‘Black Mirror: Bandersnatch’ has reignited the debate about interactive TV. Is this something viewers genuinely want? Or are branching storylines simply a gimmick? Make your choice now…
Released on 28 December 2018 by Netflix worldwide, ‘Black Mirror: Bandersnatch’ is a standalone episode of the Emmy Award winning science fiction anthology series ‘Black Mirror’. Created by Charlie Brooker and Annabel Jones, essentially Black Mirror is a British Twilight Zone for the 21st century; one that examines the often unintended and frequently dystopian consequences of new technologies in either near future or alternate timelines.
It’s smart, clever TV with a pronounced satirical bite. Now located on Netflix, the team decided to take it a step further and utilise the OTT platform’s technology to produce an interactive show where the viewer controlled the action. ‘Bandersnatch’ was the result, a show with a viewing time that can be anywhere between 40 minutes and over two hours (the average is 90 minutes) depending on what actions the viewer takes when presented with a series of A/B choices. The show has, depending on who you talk to, between 5 and 12 endings buried amongst its 250 different scenes. There could well be more.
The below flowchart illustrates the journey the viewer has to take to reach certain endings (we’ve kept it deliberately small and even out of date to avoid spoilers: the original is here).
Interactive TV Future, Videogame Past
If all this feels very much like a videogame, then that’s because branching video content was pioneered in the 1980s when the first arcade cabinets powered by Laserdisc came out. This was the first delivery mechanism that allowed branching paths to be presented and then executed with a lag that was deemed acceptable to the viewer.
1983’s pioneering Dragon’s Lair has since been immortalised in ‘Stranger Things 2’. In it, the player guided a Hollywood-quality animated character, Dirk The Daring, through a series of obstacles to rescue a Princess; jumping one way to avoid flames, another to avoid rocks, drawing a sword to kill a monster etc. etc. Get it wrong and it was game over. It was fiendishly difficult to get the timings right and memorise the path, it was twice the price of any other arcade game, and it was also ridiculously popular.
“I hate this overpriced bullshit!” yells Dustin (Gaten Matarazzo) at the screen in ‘Stranger Things 2’ shortly after watching Dirk die yet again (above). Anybody that played the game in the 1980s will sympathise.
And anybody looking at producing interactive branching content in the here and now might well sympathise too, as there are numerous difficulties that occur across the length of the process from script to delivery.
These include, but are no means limited to:
The usual rule of thumb in television production is that one page of script equals one minute of content. Bandersnatch’s script ended up being 170 pages long. Even generating that was troublesome, with Brooker first writing it as a flowchart before learning about Twine, an open-source tool specially developed for telling interactive non-linear stories. This was then superseded in turn by a Netflix-developed tool called Branch Manager, which at one point crashed, taking the entire story outline with it.
It’s a good name though, as it emphasises that not only is the viewer following a branching pathway but also that content grows out horizontally, just like a tree. As scenes get added, interactive work doesn’t just get longer, it gets broader as well.
Netflix’s first experiments in interactive content were in children’s animation. The Dreamworks animated ‘Puss in Book; Trapped in an Epic Tale’ debuted in June 2017 and was followed by a couple of other children’s shows. Animation keeps it simple and makes the task of continuity much simpler, whereas the complexity of a live action interactive shoot introduces a whole new level of difficulty.
This is partly due to the nature of a branching narrative. When done properly, not only does the viewer have to choose between, say, paths C and D, but the decision they made when they chose A or B at the start of the process will have implications too. At its most logical conclusion, this means that scenes Ca, Cb, Da, and Db would all have to be filmed.
Practical considerations — time and sanity being two of them — mean that not every branch can have its ‘breadcrumb’ trail of previous decisions; this only occurs when it makes narrative sense. But even so this has implications. On set, for instance, as Director David Slade told Wired, the actors found it difficult to just switch from one branch to another and they had to laboriously shoot each scene in linear order. “With each scene we had to complete it before we went back and did it again [in a new version]. Because the reactions and where you came from could be different.”
