The first season of Stranger Things took me by surprise. Literally. My friend introduced me to it and, seven hours later, I was still sitting on the couch. The second season of Strangers Things failed to captivate me in the same way.
But one episode was particularly egregious: the Stranger Things Chicago episode. It’s bad. Really bad.
But it’s bad for a reason: Netflix is starting to experiment with the data about how people watch TV, not just what they watch. The Chicago episode, the seasonal setting, and the disorienting tonal shifts of the season are prime examples of how new TV shows are being created.
The Chicago Episode was a Pilot for a Side Series
The Chicago episode was testing how audiences would react to a spin-off series.
Even show creator Matt Duffer admitted as much, in maybe not the exact words: “It’s almost like doing a whole little other pilot episode in the middle of your season, which is kind of a crazy thing to do,” he said.
Netflix probably wanted to test out the audience’s appetite for a Hunger Games-esque, female power duo sister thing. While it wasn’t my thing, I’m sure the Eleven fans loved it. This was a bold attempt at running a pilot on the best possible test audience.
The soundtrack, the characters, even the dialogue in the Chicago episode was simplified, watered down to Young Adult Touch. We got to see the equivalent of an 80s training montage as Eleven gets more powerful. We got to think about how this quest of killing all the men who hurt people could be a long, barely watchable series.
And then, as fast as it happened, the Chicago episode is finished.
Blog posts (ahem) criticize it. Magazines criticize it. On Twitter, people yell about it.
The pilot aired and people didn’t like it. What does Netflix care? They won’t pursue it.
The Seasonal Setting is Built for Binging.
As far as our culture goes, I’m not yet in the throes of a Netflix addiction. I don’t even have a subscription. So maybe this has already happened… but I was amazed at the forethought of setting the series during Halloween. Then releasing it on Halloween. And ending it right as Christmas gets started.
This is undoubtedly based on the research of how people are going to watch the show. Will they watch the whole thing on Halloween itself? Good, got it covered.
Will a few episodes – especially after the head trauma of the Chicago episode – get watched closer toward December? Perfect, you’ll wrap up Stranger Things 2 just as the holiday season gets going.
In this way, Netflix is far, far ahead of the innovation and stagnation of book publishers. This data-driven setting that was created and plotted with actual viewing habits in mind. Stranger Thing 2 is a seasonal experience. When the kids dressed up for Halloween, there were kids in costumes on my own street.
Where else can you take this kind of behavior-based media?
I wouldn’t be surprised if we start seeing show-themed products in the show that are timed for the show launch. Or a show that airs during another major event – Presidential Election, New Year’s Eve, whatever – that again reflects the seasonality.
Movies have been themed for holidays for a long time. Now that Netflix knows people will watch entire series in a single month, show creators can get even more granular in terms of timing and setting.
The Bizarre Tonal Shifts are Data-Driven.
The Chicago episode was clearly built from data mined across different young adult sensations. There was a dash of the Hunger Games, the sister-love of Frozen, and the heavy training of Divergent.
Every other episode seemed to lean hard into data-driven TV writing, too. And, sometimes, especially later in the season, it didn’t work out.
Was Stranger Things 2 a thriller? A comedy? A horror? Was it meant for kids? Adults? The whole family?
In one scene, characters seemed to be acting almost in separate genres. The dialogue went from predictable to cringingly bad. The chemistry that made the boys so enjoyable to watch in the first season vanished, replaced by ham-fisted attempts at pasting incongruent personalities onto each one of them.
You know it’s bad when it’s not the aliens that are the true antagonist, but a bully whose cruelty is irredeemable, one-dimensional, and boring. Oh, and after said bully beats another character to a bloody pulp?
Well, that character becomes a slapstick idiot for the rest of the season. So no one cares.
Basically, the later episodes, guided by trying to appeal to so many niches at one time, became something that almost seemed to have been written by a robot. An algorithm.
Live by the Algorithm, Die by the Algorithm
I’m very interested in how Netflix continues to experiment with data-driven creation.
Take a show like Godless. Westerns are popular. But how can the show stand out? How can it end up in different genres and appeal to different viewers? Well, market it as a “feminist Western.” I haven’t seen the show, but the concept alone seems to be almost created for cross-genre pollinations.
Or how about Bright, the upcoming movie starring Will Smith, a hardened police officer and his sidekick… an orc? This is a smashing of fantasy and buddy cop comedy that smacks of the same strategy.
It seems that Netflix original series are being built not for audiences first, but algorithms.
This is the equivalent of me writing “Stranger things Chicago episode” over and over so anyone Googling for self-affirming anger over the Stranger Things Chicago episode ends up on my blog post, first and foremost.
Netflix is reinventing the genre. There are dozens, if not hundreds, of secret genres – from “Cerebral Scandinavian Movies” to “Critically Acclaimed Animal Tales” and “Goofy Courtroom Movies.”
What Netflix is doing now is finding ideas that cross genres and, therefore, hit markets in new ways.
Creating for the Machine
As we favor algorithms that help us discover our next favorite musician or movie, it begs the question of how to get discovered as an artist.
JK Rowling wrote The Cuckoos’ Nest under a pseudonym and sold 1,500 copies in four months. Harry Potter & The Deathly Hallows sold 8.3 million copies in one day.
Alec Hutson, self-published author of the The Crimson Queen, even attributed Amazon’s algorithms to his book’s success. I suspect something like “The Werewolf Love Story Box Set” is created with user searches in mind, too.
And as Amazon gets more sophisticated, maybe the engine crawls the book itself to continue to personalize search results. Are these werewolves feral or mature gentlemen? Is the setting 1970 or 1850 or 2018?
Let’s not even get started on Spotify and Pandora.
People are discovering their new favorite TV shows, movies, books, and music not from human recommendations but algorithms. Don’t forget it.
For artists everywhere, that means you need to be even more careful about how you title your work, describe it, and, ultimately… what you create in the first place.