Homepage » Acting (on) the Text: the Case of New Media (6. 2014) » Kyle Eveleth

Issue 6


When Players Feel Helpless:
Agentic Decay and Participation in Narrative Games


Kyle Eveleth



The ergodic (participatory) element of games is often cited as the core barrier to overcoming a perceived divide between good ludic (gaming) design and powerful storytelling. The present study examines two Indie games, Braid and Actual Sunlight, and their nuanced treatment of player participation in service of effective storytelling. These games in particular test the limits of player agency by asking the player to make ethically and morally problematic decisions, such as killing the main character, en route to completing the narrative. Such unusual narrative methods allow Braid and Actual Sunlight‘s game designers to unveil the mechanisms that afford, constrain, and ultimately revitalise the player’s agency within the bounds of ergodic interaction. Narrative here, rather than restricting gameplay, instead enhances it, offering a tragic moment of cathartic relief as the player is exculpated for his or her decisions during the game. The insights drawn from these two examples and larger-studio offerings like Bioshock and Assassin’s Creed suggest a deeply traditional mode of storytelling at work in many narrative video games, an assertion that allows the ludological/narratological divide to be reknit and sets up ergodic media as a whole (video games, physical roleplaying games, interactive books, and more) for critical reconsideration. 

Culpable: deserving blame.
Sometimes you’re just as culpable when you watch something as when you actually participate


The darkness on-screen opens on a sombre scene: a vaguely Manhattenesque skyline, lit by what seem to be thousands of fiercely burning fires. The word “Braid,” clearly engulfed in flame, hovers in midair just in front of a shadowed figure. In white, not part of the story world, is the phrase “Use the ARROW KEYS to move.” All is still save the movement of the flames and the swelling of cellos as that instructional text fades to display another prompt: “Press ESCAPE if you want the menu.” When the viewer-cum-player crosses the ludic threshold and presses an arrow key, the shadowed figure immediately follows the order. Jeff Keaton, one of the participants for the Youtube channel, Graphic Gaming’s “Let’s Play” series, sums up the shift from inactive viewer to active player nicely:

And— Hey—Oh, you—this is an, this is an, a, uh—isn’t just an opening, I thought it—I        was—we’re actually playing! This is— oh is this— is this, this is, obvious— this has got   to be the start, right? Do I get—ooh, hello Mr. Suited Man! (“Let’s Play Braid – INTRO-Part 1”)

Unlike many other platforming games on which Braid bases its style and gameplay expectations (solve the platforming puzzles and beat the game), in which a title screen greets the player and offers and allows them to choose when to begin (usually by pressing the start button), Jonathan Blow’s offering instead tosses the player headlong into the game. In this artful opening gambit, Blow’s Braid (2008) sets a demanding precedent for the remainder of the game: player participation is critical. Braid is popularly considered an abstract “art game,” or a video game that “work[s] off a few basic assumptions: Games have rules, rules have meaning, and gameplay is the process by which those rules are tested and explored” (Bissell 96). One of those rules is that games are ergodic–requiring nontrivial effort to be experienced (Aarseth 1)–and dynamic–requiring “action to be encountered, closed, and dealt with” (Eskelinen 39). Thus, unlike other narrative media that places only ‘trivial’ effort on the reader (such as film and literature) (Aarseth 1), games require participants to perform three additional functions above and beyond usual interpretive practices: exploration (deciding which path to take), configuration (partial selection or creation of “scriptons,” possible storylines), and textonic actions (the act of traversing the textual elements of the game) (Aarseth 62, 64). 

Many critics and game designers argue that this systemic mandating of participation, inherent to the gaming medium, lies at odds with the medium’s functionality as a narrative one. Bissell notes, “the video-game form is incompatible with traditional concepts of narrative” (93).  Further, Eskelinen argues that the fundamental structural difference between games and narratives is that “in games, the dominant temporal relation is the one between user time and event time and not the narrative one between story time and discourse time” (37). Because of this structural conflict, critics and designers contend that constructing a sound story to match a sound game is “impossible” (Blow quoted in Bissell 93). Sophistication in game stories thus often comes at the price of lessened participation, resulting in “problematic, largely unreplayable, story-game hybrid[s]” (Aarseth 10). Conversely, the most elegant of games tend to have inelegant and simplistic stories (for instance, “big, dumb, loud action games” [Bissell 96]). 

The fundamental implication we are to take from such theories of game design–told primarily by respected game designers and game theorists–is that gameplay is anathematic to good storytelling. A tradeoff between story and gameplay must occur. Eskelinen argues that the two narratologically critical elements of narrative, a temporal sequence of events and a narrative situation, are either compromised or nonexistent in games (37). Ostensibly, plots are disturbed in that the timing of the plot hinges upon the player’s actions, which can deviate from narrative timing, and narrative situations (narrators and narratees) are absent from the gaming medium (Eskelinen 37). In a similar way, viewing games through a strictly narratological lens tends to yield only partially significant readings: reading the plot of The Elder Scrolls V: Skyrim (2011), for example, must contend with the fact that the player can abandon a particular storyline at any time, leaving any loose ends associated with that quest hanging forever (Eveleth 2012). 

Unfortunately for many of these critics, this dichotomous view of video game interpretation is largely unsupported by the source materials, video games, which are designed playgrounds that can be potentially narrative. J. Yellowless Douglas explains that “looking to either narratology or to [ludology] for our understanding of interactives will offer only a highly limited return” (37). This recognition of the limits of either branch of game studies fuels my reading of two “art games,” games ostensibly in which “it is gameplay and not story that serves as the vehicle for meaning” (Bissell 96). This study focuses on the most fundamental thesis about games as an entertainment medium, that they are ergodic (participatory) in nature. It examines the ways in which both gameplay and narrative in two Indie (independently-published) “art games,” Braid (2008) and Actual Sunlight (2012), contribute to and also detract from the notion of games as a participatory narrative medium. In this examination, I note opportunities for player agency (that is, the ability to take meaningful action without coercion) as well as limitations to agency, moments that contribute to what I call “agentic decay,” or gradually-increasing awareness of nonagentic (coerced) choices. Ultimately, these shifting opportunities for player agency raise difficult questions about control and culpability in participatory media. 