Netflix doesn’t often talk figures for individual programs (though it has revealed some interesting data about the choices viewers make while watching Bandersnatch, see below). However, it’s not hard to guess that everything about producing interactive drama is going to take longer and therefore cost substantially more.
So, where’s the ROI? Well, to begin with it’s in publicity. Taking a geek-friendly format and adding it to a geek-friendly program as a one-off has been a masterclass in marketing for the company. While the animated shows were treated as an interesting curiosity in the industry, ‘Black Mirror: Bandersnatch’ has been lauded variously as an expensive gimmick, the future of television, and everything in-between. What’s more it has provided that rarest of things: a genuine water-cooler television moment in the on-demand streaming era.
There is another play at work here too, of course, which is repeat viewings. The sheer amount of permutations available make it unlikely that a viewer will reach all the endings in one sitting, and there is plenty of content dotted inside the depths of the show’s labyrinthine structure that is difficult to get to. Some of the audience will watch it just once and leave it at that. The hope is that a healthy percentage of viewers will be back for more and try out the different pathways.
To get interactive content to the audience you obviously need to be running an interactive platform. OTT is perfect for this, but even then Netflix has had to limit the distribution of the Bandersnatch episode. Download, for example, is out; the file sizes when you include all the possible branches are simply too big.
Nor can it be watched on older versions of the Netflix app or via Chromecast or Apple TV. This is basically an anti-buffering measure, as the latest versions of the Netflix app allow for the pre-caching of both of the upcoming two possible paths, something not possible on Chromecast or Apple TV. This is how it keeps the action seamless (though, as an aside, it also explains why the show is only available in HD — two 4K streams would be difficult to cache).
Interactive TV: Choose Your Path
Despite Bandersnatch’s success there is much still to learn about Interactive TV. During the program’s evolution there was much discussion about getting the right frequency of choices to keep the viewer engaged. Should they be allowed to cluster together or drift wider apart? What is the appropriate space for cinematic story-telling versus what counts as too much hassle; the old ‘lean back/lean forward’ argument, but this time occurring in the same single piece of media. In the end it was decided to place a decision roughly every three to five minutes.
What will be interesting is the next one. Brooker has talked about Bandersnatch as being a process of learning the ropes and that he’s having all sorts of ideas for things that he could do now he knows the process. Then, of course, there’s the data itself. Every viewer who has watched the show and pressed a button, even the ones who haven’t, have contributed to what might well be the largest dataset ever compiled for a single piece of television.
So far Netflix has been coy about most of the paths viewers have taken. But it has revealed that, for example, 73% of viewers make one key decision to accept a job; 60% choose Frosties over Sugar Puffs in the first decision to be made; and, quixotically, that while 55.9% of all viewers will throw a cup of tea over a person, that number falls to 52.9% in Britain. After all, tea is tea…
Anyone who has been following the progress of increasingly sophisticated data analytics in the television industry will know what a potential goldmine of information that data is, both in terms of consumer behaviour and in the way they are reacting to the narrative as it progresses. It’s easy to see how it will affect feedback into any future experiments with interactive TV.
Of course, there is a very real argument that says that any television show that involves the audience is interactive, especially when it comes to matters of a public vote. There is a long history of experiments in interaction both at the cinema — Interfilm and Sony trialled a system in the early 1990s which saw the audience vote on plot directions — and via entertainment consoles. To be honest, beyond the notoriety of games such as Night Trap, few of those titles ever really set the video gaming world on fire.
Maybe there is even a hybrid future in which OTT audiences can vote on upcoming branches, further boosting audience engagement for a certain genre of programming. And, with a steady increase in the realtime rendering power of Graphics Processor Units, you could even speculate on sophisticated animation software being able to render CG scenes as the branches occur in realtime.
But for now, 'Black Mirror: Bandersnatch’ represents a watershed in Interactive TV. Will it be one that leads to (A) further Interactive TV shows or (B) is it a one-off whose success will never be repeated?
You are an OTT TV exec. Choose (A) or (B) now…
Black Mirror: Bandersnatch & Stranger Things pictures: Netflix