Invisible Gameplay Participants: The Role of Onlookers in Arcade Gaming

Holin Lin and Chuen-Tsai Sun


The digital gaming experience is not limited to what takes place between gamers and screens, or among players who cooperate with and compete against each another in virtual or physical worlds. In this paper we will explore the roles of onlookers—an important yet often invisible group of game participants—in shaping the gameplay experience in an amusement arcade in Taipei. Instead of examining the specific effect of emotional arousal in players, we will adopt a symbolic interactionist perspective based on Goffman’s work (1973) to analyze how players and onlookers present themselves while negotiating and validating their identities in face-to-face encounters in a public gaming setting. Our main argument is that the game machine and public space characteristics of amusement arcades facilitate specific interactions between the players’ gaming experiences and their social relations. Accordingly, the presence of onlookers and their relative skill levels encourage dynamic oscillation between practice, recreation, and performance. As gaming situations change, interactions between players and onlookers also vary in sophisticated and subtle ways.

Our major research concern is determining how the presence of onlookers affects gameplay. In a public playground—where onlookers are frequently players waiting their turns, friends of players, or passers-by attracted by performance—how do players define situations and act upon or react to them? By moving the gameplay audience and bystanders to a more prominent position, we will attempt to define the roles they play in influencing player behaviors, maintaining the ‘magic circle,’ and shaping the gaming experience.

Literature Review

When investigating the relationship between players and other persons in their presence, researchers have usually focused on co-play relationships or performer-audience connections. Thus, onlookers have generally been placed in the background in favor of elaborating on players’ cognitive processes or describing player agency when dealing with out-of-game interactions during gameplay. We will instead consider the complex and dynamic characteristics of the player-onlooker relationship, based on our beliefs that onlooker roles are multiple and fluid and that clear distinctions cannot be drawn between players and onlookers within the social situation embedded in gaming spaces. In this paper we will mainly employ Goffman’s tactics of observing the behaviors of onlookers as presentations of themselves in social settings. Furthermore, building on the interest shown by previous researchers in shifts in degree of player immersion when an audience of strangers or companions are present, we will look at (a) how player-onlooker social processes facilitate their movement among several states of fun, and (b) how they collectively negotiate gaming norms based on skill levels in public arcade spaces.
Lazzaro (2004) emphasizes the perspective of gaming as a mechanism for players to produce social experiences. Through gameplay, players create opportunities for interpersonal interactions and societal connections. Almost three decades earlier, Mehrabian (1976) pointed out that the sense of fun triggered by interactions during gaming experiences is not limited to players but includes spectators, passers-by, and other outsiders. Group play can certainly encourage new forms of gaming behavior and rituals and add to the players’ sense of excitement, but we must acknowledge the potential for enjoyment among those acting as onlookers as well as between players and audience members.

Previous researchers have used social facilitation theory to examine the effects of audience presence/absence on players, noting that audiences can promote positive performance among good players and decrease successful performance among poor players. Kimble and Rezabek (1992) took game complexity into account and found that when playing a simple game such as pinball, good players frequently performed worse and poor players better under audience pressure; however, when playing a complex game such as Tetris, all study participants played worse regardless of skill level. They suggested that a “social choking” effect is prevalent in such scenarios, that good players are more self-conscious and self-evaluative, that self-attention is an interfering factor during play, and that good players are therefore more likely to underachieve in the presence of an audience.

Social facilitation theories for gamers were derived from theories about performer-audience relationships in traditional competitive arenas or performance theaters. In arguing that social presence by itself is not sufficient for explaining the effects of social context in gaming, Jakobs et al. (1997) considered the effects of social characteristics (e.g., communication potential and opportunities for monitoring performance) on shaping participants’ emotions in certain social contexts. de Kort et al. (2007) further suggest that the ways in which onlookers evaluate players is highly dependent on their relationships with players and their ability to monitor the performance.