Braid and Actual Sunlight have been chosen for this study because they share three similar fundamental features. First, they are “Indie” games, in that they were produced by either an individual or a small team and not a larger corporation, like EA Games or Ubisoft. Indie games are most often experimental or “art games” and are likely to actively problematise many of the assumptions about video games as designed objects in which play occurs. In addition, both games employ an “old school” style (Bissell 96), which is known within the gaming community as “retro.” “Retro” games use art styles and gameplay genres most often associated with earlier eras of gaming, most of which are named for their graphical limitations (and associated hardware limitations): 8-bit, 16-bit, 32-bit, 64-bit, and 128+ (the contemporary era in which both games were released). These eras are also most often associated with the consoles on which games were played: the Atari home system, the Nintendo Entertainment System (NES), SEGA Genesis, Super NES, the SEGA Dreamcast, the Nintendo 64, the Playstation 1, 2, and 3, and the Xbox and Xbox 360. It should be noted that, for some gamers, the traditions into which these two games fit–side-scrolling platform puzzle game and 16-bit roleplaying game–adhere to structural and formal expectations as codifying as any literary genre. These “hard core” gamers come to the console with expectations for gameplay and systems of work and reward already in mind; these expectations can and often do vary drastically from what less experienced players, “casual gamers,” bring as expectations for the game. Finally and most crucially, these games were selected because they uniquely attempt to “braid” together story and gameplay in meaningful ways. In each, I will primarily focus on the ludic structures of each game, those elements that make them a game: design, including rules and paths for completion; mechanics, such as being able to jump or not; art direction, including both avatars (sprites) and backgrounds; and music, both sound effects (SFX) and atmospheric (background). After describing these structures, I will briefly but meaningfully note how these elements dovetail with the narrative elements to reinforce or undermine the gameplay’s message about player agency. 


Jonathan Blow’s Braid has received widespread critical praise from gamers, designers, and even literary critics. Widely viewed within the gaming community as an Indie masterpiece (Whitehead; Cavalli; Benedetti; Chaplin; Quillen), Braid has also garnered some critical attention in the academy; John Kerr called it “the sort of ontological labyrinth that Jorge Luis Borges might have made” in Film-Philosophy (166). Braid tells the story of Tim, a well-dressed young man who has one goal: to find his princess, lost to his arrogance and selfishness. Tim is empowered with the ability to control time, and because of this power, he cannot die; moments in-game that should kill Tim instead cause time to freeze, prompting the player to rewind Tim to a safe point in time. Initially able only to rewind, the player unlocks more of Tim’s abilities as they progress through levels, including the ability to fast-forward time and the ability to send ethereal Tims into the future by performing an action, then rewinding. Simultaneously simple and complex, Tim’s story has generated an unprecedented outpouring of fan theories and interpretations about its significance. The most pervasive of these is that Tim is a researcher seeking the secret of his princess, the atomic bomb. Weaving together time, space, and expansions or contractions of the two, this nuclear interpretation suitably unites two of the major themes of both the gameplay and story of Braid: progress and work are desirable, but some ends are better than others. 

It is not difficult to understand why developers appreciate Braid so much. Blow, like his game, is known in the industry as “the platonic ideal of an indie game developer” (Bissell 91), able to take widely disparate ideas and unify them in interesting–and eminently enjoyable–ways. Utilising perhaps “the most archetypal video-game genre” (Bissell 97), the side-scrolling platformer, Braid nonetheless innovates by undermining the two primary directives of that age-old genre: to finish the level quickly and completely (or correctly). Most side-scrolling platformers feature three elements that execute this implicit design concept: a method of timing and rewarding haste while punishing sloth, a limited number of attempts, and a scoring system that rewards collection of objects secondary to the goal of finishing the level.

The time limit is, in many ways, the hallmark of the platformer. Because platformers have one designed path to the goal, they must present a challenge by preventing players from taking as much time as necessary to find that correct goal. In Super Mario Brothers, for example, a time limit is imposed on the player; failure to finish the level within the time limit is punished by taking away one of the player’s tries and forcing him or her to begin again. In addition, speed, quantified in the form of leftover time, is rewarded by giving additional points for each leftover unit of time once the level is completed. Haste is also implied by the game’s visual and aural art: in Mario, a warning SFX is initiated when 100 time units remain and the music doubles in tempo; in Sonic the Hedgehog, the eponymous player avatar will tap his foot impatiently should the player not make a command for a few seconds, indicating his distaste for standing still. Braid destabilises this generic convention not just by refusing to quantify time to completion but by making the manipulation of this fundamental resource the mechanical focus of the game. From the opening screen to the final sequence, the player is always in control of Tim’s actions; there is never a moment when the player is cued to take control, as in most platformers and their quasi-ubiquitous “press start to play” mantra. [1] The lack of timer gives players a chance to breathe (Tim’s idle animation is notably him taking deep breaths and looking around). Bissell notes that “half of the pleasure of Braid, at least initially, is simply to stand there, look, and listen” to the combination of the “beautifully aglow” visual world and the “celestially lovely” soundtrack of Braid (99). Though players must eventually figure out the puzzles and solve the levels, they are encouraged to take their time in doing so. 

