Matthew Schwager is a media designer and essayist. He holds an M.A. in Experimental Digital Media from the University of Waterloo, where he researched religious hermeneutic as applied to user experience. This fall he will be an M.F.A. candidate in graphic design at the Yale School of Art.

Shelter is a game about figuring out what the hell to do next. You play as a mother badger trying to guide her children to a new den. Gameplay consists of roving across predatory landscapes, securing food in the process, and feeding this food to your kids. This sounds fun. For me, it was not.

The actual process of playing isn’t so simple. I wrote down questions while I played: Why is one of my badger babies grey? Is it a special baby, or is the greyness a simulation of natural variation? Is it not being fed enough, and is that why it is grey? What is that guitar noise? As I play, I wander around a lot and a guitar strums like a curtain is being pulled back to reveal something. The guitar annoys me. I feel like it’s telling me that I’ve discovered something new, but the environment is so cluttered I can’t really see anything new at all.

Shelter’s design elements seem to have a disconcerting lack of unity, one that hampers its ability to function as a game. I consider it to be a broken game, a deliberately uncharitable conceit. I use Shelter as an indicator of what happens when a game does not deliver a stable or reliable experience—either narratively or generatively—that we can unequivocally rely upon to construct theory or make generalizations. This line of inquiry approaches games more from a designer’s perspective than from a theoretician’s or a critic’s, and the result is predicated more upon the inspired mess-making and unpredictable solutions of designing rather than the generally rigorous attitude that informs research. Rather than formulating what a game is, which can carry with it a sort of prescriptive argument about genre or mode, a designerly methodology gives us license to consider what a specific media artifact does, allowing us to critique without condemning.

Here, the (affectionately nicknamed) “broken game” consists of moments of confused level design and/or goal-setting, and, like a misprinted book with text that runs off the page, the broken game challenges the broad applicability of our methodologies and critical typologies. We might find that accepting the broken game as a conceit broadens our definitions in new directions, guiding us to idiosyncratic and indulgent ways of adjusting and applying theory. This critical openness might satisfy needs we didn’t know we had or create new ones we don’t know how to satisfy. Identifying and subverting needs in this playful way, to me, is a process essential to design. In this essay, I use my inability to wayfind throughout Shelter as synecdoche for reading games and playing theory.

Wayfinding and Flow

The most basic ideal of wayfinding is physical and consists of the locating of one’s self upon a map or sequence of mechanistic objectives. An indicator of clear wayfinding could be Csikszentmihalyi’s notion of “flow”. Flow evaluates an experience’s makeup by considering the challenges presented and one’s ability to complete those challenges. When one’s skill level is low and the challenge difficulty is high, one feels anxiety; when the reverse is true, one feels bored (Csikszentmihalyi 74). Playing tennis with a busted racket against Serena Williams is an anxious experience, and playing tennis against a toddler is boring. When skill and challenge are appropriately balanced, neither one overwhelming the other, a strange, productive state of mind sets in, a “flow” that keeps a player simultaneously capable and vulnerable.

I found that Shelter immediately hands you a busted tennis racket: the visuals. The game is rendered almost exclusively in low-contrast pastel tones, and, starting off in an underground warren, I had trouble finding my way around. This was not due to cranky control schemes but to an inability to wayfind. My badger was brown, my babies were brown, the ground was brown, food from the ground was kind of turnip-colored, the den corridors were tight, winding, without very many characteristics, and brown—these weren’t problems that could be addressed with a contrast knob on my end, but were problems with narrative space, with a cinematic and silent “opening” that I really needed subtitles for.

This could conveniently be coded as affordances, or the lack of affordances. Norman defines an “affordance” as the “perceived and actual properties of [things], primarily those fundamental properties that determine just how the thing could possibly be used”. These “clues to the operation of things” delineate relationships between the actor and the world; door handles indicate access points, stop signs establish rhythm in the flow of traffic, and curbs on a road indicate directionality (9). Here, there were no perceptible affordances available; I found myself turned around too many times to get out of a very simple part of the game.