Different types of game machines in amusement arcades are open to public performance to varying degrees, which shapes player-onlooker relations. One of the most “open” game types—dancing machines such as Dance Dance Revolution (DDR)—have received a great deal of research attention due to their characteristic of attracting onlookers, which transforms play into performance; player-onlooker interactions are highly visible for this type of game. Smith (2006) states that successful DDR experiences are highly dependent on onlookers, and therefore considers it an example of public demonstration and competition. In arguing that DDR is more than a game, he notes that visitors to game boards and websites divide DDR players into the categories of technical dancers and freestyle dancers—the latter referred to as “real dancers.” While technical dancers are skilled in terms of accuracy, freestyle dancers introduce upper body movements, directly face their audiences, and purposefully express their own artistic styles in order to create extra excitement.

Healey and Light (2007) investigated arcade gaming in the London area—including DDR and Para Para Paradise (PPP), a hand motion-focused dancing game—and observed several known features of expert player communities. They report that players belonging to different levels tend to use the machines at different times of the day, with experts choosing peak hours to ensure the presence of large numbers of onlookers. Each arcade has its own expert player community, with members aware of each other and their performances even if they do not compete at the same time. Other researchers have reported that the social contexts of these kinds of communities go beyond arcades. Andrews (2006) describes DDR players as having their own online meeting places for discussing style, courtesy, and special dance patterns. Very recently, Behrenshausen (2008) analyzed performer-audience negotiation in parallel with identity formation, bodily competition, and meaning making—all based on a clear delineation between performers and audience members.

As previous research usually treats onlookers as “others” who exist outside of clearly marked boundaries, we will purposefully put onlookers at the front of the stage in order to better understand the intersubjectivity that exists between them and players, and use Goffman’s (1973) perspective on the presentation of self to analyze onlookers, just like players, in amusement arcades as performers. We observed their use of signals (cues, texts, hints, expressive gestures, status symbols) in a recognized structure of social encounters so as to avoid embarrassment of both the players and the onlookers, which fits with Goffman’s insight that when individuals enter spaces filled with the physical presence of others, they need to take a character in that social setting and immediately begin to perform. As game genres, machinery, and spaces develop, the presence of special props and costumes can be used to identify how onlookers select characters and move between them. Other, more mobile and transitory ‘sign vehicles’ (which Goffman identified as posture, speech patterns, facial expressions, and bodily gestures) are also indispensable when investigating flexible and fluid relations in amusement arcades.

Data & Methods

Our primary data collection method was participant observation in amusement arcades, supplemented by on-site interviews with players and electronic discussions found on related game boards and websites. Our field of observation was Tom’s World, the most popular arcade in Taipei; it is located in a shopping building primarily aimed at adolescent consumers in the old downtown district of the city. Except in the early hours on weekdays, it is usually filled with players in their teens and early twenties. On weekends it is usually packed from opening to closing, and often presents a more diverse population. The two authors and two research assistants visited this arcade separately during April and May in 2008, at different times on weekdays and weekends. Each visit lasted from 90 minutes to 2 hours, for a total observation time of approximately 26 hours. Since the two authors do not fit with typical player profiles, our appearance was very noticeable during non-peak hours and we always attracted attention from the arcade staff. We therefore limited our visits to peak hours and gave intense observation training to our assistants, who visited the arcade at all other times, and changed our observation notes frequently.

We noted that game machine/interface and genre had a direct influence on the presence and composition of onlookers, and therefore chose five motion sensing music video games as our focus: DDR, Guitar Freaks (GF), Drum Mania (DM), Drum Master (aka TAIKO), and Rock Fever 5 (aka DJ). Our rationales for these choices were two-fold. First, certain game genres require specialized knowledge thresholds in order to appreciate player performances. This characteristic excludes all casual (‘pure’) onlookers—in other words, audiences for these games tend to consist of companions of the current player or of people waiting their turns. Strategy and fighting games also fall into this category.

Second, some games are inconvenient for watching. For example, experienced players of racing games are the only participants who can feel the flow and excitement of play, meaning that onlookers are mostly limited to player friends. Another example is fighting game machines, which occupy small spaces and therefore drastically reduce opportunities for watching. In contrast, motion sensing games are ‘large’ both in terms of machine space and player movement, which explains why they are generally placed in the center of an arcade, where they attract the greatest attention. For these games, player movements are considered the spectacle, rules are simple and easy to follow, and the background knowledge required to appreciate performance is minimal. Onlookers can judge players’ skills from their footsteps, other physical movements, and from the “combo values” indicated on game screens.