The second feature, a limited number of attempts, is critical in many platformers because it tacitly encourages players to hone their skills en route to being able to complete levels more quickly. Bissell notes that “mastering a platformer…is not play; it is a physically crushing process of memorisation and reflex mastery” (96). Failure to complete the level forces the player to go back, ostensibly giving them a chance to memorise the parts they completed and to implicitly begin to compile the skills needed to succeed at the more difficult levels. With this system of limited attempts in place, players are encouraged by design to recognise dangers that will cause them to lose an attempt: pipes in Mario, which could contain Piranha Plants, gaps in the floor that will end the attempt, or telltale signs of a long leap coming up. In addition, players are encouraged to learn the quirks of the system: Sonic and the other playable avatars in Sonic the Hedgehog are almost incapable of microadjustments and stopping quickly, so that players must learn to cease their command input at a different time than in the ice levels of Mario. As mentioned earlier, Braid has no death system and no tries. Players just rewind to a time when Tim was not in mortal peril. This can be encouraging for players who are not as adept at platforming, providing a less frustrating method of re-trying a mistimed jump. Even more adept players, the lack of punishment for mistakes encourages a more plodding, explorational play-style: testing the exact limits of Tim’s jumping abilities can be done with no negative consequences and at one’s own pace. 

In platformers, players can usually recover lost tries by collecting items in the levels, such as coins in Mario and rings in Sonic. Accruing a certain amount of the item, usually 100, yields another try (a 1-up), and often gives a small but appreciable number of points. Other collectibles make it easier to complete the level, such as the shield power-up in Sonic or the Tanooki suit in Mario. These power-ups usually take the player on a path away from the final goal, but make up for this temporary slowdown by enabling the player to negotiate some obstacles more quickly. The fire flower in Mario, for example, allows the player to shoot a fireball, which kills lesser enemies on contact and bounces along slightly faster than Mario’s top running speed. In Braid, there are no power-ups, no coins, and no scoring systems. The collectibles are puzzle pieces, which combine to form pictures of moments when one (perhaps Tim) might conceivably want to rewind time (such as knocking over a full glass of wine) and stars, which are incredibly difficult to find items that change the ending of the game. Without worrying about gathering resources, Braid’s players can feel free to play as they like, perhaps going back for the puzzle pieces after completing the game and finding the stars on a subsequent play-through. 

The desired effect of these features in platforming games is a reduction of play to instinctive, mechanistic, unthinking action. The prize for many platformers is called a “speed run,” or an attempt to complete the game in the least amount of time necessary and using the fewest number of resources while accruing a high score for the time spent playing. Often, these runs are measured not in the hours of playtime contributed to role-playing games like Skyrim but in minutes or seconds (even frames!). The current world record for fastest Super Mario Bros. run is four minutes, fifty-eight seconds with no lives lost and 117,000 points collected. The concentration, memorisation, and reflexes required to complete such a feat makes it difficult to consider, in-game, why Mario is hitting the Goombas in the first place, why Daisy never changes her locks, and why Bowser has such problematic hobbies as kidnapping. Unlike many other platformers, Braid invites critique in these locations. Blow notes in an interview with Bissell that the enemies were designed purposefully to have human-like faces and to make a sorrowful booing sound when killed. He tells the questioning Bissell, “that guy didn’t want to get jumped on” (Bissell 101), indicating that players should consider the consequences of their in-game actions (ibid). This effect is compounded by the increasing amount of time the player spends in-level considering the puzzle and, more often than not, reversing time to kill an enemy again in a more advantageous location. To die then be raised from the dead only to die again: this is the enemy’s purpose, and his sad booing laments his relegation in the game. 

Indeed, all of Braid’s design elements are purposefully crafted to invite reconsideration of initial appearances. Braid’s artistic style, the “arcadia of chuggingly locomotive clouds, heartbreaking dusk, small scurrying creatures, and lustrous flora” (Bissell 99) depicted in hand-painted background paintings mixed with the spritely but hand-drawn characters combines to initially appear innocent, simple, and childlike (Bissell 98). The soundtrack, licensed for the game from composers instead of written specifically for it, is string-heavy and atmospheric, ranging in motifs from the pastoral to the epic. [2] The first world, World 2 as it turns out, is lush and green and a bouncy soundtrack accompanies it. As players progress, however, the tone saddens and the art darkens. Later levels employ a darkly harmonic musical theme that can be played forwards or backwards, sounding eerily beautiful either way. The final world’s runway to the finale features increasingly anxious music and frenetic background art, filled with clashing colors and shapes. The game thus shifts from safe and innocent to fretful and foreboding, leading players to feel that this is not as happy a place as it first appeared to be.

This complex relationship between appearance and reality in Braid suggests that even the messages players have been sent regarding gameplay may require added scrutiny. Each of the gameplay techniques examined above suggest unconventional generic freedom; platformers simply do not let their players wander through their worlds. They are designed for action and progression. Though Braid harnesses these nostalgic expectations to creative relative freedom for the player to act, turning its own critical lens back upon itself reveals that these are illusory freedoms. With or without time constraints, the player is still demanded by the genre and the medium to progress through the game, and that means finding the solution to each level. The levels, difficult and yet rewarding puzzles as they are, are still puzzles that feature only a single solution. The point of the game is still to “solve” its puzzle and progress through it to the end. In order to “beat” the game–the point of any game is that it is a means (play) to an end (beating the game)–players cannot simply sit around absorbing the aesthetics of the game world. Once a player beats the game, there is not much left to do. Critics in the gaming community have levelled some criticisms at Braid for this lack of replayability, noting that once it has been completed, the player gains little from replaying the levels aside from the usual rewards: increased skill and speed. Unlike a roleplaying game, in which a player can unlock different storylines through alternate playthroughs (and sometimes can only see this content by playing differently–see Bioware’s Dragon Age and Mass Effect series and Peter Molyneux’s Fable series), Braid yields nothing substantial to subsequent play. Going back into previously-beaten levels does often allow the player to collect interesting “Easter Egg” items, but as I will contend later, these do not break the largely unidirectional nature that Braid ultimately exhibits. 

