Thoughts on “A Dark Room”

The first time I played through A Dark Room (browser version available here), there was a moment early on in the game—the village was still small, I had yet venture onto the dusty path, most everything was still mysterious—where I started to question my own virtue. Not my character’s virtue, but mine.

But I’m getting ahead of myself.

For those who haven’t played it: A Dark Room is an online/mobile text-based game, published by Doublespeak Games in 2013, which slowly reveals its true nature over the course of play.

You begin, naturally, in a dark room, with only one possible input: “light fire.” But from single point of interaction, the world quickly expands. Quoting from Michael Thomsen’s account of the game in The New Yorker (full article here):

After stoking the fire a few more times, you have a new option: collect wood, which can be used to build a cart. Once a cart is built, you can make traps and set them in the surrounding forest, and soon you’re collecting cloths and furs, which can be used to build more huts to attract others to your small enclave, allowing for the collection of even more fur and meat. You can begin to see a structure emerge from the fragments, but where that structure will lead you remains impossible to predict, and so the compulsion to keep pressing little word buttons becomes stronger.

It’s those “others” Thomsen alludes to here that I’m interested in, the villagers who take refuge in your growing community. The start of A Dark Room feels profoundly lonely. The only company you have is the builder, a stranger who stumbles into your now-lightened room who says that she can, well, build you things. Once she starts erecting huts, though, the town’s population starts to grow. A “stranger” here, a “weathered family” there: it all adds up. You as the player-character can then start assigning those villagers tasks, such as gathering wood or hunting.

When villagers began to appear in my first play-through, I was glad for the presence of extra people, even if those people were, in fact, nothing more than a number and a job description. When a wild beast attacked the village and killed several of them, I felt something resembling guilt. I, the de facto leader of this village, had failed to protect my neighbors, and now there was no evidence they ever existed.

This emotional connection did not last very long.

Around the time my village hit a population of twenty, when I had started assigning villagers to cure meat and tan leather (for reasons I was not yet clear on), another wild beast attacked. Rather than feeling guilty or sad this time, though, I was merely annoyed. The number of gatherers in the village plummeted, meaning it would take so much longer to collect enough wood to build a workshop (for reasons I was also not yet clear on). The only other option would be to re-assign the other workers to gathering, which of course meant a trade-off in resource gathering: more wood at the expense of meat, fur, etc.

Right then, in a brief flash of insight, I realized that I had stopped seeing the villagers as text-based representations of people, and had started seeing them as resources. They were merely means to my own still-unclear ends, sacrifices to some vague notion of “progress.” And then, as if that sudden doubt never occurred, I went back to pressing buttons, back to accumulating resources.

After all, there was so much of this world that had yet to unfold.

This is, I concede, not a grand revelation about the nature of player/non-player character relationships. Games consistently take an instrumentalist approach to NPCs. They are resource gatherers, quest givers, and of course, enemies. That the villagers have no lives and no function beyond their job descriptions is hardly a surprise.

What I do find surprising, though, is that the subtext of “NPCs have only instrumental value”—which is not even a subtext of the game, really, more a convention it happens to use—is brought to the level of text in the mobile version of A Dark Room.

The browser version of the game, as developed by Michael Townsend, does suggest that the player-character is a villain in the narrative of the game. They’re one of the so-called “wanderers” who conquered this world and have left it in ruins. You eventually find a spaceship in the wilderness, your ticket out of this hellscape. It’s been badly damaged, but it could be restored, and you find it fortunate that the “natives,” people like the villagers, haven’t figured out how to yet. One could read a colonialist narrative onto that story, but the game does not directly implicate the player (as opposed to the player-character) in that narrative. It’s just too oblique in its story-telling to do so.

It wasn’t until Amir Rajan adapted A Dark Room for iOS that the game’s critique of the player’s actions became overt. In the early goings—coincidentally, near the point when I had that flash of doubt—the builder begs you to stop overworking the villagers. When you keep pushing them to gather wood anyway, the game overtly relabels them. They are no longer “villagers.” They are “slaves.”

The player’s instrumental approach to the NPCs has consequences, which is certainly uncomfortable, as Rajan notes in an interview with Brian Riggsbee (full interview here):

The web version didn’t have any of the builder commentary or the slave transition…It’s funny actually, someone reached out to me on Twitter about the slaves transition and how “it wasn’t his choice.” He was pretty angry about it. His Twitter profile background was that of Fallout: New Vegas, where you can literally [be] part of a slave-driving army.

