Inscrutability in Game Design: Perils and Possibilites


Ian here—

Earlier this week, I sat down with some friends to play David OReilly’s Everything (2017). We were all suitably hyped, and ready for an evening of being different things while chuckling over the political indefensibility of object-oriented ontology. (Yes, really. We’re those people.)

But our evening hit a snag. Somewhere along the line, in our first playthrough, beginning as a cougar, we encountered a weird bug. One of the little “gain a new ability” thought bubbles failed to pop up in the environment. No matter how far we traveled, or how many objects’ thoughts we listened to, we simply could not unlock the ability to scale up or down into bigger or smaller entities. We were doomed to be a cougar. (Or, cougars, rather. We were able to expand our ranks, and become about a dozen cougars, rolling across the rocky landscape. But being a dozen rolling cougars is small consolation in a game that promises that you can be, well, everything.) For 75 minutes, we wailed at the screen. “I thought you could be lint particles and galaxies and stuff!”

We tried, to no avail, to look up walkthroughs for Everything. Everywhere we looked, we found glowing reviews of how chill the game was, how it allowed you to sit back and let the experience of being everything wash over you. We couldn’t fathom how we got stuck in a game with no real objectives. To pour salt in the wound, the game seemed to be actively mocking us. “As long as you keep moving, in any direction you choose,” thought a tree, “that will take you where you need to go.” “Over time,” thought a rock, “you might find there’s no right or wrong path to take here.” We were seriously stuck, and the only feedback we were getting were Zen aphorisms that told us to take it easy and not worry so much.

Eventually, we deleted our save file and started over. Thankfully, this solved the problem, and we spent the rest of the evening enraptured. But it did get me thinking about the trickiness presented by purposefully inscrutable game design.


I study avant-garde cinema. I write criticism of it, and I teach it in my classes. I’m no stranger to moving images that are, on the surface, inscrutable. Usually, I like things like this. I like to be flummoxed. I like to be thrown into a series of images, not knowing the rules, and having to either claw my way out through careful observation and hypothesizing, or simply let the mystery wash over me.

But I have a confession to make: I don’t particularly the feeling of being confused while playing a game. This is, for me, a quandary. To a certain extent, I consider it to be a kind of moral failing. Yes, I made it all the way through Liz Ryerson’s Problem Attic (2013), but to be perfectly honest I kind of hated the experience. I couldn’t figure things out, I couldn’t find good resources online to help me figure things out, and I ended up making a guide just so that I could prevent future adventurous players from having an experience quite so unpleasant as mine.

I used this GIF in a post before, but I don’t care: it captures the confusion of playing Problem Attic so well.

And that’s the best-case scenario. Even though I didn’t like it, I at least respected Problem Attic enough to finish it. (In fact, I respected it enough to assign it to my students.) But there’s stuff like Droqen’s Starseed Pilgrim (2013), which I couldn’t even stomach enough to play more than two hours of.

I’m not proud of not sticking with Starseed Pilgrim. Again, I can’t shake the feeling that it’s some sort of moral failing on my part. I know that a lot of people have enormous amounts of respect for the game, and fervently evangelize for it. And, given my tastes in cinema, one would think that, as a matter of principle, I respect the right of artists to be opaque, to refuse to explain themselves, to expect some genuine meaning-making effort on the part of their audience.

But unlike, say, Joana Pimenta’s admittedly vexing An Avation Field (2016), I get this weird feeling that Starseed Pilgrim actively hates me. Its design is so hostile that it bends into a weird sort of anti-pedagogy, as if its goal were to make confused players feel like stupid, filthy casuals for not understanding its true purpose. (Perhaps I take the game’s design philosophy personally since, like James Paul Gee, I’m a teacher who thinks games actually present interesting case studies in pedagogy.)

This all feels tremendously irrational to me. But I do think there are specific things about games, as a medium, that short-circuit the pleasures of inscrutability that someone like me can otherwise enjoy in other art forms. Two things, in particular:

  1. Games actually ask you to do stuff. Unlike an experimental film or non-interactive piece of video art, you can’t sit back, be confused, and just let things wash over you. You need to actually, on some level, understand the mechanics in order to progress onward, and get the full aesthetic experience. Obscure mechanics, then, present a more serious impediment to an aesthetic appreciation of games than other types of obscurity do in other media.
  2. Games can be functionally broken, just on the level of code, in ways that are easy to confuse for deliberate obfuscation. The “is this supposed to be hard to figure out, or have I actually just hit a bug?” problem is really something unique to games.

This first point is something that my mentor and colleague Patrick Jagoda is currently writing on. Patrick makes a distinction between “mechanical difficulty” and “interpretive difficulty,” pointing out the ways games can layer the two on top of each other. A Joyce novel or a Hollis Frampton film might both be difficult to interpret. Something like Problem Attic, meanwhile, is not only difficult to interpret on a thematic level: it is also difficult to figure out what you’re supposed to do, and what the rules of its spaces are.

The second point, though, is worth considering, as well. In 2008, I attended a Phil Solomon retrospective at the Gene Siskel Film Center in Chicago. We were watching Psalm III: Night of the Meek (2003, then still known as Twilight Psalm III), watching figures boil and buckle out of the liquified emulsion, coming into being and then dissolving away again. And, all of a sudden, the images began to melt a little bit too much. For a split second, we could all almost believe that Solomon had perfected a new form of melting imagery, but then it quickly became clear that the filmstrip itself was, in fact, melting. And then it was burning. And then the projector shut off. Solomon came to the podium—apologetic, apocalyptic, and somewhat bemused. “Well, that’s the end of that print,” he said, and then, with something between a winsome smile and mournful sigh, he added, “… the End of Cinema.” If one was going to ever experience a technical failure while viewing a work of art, this was the way to experience it: clearly explained as such, but still weirdly resonant with the themes of the work itself.