There was an essential challenge in our observation: since we wanted to investigate how the presence of onlookers affected players, we needed a situation with no onlookers for comparison. This was not possible due to arcade game design and space layout, even during off-peak periods, since our presence constituted an audience of a certain kind. We took two actions in response. First, upon our arrival we made a special effort to locate players who had started playing in the absence of onlookers, then compared song selections and performances at those times with songs and performances chosen in the presence of an audience later. Second, we used player discussions on game boards and blogs to understand feelings and thoughts during gaming experiences. We also interviewed three players who were waiting their turns so as to gain a better understanding of interaction rules and player reactions to audiences.

Findings and Discussion

Our findings can be summarized as follows.

1. Player and onlooker identities in amusement arcades are flexible and fluid, with few ‘pure’ onlookers. Note that players become onlookers when waiting their turns to play.

2. Relative skill levels between players on stage and onlookers do much to determine the nature of gaming scenarios: performance, practice, or recreation. When on the machine, skillful players can transform a personal recreational event into a performance stage, create an audience consisting of people waiting their turns, and interact with ‘apprentices’ wanting to learn something. When less skilful players take the stage, relationships quickly revert to someone practicing and others waiting their turns.

3. Players and onlookers demonstrate behavior patterns and models corresponding to characteristics of the gaming scenario in question. Each party has responsibility for maintaining the situation.

For convenience of analysis, we divided players into three categories based on skill level and relation to the game community: (a) experts, who appear frequently and express near-perfect skills. Some are considered ‘legends’ in player circles, and achieve high levels of fame in arcades. (b) Apprentices, who also visit arcades on a regular basis, sometimes with friends, and who have some experiences with the game up to medium skill levels. They know who the top players are and which experts have certain specialties. (c) Newbies and outsiders, who feel pressure when playing in front of others, but who are not conscious of invisible interactive rules. More likely to visit arcades on weekends, they often react strongly and outwardly to good performances by experienced players.

In the following we will analyze instances of arcade interaction and self-presentation along several dimensions, and discuss the symbolic meanings given off by these self-presentations. We have placed our reading of the symbolic meanings within brackets []

‘This is my place’

We observed that experts and apprentices alike frequently gave hints expressing an idea considered common to all members of an inner circle: ‘I know this place well.’ The most common way of conveying this message was carrying large amounts of game tokens, yet we never viewed any experts or apprentices approaching staff desks or token machines to get more tokens; those acts were limited to passers-by. When their turns came, experts or apprentices took out their packs of tokens from their bags and displayed them to the machine [I’m a regular, not a passer-by.] Experts regularly earn bonus runs for their performances, meaning that they need fewer tokens, yet we noticed that they still tended to carry large packs of tokens—perhaps evidence of their ‘long-term devotion to the game’ or proof that they are unwilling to break their concentration to complete a task as mundane as purchasing more tokens.

We also noted differences between experts/apprentices and beginners in terms of how they used gaming space. Experts were more likely to place their bags in places that did not interfere with gameplay but that were clearly part of the gaming space—for instance, right beneath the machine. Other times they gave more aggressive signs of their status by leaving their bags near the game even when it wasn’t their turn, only retrieving them when leaving the arcade. [I treat this space the same way I treat my own home.] Casual visitors and newbies always carried their bags, sometimes even when dancing; some asked friends to hold their bags while they played. If they did place their bags in a public space, they rushed to retrieve them once the game ended.

We also observed experts and apprentices sharing some local knowledge with other members of their insider circles. By sharing such knowledge, they demonstrated their ongoing relationships with the gaming space. For example, we saw a medium-level player stepping on one corner of a DDR platform while his friend was playing, declaring ‘this helps to stabilize the machine and makes it easier to dance’.