Even Braid’s lauded story, presented in bits and pieces and left up to the player to unravel, is not as liberating as it initially appears. Though players are encouraged to take an active role in traversing the text and generating its textonics, revealing the words and putting together the story on their own, the story is ultimately a linear one. It is not influenced much by the player’s actions, which go largely unnoticed in collecting the elements of the story. There is a single correct–that is, designed and intended–layout of the story, a single plot skeleton to which the flesh and sinew of the story must be affixed. Though initially a jumbled set of vignettes, progressing through the story supplies the player with clues to the order of snippets in the way that collecting more pages of a dropped essay allows the collector to reorder it appropriately. Though initially seeming to be the moment of freedom players may use to negotiate their discovery of the actual linearity of a “nonlinear” game, even this wiggling of narrative elements becomes rigidly structured at the game’s conclusion. Upon collection of the final page, the game explicitly reorders the story into its correct orientation, and in order to read it, the player walks to the right through the books–the traditional direction of forward progress in platforming games. The final level initially depicts Tim’s prized princess escaping from a monstrous knight and a wall of all-consuming fire. Working collaboratively, Tim and the princess navigate their obstacles and outrun the wall of fire. As Tim climbs a final ladder to the princess’s house, ready to claim his prize, the screen flashes and Tim finds the door locked. When the player presses the rewind button, the truth of the story is revealed: the princess was running from the monstrous Tim, trying to stop him with obstacles that he avoids, until the knight rescues her. The game’s temporal ordering is revealed to have been a jumble, and Tim’s powers of time control are questioned as real powers or as the musings of a madman. No matter the story that the player has constructed of Tim’s love for his princess; the truth is in the text, so they say. In a sense, the game names the player as the cause of Tim’s grief–he has, after all, obediently put his life on the line for the player to learn about his mistakes. Reading in reverse, as Braid encourages players, we see that Tim was moving from the complications associated with the crisis to the much simpler time of working through the trauma. Thus the common gameplay trope in platformers of increasing complexity runs contrary to the narrative, which requests simplicity but must yield to complexity in the interests of entertainment. 

In this way, Braid’s story, far from being sabotaged by its ludic elements, instead stabilises the critique they make possible. That critique is of the player’s role and ability to act without being coerced in participatory media. While Blow sets his game apart from those stately melodramas pinned to mechanically elegant games he denounces as not satisfactorily addressing game interactivity (Bissell 93), the narrative structures combined with the ludic structures show that he has in many ways crafted the very thing he loathes. Blow notes that Braid is designed to introduce doubt about the reasoning for a player’s actions (102), and using the narrative of Braid as a decoder reveals that Braid is not safe from these questions. From its pseudo-nonlinear plot to its pseudo-liberating method of storytelling to its time-reversing finale, Braid shows that the player has been very carefully led down a curving but ultimately linear path, completing challenges in specific ways and constructing the story in ways that would not betray its twist ending. In this way, Braid’s initially liberating model is exposed as entirely dictatorial, though it becomes questionable just who the dictator is. Braid seems to suggest, as I noted above, that it is actually the player’s fault that Tim has revisited the traumatic series of events he was in the process of dealing with when the player first meets him on the title screen. In the same breath, Braid critiques the stability of medium- and genre-specific traditions and expectations. In an age with fewer mechanical limitations, why must right equal forward in platformers? Other Indie games, like Fez (2012), have innovated by changing the directionality–up is forward. The stable and linear designs of levels, also largely a hardware limitation based on memory and boot times, is a tradition that has continued largely because of player familiarity and generic recognisability. Much of the success of contemporary platformer LittleBigPlanet is due to its profound level-editing engine, which has allowed a large community of level designers to continue building content for the game. If there is a dictator at fault for the design of this game, Blow reminds us that the same designer, in making a profitable and entertaining game, must follow some conventions, many of which continue primarily because of a complex web of player and developer influence.         

Actual Sunlight 

But what happens when a game claims at the outset that it is entirely dictatorial, that it is not a game because there are no choices, no variability, and no means to reach a shifting end—especially when it uses a game engine? Apparently opposed to Braid’s initial freedom is Actual Sunlight, which claims that its “gameplay is minimal, and serves only to move from one part of the (admittedly) text-heavy story to the next” (O’Neill). Will O’Neill’s interactive short story game has drawn its fair share of controversy for shutting down player participation. “Formalists have argued that games like Actual Sunlight aren’t games,” reports TheIndieMine’s Mark McAvoy (1), recalling formalist Tadgh Kelly’s note that such games generally focus on an internal conversation and are made to deliver a message, not a playing experience (1). On the other side of the debate are critics like Leigh Alexander, who contends that the use of the games medium makes Actual Sunlight a game in the same way that John Cage’s use of pauses is still music (1). O’Neill remains neutral in the debate, considering Actual Sunlight both an interactive story and a game. Indeed, on the Actual Sunlight information page maintained by O’Neill, both terms are used somewhat interchangeably to describe the experience. Players of Actual Sunlight are privy to the story of Evan Winter, a depressed young professional in Toronto. Focusing on three periods of Evan’s life, “the story is linear, unavoidable, and (hopefully) thought provoking” (O’Neill). Ostensibly removing player desires from the mix, O’Neill explains that “you [the player] experience [Evan’s] perceptions, fall under the consequences of his decisions, and meet everyone who didn’t change him.” 