This sort of non-challenge—being told to follow an affordance-less looped path it until it ends—can help deepen our understanding of play-states. Csikszentmihalyi’s “flow” is often conceptualized as opposite ends of a spectrum; they are non-overlapping experiences due to the inverse relationship of the variables (“skill” and “challenge”) involved. But, I found that Shelter made anxiety and boredom set in simultaneously—I did not have the skills to intuit where I was supposed to go and what I was supposed to accomplish, and so anxiety was running fiercely in my blood. However, once I scaled my attention up from the diegetic level of the game, I found myself, strangely enough, bored at the exact same time, for the simple reason that I knew a higher level of skill would not have altered anything about my experience. A confusing path is a confusing path. The opposite, seemingly irreconcilable ends of the flow graph coincide, and not due to easily changeable variables that are part of the traditional dynamic of flow. It’s important to note that I did find one period of excellent flow during the game when hiding from a bird of prey and darting from one section of grass to another. I found myself dropping all my cynical detachment to fully apply my attention to surviving this particular landscape, because I was offered clear risks and clear ways to avoid said risks. However, this state of flow was short-lived, an oddity in Shelter’s universe. The initial struggle in the home den was the game short-circuiting, folding Csikszentmihalyi’s chart over itself, adding an unhelpful third dimension to his formulation, and in a way making it harder to wayfind on Csikszentmihalyi’s chart itself.

Existential Wayfinding through Failure

There’s an invisible and existential level to wayfinding, too—understanding abstract win and loss states as described and prompted by landscapes, which are artifacts of level design. During a night-time stage, amidst more offensively cheerful guitar strums, I hear the sound of enormous jaws clapping shut. Should I be concerned? I can’t tell: an enemy would be a new element. I don’t see a predator ahead of or behind me, and I don’t see any affordances in the landscape, any clues that might ask me to go through a routine of hiding and risk-taking. What are the mechanics here? More jaw snaps. I don’t see a path forward or to the side. I am frantically waiting to be told what I should be frantic about. The guitar keeps happily strumming for no reason. I am rootless, confused—not for my children, nor for momma badger, not in dramatic terror, but because my perceptions, my keyboard, my avatar, and the landscape all seem to be unplugged from one another, pulling in different non-directions, the mental equivalent of a limb fallen asleep. Everything seems out of place and nondiegetic, and I’m not sure what remains when everything in a game is nondiegetic. This is a nightmare, not because I am about to be killed, but because my interpretive faculties are being pulled apart. I book it in a random direction, leaving my children to deal with the darkness, my badger mother yelping the whole way to the next level.

Once I reach stable ground, the level ends up feeling like a crude prank rather than a simulation of survival. I am unproductive, and I use that word carefully. As Jesper Juul writes, there are multiple origins of failure during gameplay—maybe we are inadequate players, perhaps an unfair game is to blame, or maybe there’s just some bad luck in between the two (17). This typology is productive for conceptualizing failure. It allows us to decide what to do after we fail, no matter if blame is laid at the correct feet. Juul makes these designations clear: “A player who loses a game can be bad at this specific game or at video games in general, claim that the game is unfair, or dismiss failure as a temporary state soon to be remedied through better luck or preparation” (19). Juul has a broader point to make about blame, deferral of responsibility, as well as personal insecurity, but it is enough to see that his typology gives us solutions and prompts actions, usually indignation. Even if I did steer my badger wrongly on multiple occasions, I had little to lose. During gameplay, there seemed to be no consequences for losing children. Perhaps I missed something. The avatar did not seem to experience any losses. If there was any overall danger posed to the player, I was unsure what it might have been, and this, combined with the simple fates of my sadly bite-sized children, contributed to a dysfunctional gameplay experience that left me more empty than contemplative, irritated than invested or disgusted.