Strong awareness of the presence of spectators

With the exceptions of experts, we noticed that all players were very aware of the presence of an audience. Due to the large number of machines in an arcade, it is not easy for players who are facing the screen to determine whether the people behind are their audiences. Almost all non-expert players took a quick glimpse before starting a new run to confirm the direction of onlooker gaze, and perhaps to determine whether or not they were being watched by experienced players. Once the presence of an audience was confirmed, they quickly withdrew their glimpses and focused on their performances, giving the impression that they were neglecting their audiences. No further eye contact took place until the end of the round, when another glimpse confirmed the presence or absence of onlookers.

The general atmosphere around a motion sensing game is one of performance, thus giving off a holistic feeling that ‘everyone is good at dancing (or drumming, or playing)!’ Although experts and beginners mix while waiting to perform, poor players usually falter and only occupy the stage for a short period of time—often without attracting much attention. In contrast, experts consistently earn bonus songs, thereby making themselves highly visible, giving beginners the feeling that ‘we are surrounded by good players,’ and heightening the general awareness of skill level and audience-performer contexts.

In contrast, experts consistently gave the impression that they had arrived out of thin air—jumping onto the platform, not glancing at onlookers, selecting songs decisively, dancing immediately on the first note, and leaving the moment the game ended often with no evidence of physical exertion. [I’m just passing by and have quick fun, this is my effortless performance, I don’t care if there is audience or not.] They neither hesitated nor made eye contact with onlookers. However, instead of interpreting this as evidence that they were unaware of their audiences, it is more like that they are pretty sure whenever they are on stage, there will be an audience. On one occasion, we noticed that during the performance of a pair of players, four high school student onlookers left the audience of eight. Even though they left quietly, one player turned his head and give a quick glance toward those leaving.

Presentation of status: Choosing a song

Arguably the best example of how the presence of onlookers affects gameplay is in song selection by apprentices. The primary concern for any insider player is deciding between songs for practice and songs for performance. Players cannot progress if they don’t choose more difficult songs for practice, but choosing complex and unfamiliar songs sharply increases the potential for failure. According to our informers and website postings, regular players frequently download software to their own computers so as to familiarize themselves with the rhythms and step patterns of available songs, and then practice in private before testing themselves on arcade machines. We found that the appearance of an audience influenced many apprentices to deemphasize practice in favor of performance by selecting songs they were very familiar with and/or songs with better performance impact. In such situations, they refrained from picking the ‘most difficult’ songs or challenging themselves by trying something new. The general rule we found is - selecting a song as difficult as possible, which still allowed the player to give a perfect performance. On one occasion we watched two junior high school students playing DJ in front of some onlookers; they stuck to safe songs at levels 4 or 5 (with 9 the highest level of difficulty) and performed perfectly. They only felt comfortable experimenting with songs at higher levels when no one was around to witness their errors.

In another situation, we were able to watch two DDR players, both of whom gave failing performances on a song they had chosen upon our arrival. A third player appeared about the same time as we did and took a turn at the machine, after which the three of them took alternate turns and traded perfect performances. Our assumption is that the first two players were practicing dances to some unfamiliar songs, then switched to performance mode when an element of competition was added.

Audience influence on song selection does not apply to experts and newbies. To the untrained eye, an expert’s practice and performance dances are indistinguishable when done perfectly. And whereas apprentices can choose songs to match different situations, newbies do not have new-and-unmanageable versus old-and-familiar alternatives available to them. In situations where two consecutive players are in direct competition, there is greater chance that the second in line will pick more challenging songs if the first has played well. However, restraint is a valued quality among players. On one game board someone criticized ‘arrogant’ experts for selecting the exact same song used by the previous player but with much higher degree of difficulty, [I used your song to beat you, I’m much better.] Our impression is that these ‘arrogant experts’ are considered distasteful by the general arcade gaming community.

One DDR apprentice posted his embarrassing experience on a game website. When his turn came and he started to play, the two experts who had immediately preceded him did not leave (as is usually the case), but stayed to watch. Shortly after selecting his second song and starting to dance, he noticed that the two experts were laughing, which unnerved him. The two experts later replied to his post, explaining that they had predicted which song he would choose next, i.e., to repeat the first song but of the highest level of difficulty, and were laughing because their prediction was correct. They apologized for affecting his performance, adding, ‘One should choose new songs on a new machine. We are all DDR fans, no one will make fun of you =ˇ=.’