The burning print of Night of the Meek was deeply tragic and oddly fitting. By contrast, the bug my friends and I encountered while playing Everything was nothing but frustrating. We weren’t transfixed. We were infuriated. Everything is hardly the most inscrutable game out there (it’s not Starseed Pilgrim, let’s put it that way), but it does pride itself on giving players little in the way of goals or direction. It was, in other words, just open-ended enough for everyone in the room to become deeply confused as to what, exactly, we had missed: to think that our confusion was somehow our fault, or OReilly’s fault, rather than the fault of some accidental bug. And this was not the intriguing sort of confusion. It was just the frustrating sort of confusion.

* * *

Wow, 1,300 words in, and it feels like all I have been doing is whining. One of the reasons I wanted to write this post, though, was that, in addition to playing Everything, I have also been playing Stephen L. Clark’s Rooftop Cop (2014). And I think it’s really something special. It’s something that clears a personal hurdle for me, being inscrutable without being frustrating. And I think it has emerged as one of the games that, in the future, I would be most likely to recommend to people who like video art but aren’t sold on the whole vidoegames thing.

Rooftop Cop consists of five levels, which the game bills as “ambient vignettes videogames.” In the first vignette, “A Proud History,” the player takes control of a cop walking a beat, while other cops look on from rooftops. The player has several options: they can “cite” citizens by first clicking on them and then physically apprehending them. This nets the police department a certain amount of money, based on how many times the player clicked on the citizen before grabbing them. They can also “watch” citizens, and construct a “fence.” This last mechanic utterly perplexed me. Building fences didn’t seem to do anything at all. It didn’t constrict the movement of pedestrians in any way I could observe.


At the end of a set amount of time, the police department collects whatever money that has been levied in fines from citations. But there are also an enormous amount of expenses deducted, more than could ever be reasonably recouped in fines, so the department operates in a continually deepening deficit. The game is, according to Clark, structured as a “loose metaphysical timeline in which the Cops slowly lose their way.” So I thought, okay: this first level is showing that street-level policing is expensive, and that economic forces have pulled cops off of beats, sowing distrust, pulling them away from the public they are employed to serve, and establishing a militarized police presence in the place of responsible community engagement. But there were so many ragged ends. What did “watching” actually do? Why did the “fence” mechanic seem to be utterly broken? What wasn’t I getting? “I wish this was more like Papers, Please,” I grumbled to myself, “so that I could enjoying airtight mechanics that perfectly express what it is like to operate under a repressive regime.”

I pressed escape to exit the first vignette, and started up the second, “Capture the Flag for One.” As the name implies, all there is to this one is for your character to run back and forth on a field, carrying a flag back to your home base without competition.

Except … you don’t. After one successful attempt, you find that the ground is getting soggy. It’s sinking, like a slowly and sadly collapsing bubblebath. Gravity is pulling you down, and the sides of the playing field are now too steep for your to ascend. You’re stuck in the mire, helpless to escape this slowly sinking fate.


I had no idea what this was a metaphor for. But I did know that I loved it. Even as I was contributing input on my keyboard, I was able to appreciate what I was seeing on my screen as a piece of animation. Color, composition, and movement were coming together to weave an overall mood, a mood that was more central than any message.

From here on in, I decided to embrace confusion. I wouldn’t let Rooftop Cop frustrate me. I would, instead, let it wash over me, trying to gain as much meaning from its animated vignettes as I could, while at the same time acknowledging that I wouldn’t grasp everything. I decided to approach it as I would a Yuriy Norshteyn film, or a Janie Geiser film.

And it worked on this level! The fourth vignette, “God Bless Everyone,” put me in charge of two figures on a sort of floating island. They could feed fuel to their fire, and use debris to expand the size of their island. If the island collided with the buildings rising up from the inky blackness around them, I could have them climb onto them, grabbing more debris or even pick up more passengers.


I had no idea what was going on here, but I didn’t care. The vignette invoked a deep feeling, rather than a concrete thought, and for that I was appreciative. I have to say, I liked it better than This War of Mine (11bit Studios, 2014), with which it shares some broad mechanical similarities. In being far less prescriptive, it actually opened up more room for me to feel things.

What does Rooftop Cop get right? Why did I stick with it, working through my initial frustration? I see several factors at play here:

  • In describing its sections as “endless ambient vignettes,” Rooftop Cop makes it clear that whatever goals these vignettes have will be set by the players themselves. This removes some of the hurried confusion that games such as Starseed Pilgrim can engender, and opens up healthier avenues for mystery and exploration.
  • Unlike Problem Attic, there is no gating of any content. If you’re not feeling a particular vignette, you can always hit escape and pull up the next one. It resembles a a book of poetry, where you can skip around, more than it does a traditional game in which experiencing the entire work is contingent upon you successfully overcoming challenges.
  • It spaces out the vignettes that really do seem to be genuine challenges, with a typical game-like structure (the first and third vignette, in particular) with vignettes which are clearly better thought of as digital animation, meant to be visually appreciated (the second vignette).

It’s possible that someone else will have a completely different experience, of course, but these things together allow Rooftop Cop to circumvent the usual problems with inscrutability for me. I am intrigued by it in ways that I am normally intrigued by experimental cinema, rather than driven to ragequitting.

So, that’s Rooftop Cop. I might post a few more thoughts on Everything down the line, now that I’ve finally gotten it to work.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s