Juul is writing about clear-cut failures, though, and I’m not entirely sure that I failed during this level. I may blame the game for my inability to wayfind, and I may fairly or not call it poor level design, but I am not exactly sure what I am blaming the game for or what it was that level was trying to do. It reads to me as a level that’s purposefully broken, intentionally including those mismatched guitar strums, meant to make you inappropriately aware of yourself, or meant to make you misread and panic like an animal—the same as printing a book poorly and saying that the point of it is to make you feel frustrated while reading. It’s not that I need a clear goal or nicely written message to enjoy a game. But, when a bird finally picked me up and flew me away near the end, my sentiment was thank god that’s over with.

Magic Circles within Circles

Useful here, and just as open to complication (or contemplation) as previous concepts, is the magic circle. The magic circle, as the liquid boundary between the creative state of play and the hierarchy of everyday life, is the physical, social, or abstract place in which the game occurs and where its rules are legitimate. The magic circle is understood as a social boundary, not a literal one, a contract recognized by abstract “proper boundaries of time and space according to fixed rules” (Huizinga 32). Inside the boundaries of the magic circle, play happens; meanings are formed and performances executed, often using everyday devices, but ones that are nonetheless different than the meanings and performances found in the everyday.

Huizinga pairs the abstract “magic” location with architecture, which give us concrete boundaries with which to form abstract boundaries (10). To contemporize this pairing, I suggest that these “proper boundaries of time and space” are, in video games, delivered by level design, and the magic circle may be more deeply discussed as a system of affordances or, more simply, a graphic design. The magic circle is demarcated by rules, yes, which are easily represented in text, typologies, and what it means, socially, to play a game. But the magic circle is also rendered through literal boundaries, the visual rules of the landscape that craft affordances and direction that allow a game to exist in the first place.

This take on the magic circle is rendered through the language of forms and affordances and project-to-project critique. Once we think of the magic circle not only as an abstract social ratification but as a literal guide for wayfinding, of designed boundaries of when and where to play, we expand its definition. The performances executed in-game have different meanings depending on where players are in the game’s landscape, some motivations important at some times and not at others, with the difference indicated by affordance. There are certainly subsets of play within the broad category of play, each of which require their own modes of demarcation—magic circles within magic circles, all the way down. A reading of Shelter that focuses on the management of these circles, evident through wayfinding, can help us arrive at this perspective of the magic circle—of how and when to play—as a crucial formal tool that keeps flow, immersion—the game itself!—continuing, and how the mismanagement of these boundaries can result in not a bad game but a ludological and narratological null value, circles that become spirals, an experience that eats itself, a “broken game.”

It should be note that part of this brokenness is my fault—as a non-serious gamer trained by one-time playthroughs of Bioshock or watching over people’s shoulders as they organized spreadsheets for EVE Online, I expect environments that reward, not ignore, the player that gets tied up in knots or forsakes objectives in favor of exploration. I am not a member of Shelter’s intended audience. I expected too much, gave too little. Shelter is a small game, which disoriented me. I could not intuit the game’s deeper structure, read its affordances and place them into broader relationships, or match its expectations, and I had this problem until very late in the game.

But, to be fair, Shelter’s smallness demands a certain economy of design. It takes very little to upset the balance or ruin the player’s palate, and my palate was ruined very early on—indifferent is a good word to describe Shelter‘s design ethic, and not in a way that advances its conclusions about harsh landscapes. I found that Shelter (or at least what Shelter has to show us about flow and failure) contained more Mobius strips than circles, more optical illusions than dependable paths, more non-diegetics than diegetics. These aren’t necessarily bad—other games predicated on wayfinding, such as the Portal series or The Stanley Parable, can successfully own such contraptions—but Shelter‘s casual inclusion of self-destructing design is more than enough to make me struggle.

Works Cited

Csikszentmihalyi, Mihaly. Flow: The Psychology of Optimal Experience. New York: Harper & Row, 1990. Print.

Huizinga, Johan. Homo Ludens: A Study of the Play-element in Culture. Boston: Beacon, 1955. Print.

Juul, Jesper. The Art of Failure: An Essay on the Pain of Playing Video Games. Cambridge, MA: MIT, 2013. Print.

Norman, Donald A. The Design of Everyday Things, Doubleday, 1988. Print.