Speed in song selection is another signal of skill and status. One clear mark of a newbie is hesitating to pick a song; they tend to listen to short samples of many songs one by one [I’m not good at this]. In one case, after sample-listening to songs, a player eventually selected the same song he played just a few minutes ago. During prolonged selection processes, pure onlookers tend to leave to visit other parts of the arcade.

Presentation of status: Performance

In an amusement arcade, all visitors want to play their roles well. When true newbies are on stage, their awkward movements and repetitive errors make the gameplay atmosphere one of pure recreation. When apprentices perform, their body language and non-verbal signals reveal nervousness and self-criticism: constant sighing, head shaking, tongue spitting, and occasional glimpses toward onlookers. Apprentices constantly alternate between practice and performance. When a performance in front of an audience is less than perfect, their body language is one of apologizing to onlookers, [I know I’m not doing great; I’m actually much better than this.]

The situational feature of gameplay among experts is quite clear: except for those uncommon situations when others at the same master level are present, almost all expert performances are given in front of onlookers who are less skilled. As with many performers, they tend toward using exaggerated movements to enhance the effects of performance, to show off their talents, and to create new challenges. Individual experts have their own tricks for enhancing their performances: to show a lack of need to see the game screen they may close their eyes (even lower their heads), use the DDR ‘invisible’ mode, or dance with their backs to the screen. We even witnessed one master player making small talk with a friend while performing at the highest level of DDR difficulty, [I have memorized the whole song.] Even when they are practicing, experts are concerned with performance. For example, although PPP scores only count on hand movements, we observed several full-combo PPP players practicing their footsteps again and again, to make their performance even more spectacular.

The functions of companions

Companions can enhance enjoyment as cheerleaders, provide cues for play, and reinforce the players’ sense of achievement. In addition, we found they also serve the functions of conveying spoken and unspoken messages to pure onlookers, and alleviating performance pressure. Newbies frequently visit arcades in groups to overcome the sense of unfamiliarity and embarrassment they feel. On one occasion we witnessed a group of junior high school apprentice DJ players. One female played alone, with a male friend hitting the set of buttons next to her, even though they were not activated. When the boy started to leave, the girl cried out, “Don’t go, or I’ll lose.” The boy had done nothing to help her, but by asking for help the girl sent out a message to the onlookers [I’m not really good, don’t expect a first rate performance]. In contrast, companions provide another means beside body languages for experts to show off their status: after a performance an expert may linger on the DDR performance floor, leaning on the machine handles and chatting with companions.

Waiting for one’s turn

Rules for lining up to play a game are complex, unspoken, and difficult for outsiders to decipher. Because of spatial limitations in arcades, actual waiting lines are impractical. “First-come, first-serve” is the only principle that we could clearly observe; all other guidelines are achieved via player consensus. While quarrels are not unheard of, they are amazingly unusual for such a complex, noisy environment operating according to ambiguous rules. Getting access to the most popular games requires considerable patience: 30-minute wait times are common during peak periods.

It is difficult but crucial to distinguish between real audience and waiting players among onlookers. Some interactions can be explained through this angle. For some games, placing a pile of tokens near the token slot is sufficient for holding one’s place; for DDR, players can pre-insert tokens to the machine for the same purpose. Again, restraint is appreciated: aggressive acts of token pre-insertion that might affect the current player are frowned upon, especially when a game is in progress.

On the other hand, there is considerable flexibility in queuing behavior, which involves negotiation of playing rights. One guiding principle is the acknowledgment that experts are capable of giving performances that everyone can enjoy, therefore queuing rules are sometimes set aside when an expert is on stage. At such times, waiting players become part of the audience. Another unwritten rule seems to be that no one should place a token on a machine when an expert is playing.