It’s all fun and games until the game points out that your progress has come at the expense of someone else’s autonomy.

To include the slaves transition was an editorial decision on Rajan’s part, an act of interpretation as well as adaptation. But does it pull the adaptation too far from the source material? Hardly. If anything, it simply reinforces the in-game narrative. Just as the wanderer uses the people they conquered to serve their own ends, the players use the NPCs to gradually satisfy their curiosity.

After all, how else can this story unfold?

Broadcasting Sports Injuries

The first few days of the 2016 Summer Olympics in Rio have been marred by serious, gruesome-looking injuries. On Saturday, August 6, French gymnast Samir Aït Saïd broke his leg while attempting a vault during the men’s qualification round. The following day, Dutch cyclist Annemiek van Vleuten crashed head-first into a curb during the women’s road race, resulting in several spinal fractures.

I watched both of these injuries happen live: Aït Saïd’s over an Internet stream, van Vleuten’s on the local NBC affiliate. After the initial shock and revulsion had passed, a question began to bite at me: what is the appropriate way to cover such injuries when they occur on a live broadcast? Given the inherent risk of physical injury in sports, and the inherent uncertainties of broadcasting events live, this is a question of some significance.

Broadcasters appear to face two conflicting imperatives. On the one hand, we expect that a sports broadcast will present an accurate (and often total) account of the event; on the other hand, we expect that broadcasters will maintain “good taste” — that they will not exploit the misery of participants or spectators (beyond, one might clarify, the misery that comes with defeat).

In most instances, these two imperatives do not come into conflict. Even in violent sports such as American football and ice hockey, the injuries players suffer do not generally inspire shock and revulsion. Concern and pity, perhaps, but nothing out of the ordinary. Thus, during most games, media outlets can broadcast the events in their entirety, injuries included, without ethical worries.

But sometimes, unexpectedly, a player with sustain a gruesome injury, either as a tragic part of normal play (e.g., Joe Thiesmann’s compound-fractured leg on Monday Night Football) or as a result of a violent infraction of the rules (e.g., Steve Moore’s broken neck at the hands of Todd Bertuzzi). Such injuries inspire immediate revulsion, disrupting the constructed narrative of the sporting event and, oftentimes, the event itself. In these scenarios, the imperatives of accuracy and good taste do come into conflict. How, then, should this conflict be resolved?

The Aït Saïd and van Vleuten incidents offer two different possible solutions. In the case of Aït Saïd, the broadcast strategy was to avoid showing the aftermath of the failed vault, once the extent of the fracture became clear. The camera quickly cut away from the scene, and they did not show replays of the failed vault. When the broadcast did return to the scene, it was always in the context of Aït Saïd receiving medical attention, framed in a manner which concealed the severity of his injury.

In the case of van Vleuten, the broadcast team was more willing to dwell on the fateful crash. This was in spite of the fact that the site of the crash, unlike the site of Aït’s Saïd’s, did not have a dedicated camera, indicating a degree of choice on the part of the broadcast. Not only did the broadcast show replays of the collision, but the commentators continued to discuss van Vleuten’s injury for the remainder of race. relaying information regarding her condition.

So which is preferable: drawing coverage away from the injury, or making the injury a feature of the sporting narrative? They are markedly different, but I will say that in their respective moments both approaches felt reasonable.

One must remember that the two injuries happened in different contexts. Aït Saïd’s broken leg did not impact the outcome or proceedings of the qualification round. He was a relatively marginal figure in the competition (though he would have qualified for the rings final), and there were five other apparatuses the broadcast team needed to cover. Even if the broadcast wanted to dwell on his injuries, there might have been too much going on to do so.

Contrast this with the road race: van Vleuten was nearing the final stretch, leading by a significant margin. Her crash completely altered the medal situation, giving four other cyclists behind her a chance at gold. Further, in covering the progress of the race, the broadcast team had to film the other cyclists passing the scene of the crash. Had van Vleuten’s injury happened earlier in the race, they’d have little cause to dwell on it.

This pragmatic variance does allow us to keep both the accuracy imperative and the good taste imperative. There might be a case for dropping one of the imperatives instead. Pain is pain, regardless of context, and adjusting the guidelines based on the competitive conditions of a game may well trivialize that pain. Yet a consistent response might well do the same, treating such human misery in a dispassionate manner.

In the messy world of live broadcast, perhaps pragmatism is the best that can be done.