At first blush, the story of Actual Sunlight appears to completely subsume any freedoms the player may have going into the experience. Using the genre of the 16-bit role-playing game, Actual Sunlight makes its alterity from reality immediately known: players face a familiar title screen and press a key to begin the story. Interaction in the game is limited to the familiar gestures of the genre: move the character, in this case Evan, around the map and interact with things and creatures by pressing a key. Like a good Hitchcock film, the maps, or playable levels, in the game are extremely small, cramped, and absolutely impermeable until you are forced by the narrative to move into or out of them. From time to time, the game will offer the player the freedom of selecting the next action from a traditional RPG dialogue menu box, asking if Evan would like to buy an unhealthy meal or a healthy one or asking if he wants to play video games or go out. Choice is not really given to the player in these situations, however; though the cursor can indeed be moved and different choices can be made, selecting the incorrect answer, the one that does not progress the story along its linear path, results in Evan refusing to execute the player’s selected action. The player is stuck until they choose the linear choice, which is always the one that leads to the game’s tragic ending: Evan’s isolation and eventual defenestration from the top floor of his apartment building. 

Like Braid, Actual Sunlight subverts the generic conventions many players bring to the experience. Role-playing games, unlike platformers, usually provide many opportunities for player investment and truly agentic participation. Players often can name their characters, control their progression, and dramatically change the stories that befall their avatars. In many of these games, players are tacitly encouraged to identify with their avatars, feeling that they are an acting extension of the player into the game world. Exploration and configuration are heavily centered on the player, and traversal is often nonlinear. In this way, role-playing games are usually more ergodic than platformers (they offer more avenues for meaningful and ‘nontrivial’ participation). This is a main reason why these games usually offer ‘save points,’ or ways to pause the action and retain whatever the player has accomplished, which most platformers lack. A secondary but related reason for this is the drastically greater amount of time required to finish most role-playing games, which can take days of game time to finish satisfactorily. Actual Sunlight systematically subverts all of these generic conventions. The player is never given an opportunity to customise or configure Evan, and they always remain something of a third person to the action; avatar identification is purposefully minimised, and Evan is not to be viewed as an extension of the player. Exploration is likewise minimised, with most of the maps occupying the space of a single screen or less, making the entirety of the map viewable by the player at once. There are no secrets to be discovered, no items to collect, no monsters to fight, and few conversations to have outside of the required and scripted ones. When people are finished speaking, they cannot be spoken to again. Though players have the option to save, doing so is presented not as a way to protect progress that has been made but to offer a ludic ‘way out’ of the action: the save screen seems to say to the player, ‘it’s okay to stop now–go do something else.’ In total, the game takes about fifteen minutes to complete–about as long as a fast speed run of Sonic the Hedgehog and about as long as avatar customisation in Skyrim

Each of these subversions nullifies the liberating elements of conventional role-playing game ludic structures. The final sequence in Actual Sunlight make explicit that the player is not in control. The player, controlling Evan, walks toward the top of the screen, through a series of doorways. Evan’s passage through each door causes it to lock; attempting to turn back tells the player that there is no turning back. When Evan reaches the window, the player is presented with a dialogue box containing four selections. Each is identical; each says “kill yourself.” The player cannot back out of this selection box, and the only way to not decide to kill Evan is to stop playing. Selecting the only choice available causes the screen to fade out, giving the player the usual “game over” associated with avatar death. 

There are slight glimmers of hope for avoiding the tragic ending of Actual Sunlight, but these are all essentially trivial. In some situations, the player can rebel against the structured game experience by not taking the actions Evan tells him or her to take. For example, Evan returns home drunk and angry, and the player is instructed to destroy objects in the apartment. The player can complete this scene by going to Evan’s bed and walking into it, refusing to destroy anything. [3] The player can also refuse to progress in the story, but this, like not selecting the critical choices, is not really playing as it does not use the available means to reach the end of the game. Regardless of the rebellion, the game glosses rejections of instruction and continues on its relentless path to the unavoidable conclusion.

The imprisoning constraints made through the ludic elements of Actual Sunlight are replicated in its aesthetic experience. The music is of the same kind and quality of classic 16-bit MIDI tracks, tinny on the high end and muted on the low. Unlike an RPG, however, there is very little music in the game. Music in RPGs often serves as an auditory indicator to the player than something has changed, whether it is the map (from town to world or battle), the narrative situation, or the game’s state and conditions (game over, victory, gain or loss of an important plot element). In Actual Sunlight, the music is as atmospheric as its classic inspiration, but it is not differentiated much. For the most part, the tracks are dark, sombre, and match the tragedy of the narrative: they feel like a crushing weight in the way the gameplay weighs on player choice. Sound effects are harsh, played at a higher volume than the background music, and represent atmospheric effects, such as Evan destroying his game console or the bus revving as it drives him to work. Similarly, the visual style is sparse and largely unmodified from the standard RPG Maker chipset. Partly a result of budgeting concerns, the game nonetheless inflates this lack of customisation and builds it into the sterile, dark atmosphere of the game; there is no fertile ground for players to find identification with anything but Evan’s tragedy. Even these urges to identify with the controlled player character, often enhanced by the cartoonishness of artistic styles, are ultimately halted by Evan’s distancing from non-player characters, who remain flat, and from the player controlling him, who seems not to have any control at all. The inclusion of more realistic representations of faces for speaking characters, including Evan, reinforces this separation between player and gameworld: players are peeking in on the world, not actors within it even if they are the ones making Evan move. 

Like water, however, constricted freedom will often find other wellsprings through which to flow. Constricted by both ludic and narrative structures, the opportunities for agentic action decentralise from the game, shifting from the configurative, explorative, and textonic functions of player experience into interpretation–the realm of all signification, regardless of medium. A large and passionate community has risen around the Actual Sunlight experience, crafting essays and pushing for its eventual publication on Steam (O’Neill). The game has received impressively positive reviews given its content, most often centered on its treatment of participation and predestination. Actual Sunlight’s constricting gameplay and inescapable narrative exposes that the key ingredient of games as a unique medium, their ergodic nature, is not essential to the medium. 