We noticed the use of body language to send signals about one’s status as a ‘dedicated audience member’ or ‘indifferent waiting player’. During high-level performances, common apprentice behaviors include staring at the performer and screen, keeping rhythm with head or foot movements, playing ‘air drums’, taking pictures or videos, and giving silent applause when the performance is over. When the current performer is at a lower level the waiting apprentices eliminate any behaviors that might signify them as audience members and instead express behaviors that mark them as waiting their turn or being more skilled. One way of showing indifference to a performer is talking to friends about the best moves for the current song or directly commenting on the current performance.
When waiting for lower-status players to leave the stage, apprentices have ways of expressing the idea of ‘being there, but not making the effort to be an onlooker’. They include watching players of other arcade games, playing with their Rubik’s Cubes, portable game consoles, or mobile phones. By doing so, onlookers remind the player with visible signals that ‘we are not watching your performance, we are just waiting for our turn to play,’ and at the same time send out message to other potential players that ‘I have been waiting, don’t cut in line.’

Newbies and lower status apprentices give many signals that they are not comfortable on stage. They spend a great deal of time making sure no one is waiting behind them, constantly looking around to see who is watching, intentionally leave the platform after one round, and wandering through the arcade until they are certain that no one is waiting before returning for another round of play. After repeating this pattern several times, one DM player moved toward the exit, seeming to leave the building. At the last moment he took one more look at the area and returned to play another round. Such actions seem to be aimed at luring ‘waiting players’ to present themselves. Only when they are certain that their audiences consist of pure onlookers are they willing to reclaim their positions as current players.


Social interactions among gamers and onlookers in amusement arcades are complex and dynamic. Onlookers play an important role in gaming enjoyment and understanding the nature of gameplay. The presence of onlookers, their behaviors, and their relative skill-based rankings shape the magic circle around game platforms. Onlookers assist in enhancing and expanding gameplay and magic circle boundaries when they accept the roles of focused audience members or learning apprentices, thereby turning play into public performance. On the other hand, they can also use their repertoires of gestures to highlight their positions as disinterested players waiting for their turns, thereby shrinking the magic circle and redefining gameplay as an activity limited to the current player and the machine.
On-stage gaming activities are the products of social interactions between players and onlookers rather than personal player behaviors. Based on their understanding of relative skill and status levels, players and onlookers negotiate rights of play, what to play, and how to play. In a similar manner, being part of an audience is much more than merely ‘being there’. Through the use of gestures and body language, onlookers shift between their roles as concerned audience members and disinterested players waiting their turns.


Andrews, G. (2006). Land of a couple of dances: Global and local influence on freestyle play in Dance Dance Revolution. Fibreculture, 1(8).

Behrenshausen, Bryan G. (2008). Toward a (Kin)Aesthetic of Video Gaming: The Case of Dance Dance Revolution. Games and Culture 2(4), 335-354.

de Kort, Yvonne A. W., Wijnand A. IJsselsteijn and Brian J. Gajadhar (2007). People, Places and Play: A research framework for digital game experience in a socio-spatial context. Presented at DiGRA 2007 Conference.

Goffman, Erving (1973). The Presentation of Self in Everyday Life. Woodstock, New York: The Overlook Press.

Healey, Patrick G. T. and Light, Ann. When Scoring doesn’t Matter: The aesthetics of performance in arcade games. CoDesign, 2007, Vol.3. Supplement, pp.91-96.

Jakobs, E., A. Fischer and A. Manstead (1997). Emotional Experience as a Function of Social Context: The Role of The Other. Journal of Nonverbal Behavior 21(2), 103-130.

Kimble, Charles E. and Jeffery S. Rezabex (1992). Playing Games Before an Audience: Social Facilitation or Choking. Social Behavior and Personality, 20(2), 115-120.

Lazzaro (2004). Why We Play Games: Four Keys to More Emotion Without Story.

Mehrabian, A. (1976). Public Places and Private Spaces: The Psychology of Work, Play and Living Environments. New York: Basic Books.

Smith, Jacob (2006). Digital Dance Hall: The Fan Culture of Dance Simulation Arcade Games. In Kenton ÓђةHara and Barry Brown (eds.), Consuming Music Together. Springer Netherlands.