shows actually are, then they can attract subscribers based on the perception that they have great, popular content. They don’t have to worry about failure narratives when their shows’ numbers start to dip the way that a series like The Walking Dead does.

For its competitors, or even for writers covering television, Netflix’s secrecy is a constant annoyance. It’s hard to track how you’re stacking up against a competitor when there are no metrics to measure by, and it’s hard for writers to know what shows are actually popular beyond general buzz.

Beyond that, Netflix’s secrecy is proving to be an obstacle in keeping shows on the air. With the rise of online communities, devoted fans have been able to organize and demonstrate, en masse, their support for a series they love that’s “on the bubble.” They’ll send boxes of fake bananas to save Arrested Development, or buy loads of Subway sandwiches to save Chuck. People have ways to act out when something they love is in danger. But they have to know that it’s actually in danger first.

Netflix may be able to look at numbers and see how many people are viewing a particular show, but they have no way of actually gauging the passion behind people’s viewing habits. The Big Bang Theory and Law & Order: SVU may be ratings juggernauts, but you just don’t hear people talk about them the way they do with Game of Thrones, or even something like The Good Place. It’s unlikely that CBS cares one way or another, since they’re entirely ad-supported and ratings are their everything. But given that Netflix is a subscription model, you’d think that the people in charge would care about this.

But the people aren’t really in charge at Netflix. The algorithm is.

This isn’t news; Netflix has been making its biggest decisions using their algorithm since they ventured into original content. House of Cards was famously greenlit because they saw an overlap in their viewership of the British original, David Fincher movies, Kevin Spacey movies, and other political thrillers. This reliance on data didn’t just drive their series pick-ups, either; Netflix will actively tell showrunners how to develop their series based on their algorithmic conclusions, as Maniac creator and True Detective season 1 director Cary Fukunaga shared.

“Because Netflix is a data company, they know exactly how their viewers watch things,” he said in an interview with GQ. “So they can look at something you’re writing and say, We know based on our data that if you do this, we will lose this many viewers. So it’s a different kind of note-giving. It’s not like, Let’s discuss this and maybe I’m gonna win. The algorithm’s argument is gonna win at the end of the day. So the question is do we want to make a creative decision at the risk of losing people.”

The same rationale goes for cancelling series as it does for picking them up and creatively guiding them. Netflix is taking all of its data regarding cost of production and series backlog and weighing it against viewership numbers. It’s crunching the numbers every year, and the math changes substantially as a show goes on. According to Deadline, “Netflix’s deals include bump/bonuses after each season that are getting progressively bigger. While the payments are relatively modest after Season 1 and a little bigger after Season 2, I hear they escalate after Season 3, especially for series owned by Netflix — sometimes from hundreds of thousands to millions of dollars — as the studio starts to pay off the shows’ back-end. For series from outside studios, which I hear cost about 20% more than their Netflix-produced counterparts, I hear the built-in payment increases do not skyrocket as much but still are bigger after Season 3, Season 4 and beyond.”

This leads to a culture where series survival is purely a numbers game, and making it to a 4th or 5th series is unlikely for all but the most successful series. There is no human factor in determining a show’s fate, just the almighty algorithm. Victor Fresco’s comments to The Hollywood Reporter at the start of season 3 back this up: “We’re aware that the show gets more expensive every year; we’re aware of what seems like templates of [Netflix’s] studio stuff now. It looks like mostly three-season stuff. We know they like the show. I like to say the humans there seem to love the show; I don’t know how the algorithm feels about us, and the algorithm makes a lot of decisions, ultimately.”

It’s somewhat ironic, then, that a subscription service that ostensibly benefits from not being beholden to ratings and numbers is now even more strictly controlled by their internal viewership statistics. It’s also, in my personal opinion, a very bad business decision. Netflix has become huge largely because they had no direct competition. They managed to obtain a huge library of movies and TV series before they had to compete with other platforms for licenses, and they used their competitive lead to create enough original content to keep the original subscribers on their platform, even amidst price increases and a dwindling library. But they’re about to see the rise of numerous competitors, chief among them Disney+.

Disney+ may not have as large of a library straight out of the gate, but they’re cheaper (only $7.99/mo), and they have content that people genuinely give a shit about. Not only will they have nearly every Disney movie available, they’ll also have all the Marvel movies AND all of the Star Wars movies. They’ll also be home to a whole slew of original series based in these worlds, which audiences are already deeply invested in. Disney has also proven themselves extremely adept at understanding their audiences and giving them stories that they continue to love. You can’t release 22 Marvel movies in 11 years based on the recommendations of an algorithm; you need a human understanding of storytelling and catharsis.

Of all the series Netflix has, the only currently-running household name is Stranger Things. Don’t get me wrong, I really enjoy that show. But the competition is fierce, and Netflix isn’t doing itself any favors by cutting the heads off of their best shows right at their midpoint. It’s the sort of decision-making that makes perfect sense to a machine, which views stories as content and audiences as consumers. But it completely undervalues the emotional attachment to stories, something their competition knows all too well.