Taking a cue from Actual Sunlight, players may read the limits of agency in both ludic and narrative structures in other games that are more widely considered ‘games.’ In all games, I suspect, there becomes a moment when the player’s status as agent becomes questioned. This moment marks the beginning of agentic decay for the player, or the realisation that one’s actions thus far have not been undertaken freely but instead have been the result of coercion on the part of the game. A game’s structures embody and enact a specific ethics in that they encourage and discourse certain player behaviours (Sicart 25). That is to say that a game’s rules, methods of play, and means of winning are all bound up in ethical considerations. Sicart’s examination of XIII exemplifies how these ethics alter player relations to the game. In XIII, Sicart notes, the player is an amnesiac assassin; killing is thus an expected element of the game (though not necessarily, as Sicart carefully explains, the only course of action). However, the game does not allow the killing of police officers and civilians, a limitation that exposes a certain ethical stance: killing police and civilians is bad, so bad that it cannot even be done. Likewise, Assassin’s Creed imposes serious ethical limitations of the same kind on players, kicking them out of the simulation via game over if they kill innocents. By reading these limitations, players can better understand the values and ethics of the game, which is a hint to how it expects players to think and act. The act of reading these structures contributes to agentic decay because they critically question the validity of the assumptions underlying these ethical structures. For example, the game Metal Gear Solid discourages frontal aggression by limiting the player’s resiliency, access to weaponry, and ability to fend off multiple attackers as well as giving the enemies unlimited numbers and unlimited ammunition. Instead, Metal Gear Solid encourages stealth by easily handling variables like line-of-sight, loud noises, predictable enemy movements, and the ability to hide from pursuers relatively easily. The player may be much more readily equipped (through cheats or familiarity with shooters instead of stealth simulations) to deal with the game using a more frontal approach. Thus, it becomes known to the player that the fullness of their choice in how to play the game is being limited. This awareness is the decay of agency. It is an element of all games: football disallows the use of hands except for goaltenders, draughts relies on diagonal motion, and blackjack imposes a strict upper limit on the total value of cards. It is my contention that, upon first entering a game, an uninitiated player feels as though they have more agency than a player who has learned the rules; this is a natural element of games, which by their nature restrict actions to be ludic. This decay has a limit in non-narrative games, however, because the player can become familiarised with the rules and learn to play in ways that allow for agency to be regained. In narrative games, however, decay has a much larger limit because the player plays in service of telling the story. The only way to rectify agentic decay in narrative games is to dramatically seize it from the game, an action that can only be undertaken by breaking the game’s rules or by refusing to play. Play within the rules can be masterfully executed, as exhibited by chess grandmasters, but nonetheless the fullness of their agency is limited by rules. 

Extending this ethical question, Braid’s ludic design subtly discourages the very playing style it attempts to overtly encourage: leisurely exploration. This is a result of its generic conventions and the rules governing platformers, namely those requiring solution of levels to advance, and advancing in order to win. Because leisurely exploration does not contribute to the overall goal of finishing the level to the extent that experimenting with the mechanics and abilities of the game’s engine, it is valued less than those actions. Though struggling through the puzzles is a part of the pleasure of the game, this pleasure contends with the pleasure of solution and completion. 

This is not the grand sum of Braid’s gameplay experience, however. The aesthetics at work, specifically the backgrounds and the music, encourage a kind of timeless experience within Braid. More to the point: there is no penalty for playing slowly. As I pointed out earlier, Braid does not feature the timer by rule that other platformers utilise; there is a “speed run” option, but this is not part of the normal gameplay experience–it is instead a separate mode that can only be chosen, logically, after beating the game once. Braid does offer a reward to players who ignore generic conventions in the form of extended information. Initially, players can find puzzle pieces that, when put together properly, depict scenes from Tim’s past in which he used his powers, situations which are alluded to obliquely in the text of the story. Furthermore, players who commit to the idealised playing style envisioned by Blow when designing Braid can collect eight stars that change the ending of the game. Instead of finding the door locked in the final sequence, Tim is able to catch the princess, who immediately explodes, killing all the other living things in the world of Braid. This alternative ending does not reverse the story thus far, but instead adds context to the story, just as the puzzle pieces do. Finding the stars adds text to the epilogue that reveals the princess to have been the atomic bomb:

She stood tall and majestic. She radiated fury. She shouted: ‘Who has disturbed me?’ But then, anger expelled, she felt the sadness beneath; she let her breath fall softly, like a sigh, like ashes floating gently on the wind.

She couldn’t understand why he [Tim] chose to flirt so closely with the death of the world. (Braid epilogue)

In the context of the game, however, this ending is the linear, unavoidable, thought-provoking conclusion to which the game comes regardless of player intervention. Epilogue text seen regardless of the player’s possession of the stars alludes to the Trinity test in the Jornada del Muerto desert:

On that moment hung eternity. Time stood still. Space contracted to a pinpoint. It was as though the earth had opened and the skies had split. One felt as though he had been privileged to witness the birth of the World… 

Someone near him said: ‘It worked.’
Someone else said: ‘Now we are all sons of bitches.” (Braid epilogue) 

The gaming community has astutely observed that Tim is a researcher working on the Manhattan Project, and that the Princess is the bomb. This implication, made in the narrative, is also present in the design of the game, which opens with the same burning Manhattan skyline that is seen when Tim emerges from the final level. The player becomes enmeshed in a pun-laden web of complicity. Time has stood still, and the player’s play-throughs represent Tim’s endless attempts to gather the uncollectible end, the princess. Yet, had “Tim(e)” only stood still, not been moved by the player, then this historical moment would never have been realised. Conversely: without finding the content, the story would have remained largely unintelligible and uncritical. Thus the player is complicit in both the horrific destruction following the bomb’s realisation (because they pushed Tim to the conclusion) and in the freedom of discovering this ending, interpreting it, and finally arguing it. The player’s ethics of gathering and completion—the same ethics that caused them to ignore the conventions of the genre at hand, exploring in Braid and giving up their freedom in Actual Sunlight—are thus called into question for bringing about the destruction of the Braid world in the same way that Actual Sunlight accuses the player of killing Evan Winter, even as they are recognised as the sole compelling element in the game that has allowed this narrative content to be unlocked. Far from condemning player participation, these works draw attention to the ways it is modelled, influenced, and structured by medium expectations, generic conventions, and personal ethical beliefs.            

The “Blame Game” 

Ultimately, these games raise a thorny question about participatory media: at what point does participation, especially if it is not known to be coercive, implicate a player in the events that take place in its narrative? Though, as Richard Schechner notes, all media requires some level of tacit participation, from interpretation to identification to choosing to continue watching or reading the text (192), these “trivial” (Aarseth 1) actions can often be used as defenses against culpability. Despite being a voyeur, the viewer of Rear Window (1954) is not responsible for the actions James Stewart takes; they are commonly considered inculpable for the purposes of determining the ethics espoused by the film. For games, however, which require ‘nontrivial’ and thus intentional participation, the question of culpability becomes a valid one. 

Indie games are not the only ones asking these difficult questions. Many large-developer games contemporaneous with Braid have raised these questions in meaningful and similar ways. 2K Games’ Bioshock (2007), Ubisoft’s Assassin’s Creed (2007), and Rockstar’s Grand Theft Auto IV (2008), all raise interesting questions about the desires of players and the desires of protagonists. In Grand Theft Auto IV, the player’s character Niko often reflects on the horrible things he had to do as a veteran of an unnamed Eastern European war, usually indicating that he hopes moving to Liberty City (a fictitious New York City) will obviate any need to retain the skills he honed there. As it turns out, his unique set of skills is exactly what makes him the ideal protagonist of GTA IV, and the player’s desire for completion stands starkly in opposition to Niko’s desire for freedom from his past. In Bioshock and the Assassin’s Creed series, the question of coercion is expressed explicitly: the player of Bioshock is revealed to have been a puppet for the mastermind villain, slavishly carrying out orders. Players of Assassin’s Creed are asked to consider the similarities between their situation, controlling protagonist Desmond Miles in a platforming game and gaining necessary skills to follow more complex orders, to the protagonist’s situation as puppet for the player and a god-like force, Minerva. In these games, the question of culpability is perhaps easier to ask because these games make explicit their connection to the world of lived experience. The level of realism may be used to supplant the player’s reality with the game world (as Bioshock does), to rewrite extra-textual history (as Assassin’s Creed does), or to align real-world locations and real-world problems with seemingly identical game-world locations and problems (as in GTA IV). Players are encouraged to think critically about how their actions relate not just to the fictional worlds they enjoy but also to the real world from which they were created and in which they exist. 

Making players aware of the real-world implications of their actions seems to increase the culpability felt by players in these games. Jesper Juul notes that as readers we tend to “accept regular, noninteractive tragedy in part because we lack any responsibility for the suffering” (112). In games, however, we feel an equal and sometimes aligned, sometimes opposed set of responsibilities, for the failure and suffering of both player and character (ibid). Juul notes that though players may feel guilt, culpability usually ends there:

Still, we probably do not feel entirely responsible for tragic events in a game since they are neither real nor entirely within our control.
After all, for any linear game, it is the game designer who designed the suffering and made it unavoidable…the tragic moment in the game is the one where we are relieved of agency. (Juul 113) 

Although Juul is here specifically referring to tragic games, this argument may be extended to the tragic ludic moment at which the player discovers that his or her actions have been predestined by the design of the game. For, in narrative games, play is in the service of a story, a significant removal from the playing style of chess or backgammon. This narrative tragedy, dislocated from game character to game player, completes the process of agentic decay. Agency has been reduced and the player’s actions, though meaningful, cease to have been freely taken. In some games, like Bioshock, this results in the player taking an action that is ultimately bad for them. If only the player had not helped Atlas, many of the later trials and tribulations–indeed the entirety of the final encounter–would never have happened (though we cannot know what this would look like because the game does not allow it). In others, like Actual Sunlight or Juul’s thought-experiment Anna Karenina game, it results in the death of the player’s avatar. But in these situations, unlike in non-narrative games, the player is relieved of culpability; the player has played well, and the end results of their actions, even if tragic, are what have allowed completion of the game. In solitaire, loss is the tragic event, but it is tied to player failure, and the player is responsible for it (not having the right skill or practice to win); in Bioshock, being duped by Atlas is playing properly. When the curtain is pulled back and the player is relieved of agency, he or she is likewise relieved of culpability for the actions on-screen. 

 In this way, the exculpation of the player significantly alters the player’s participation in ways that can be fruitful for more meaningfully examining games as a narrative medium. Because games are designed systems of affordances and limitations, all of which are governed by a distinctive and expressed set of ethics—the rules—players may only be considered culpable in the boundaries of these affordances and limits. Further, when those games include narrative elements that place more constraints upon the player, the boundaries of culpability narrow. Players of Pokemon may, for example, choose any number of self-imposed gameplay limitations to increase the difficulty of the game (playing only with Pokemon given to the player, by never buying items, or, as recently done on the website Twitch, by crowdsourcing the input commands), and those self-imposed actions within the boundaries are ones for which the player must be held responsible. Players of Assassin’s Creed cannot, however, choose to be merciless killers of the world as a whole—the game’s narrative and affordances do not allow it. The player must not be held responsible for not committing actions not afforded by the game or for doing what the game mandates. [4] 

This holds two important ramifications for games as a narrative medium: first, that they are not so different from traditional narrative media forms as to be unintelligible to all critical practices used by those forms, and second that they are distinct enough by virtue of the different participatory demands to be examined thoroughly in medium-specific ways. The first is a measure that should by all means be taken as a sign that games are a worthwhile medium of study in larger fields of narrative analysis. I maintain as Schechner argues: we must begin to consider games as “not narratively revolutionary but deeply traditional” (192). 

The second, however, is a more fraught position because it requires more work going forward for game studies specifically. Having established that games are rather traditional in their method of storytelling even as they put special demands on the player for participation, we must consider that in games, as in other media, form and function are not necessarily divorced. This means that, in addition to being included in the usual canon of acceptable artefacts for scholarly inquiry, games studies focusing on narrative games must dissolve the instrumental divide between ludology and narratology. Though these approaches have been separately productive in the past, the increasing sophistication of design and complexity of structure demands the development of a collaborative method of holistic structural examination somewhere between the two extremes. Moreover, scholars must redefine what it means for a medium to be ‘participatory.’ Indeed, there are some games, both narrative and not, that are much more conducive to meaningful interactive experiences for players, and to better understand the nature of participation in both traditional media and new media, scholars need to focus on these sites for clues. Does culpability play a role? Is agency a consideration, given the medium-specific limitations and affordances that games utilise? A potential area of inquiry that is steadily growing, draws from multiple media formats, and focuses primarily on issues of participation is transmedial studies, particularly transmedial storytelling as Henry Jenkins describes it (1). This avenue offers fertile ground for further examination of the relationship between medium-specific limitations, consumer participation, and the role of agency in interactor-medium experiences.  


[1]The naming of the “start” and “select” buttons is meaningful: once upon a time, “start” actually started the game, while “select” was used to choose different modes of play, like one-player or two-player cooperative.


[2] The soundtrack can be listened to at http://magnatune.com/artists/albums/braid-soundtrack/. It is highly recommended.


[3] It is worth noting that the build I played (March 2013) allowed this; it may no longer be permitted in current builds of the developing game.


[4] At least within the game they should not; games in which questionable practices occur and contact extra-ludic ethics codes, like Morton’s List, must be considered logically in relation to the ethical systems they face outside the game world.


Works Cited


Bethesda Game Studios. The Elder Scrolls V: Skyrim. Bethesda Softworks, 2011. Xbox 360. 

Blow, Jonathan. Braid. Microsoft Game Studios, 2008. Xbox 360. 

Irrational Games. Bioshock. 2K Games, 2007. Xbox 360. 

KCEJ. Metal Gear Solid. Konami, 1998. Sony PlayStation. 

Nintendo R&D4. Super Mario Bros. Nintendo, 1985. NES. 

O’Neill, Will. Actual Sunlight. Steam, 2012. Windows. 

Rockstar North. Grand Theft Auto IV. Take-Two Interactive, 2008. Xbox 360. 

Sonic Team. Sonic the Hedgehog. Sega, 1991. Sega Genesis. 

Ubisoft Montreal. Assassin’s Creed. Ubisoft, 2007. Xbox 360. 

Other sources

Aarseth, Espen. Cybertext: Perspectives on Ergodic Literature. Baltimore: Johns Hopkins UP, 1997. Print. 

Benedetti, Winda. “Indie Game Designer Earns Raves for ‘Braid’.” MSNBC. NBC, 22 Aug 2008. Web. 7 Oct. 2013. http://www.nbcnews.com/id/26336877#.Ul3jgBtNHMo 

Bissell, Tom. “Braided.” Extra Lives: Why Video Games Matter. New York: Pantheon Books, 2010. 91-103. Print. 

Cavalli, Earnest. “Review: Braid.” The Escapist, 21 August 2008. Web. 7 Oct. 2013. <http://www.escapistmagazine.com/articles/view/editorials/reviews/5157-Review-Braid

Chaplin, Heather. “Xbox’s ‘Braid’ A Surprise Hit, For Surprising Reasons.” NPR. NPR, 27 Aug. 2008. Web. 7 Oct. 2013.http://www.npr.org/templates/story/story.php?storyId=94025221 

Douglas, J. Yellowlees. “Response to Markku Eskelinen.” Wardrip-Fruin and Harrigan 35-36. 

Eskelinen, Markku. “Towards Computer Game Studies.” Wardrip-Fruin and Harrigan 35-44. 

Eveleth, K.W. “Interconnected Indebtedness: The Anarchy of Debt in Open-World Games.” Midwestern Modern Language Association, Loyola University of Chicago. Cincinnati, OH. 10 November 2012. Conference Presentation. 

Graphic Gaming. “Let’s Play Braid – INTRO – Part 1.” Online video clip. YouTube. YouTube, 22 May 2013. Web. 8 October 2013. <http://www.youtube.com/watch?v=P69UKoZKIrs

Jenkins, Henry. “Transmedia Storytelling 101.” Confessions of an Aca-Fan: The Official Weblog of Henry Jenkins. Weblog. 4 Oct. 2013. 

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

Kerr, John Finlay. “Review: The Pleasures of Computer Gaming.” Film-Philosophy 13.1 (2008): 165-175.Web.<http://www.film-philosophy.com/2009v13n1/kerr.pdf>McAvoy, Mark. “What Makes A Game?” The Indie Mine. n.p., 23 Apr. 2013. Web. 10 Sep. 2013. 

O’Neill, Will. “About & Reviews.” Actual Sunlight. n.p, n.d. Web. 8 Oct. 2013. 

Quillen, Dustin. “Braid PC Listed for March Release on Impulse.” 1Up.com, 18 Feb. 2009. Web. 7 Oct. 2013.  

OxfordDictionaries.com. O U P, 2013. Web. 13 October 2013.  http://oxforddictionaries.com/us/definition/american_english/culpable

Schechner, Richard. “Response to Markku Eskelinen and J. Yellowlees Douglas and Andrew Hargadon.” Wardrip-Fruin and Harrigan 193-6.Sicart, Miguel. The Ethics of Computer Games. Cambridge: The MIT Press, 2009. Print. 

Wardrip-Fruin, Noah, and Harrigan, Pat, eds. FirstPerson: New Media as Story, Performance, and Game. Cambridge: The MIT Press, 2004. Print.