The Daedalus Project

Applying Psychology to MMORPGs: Automatic Mimicry

Research in social and organizational psychology has consistently shown for many years that when people or groups interact, many of their verbal and non-verbal cues become synchronized almost immediately. For example, the timing of gestures becomes synchronized (Kendon, 1970), and group members mirror each other’s posture and mannerisms (LaFrance, 1982; LaFrance & Broadbent, 1976). Individuals in a conversation will also mirror each other’s accents and speech patterns (Cappella & Panalp, 1981), and syntax (Levelt & Kelter, 1982). In fact, many human behaviors seem to be contagious, such as yawning (Provine, 1986), laughter (Provine, 1992), and even moods (Neumann & Strack, 2000). Researchers have suggested that this synchronization is an automatic human behavior that functions as a regulator of trust and rapport in social interactions (Kendon, 1970; LaFrance, 1982).

Recent research has demonstrated more precisely that when people interact, they in fact unconsciously mimic each other’s behavior. In one study (Chartrand & Bargh, 1999), subjects interacted with a confederate (a research assistant who pretends to be another subject) in a collaborative task. The confederate performed a series of movements (foot shaking and touching their face) and it was found that subjects would unintentionally match those behaviors themselves. More importantly, in a different part of that study, confederates were asked to either mimic or not mimic the subject’s behaviors and it was found that subjects judged confederate mimickers as more likeable than confederate non-mimickers.

Instead of merely influencing attitudes, automatic mimicry has also been shown to influence observable behaviors as well. For example, waiters who verbally mimic their customers’ orders (by repeating the order) receive bigger tips than when they say something else instead (like ‘Coming right up’) (van Baaren, Holland, Steenaert, & van Knippenberg, 2003). In fact, when a person is mimicked, they become more generous not only towards the mimicker, but to everyone else in general (van Baaren, Holland, Kawakami, & van Knippenberg, 2004). Mimicry increases an individual’s prosocial behavior. This process also happens the other way. Affiliation goals increase the frequency of automatic mimicry in interactions (Lakin & Chartrand, 2003).

====

This has led researchers to hypothesize that automatic mimicry is an evolutionary adaptation to facilitate and express social affiliation and that the process is bi-directional - mimicry facilitates affiliation and prosocial behavior and affiliation goals increase mimicry (Lakin, Jefferis, Cheng, & Chartrand, 2003). This theory is also supported by studies that have shown that very young infants will mimic many facial expressions they perceive, such as sticking their tongue out, smiling and opening their mouths (Meltzoff & Moore, 1977).

Virtual environments in fact provide a perfect setting for embedding subtle mimicry behaviors in NPCs because details in the environment can be rendered differently for each user. The goal of embedding mimicry would be to increase prosocial behavior in general in the community. After all, loyalty and bonds with other players is what keeps players in a community.

Examples of this embedding range from the simple to the complex. Of course, the following are not meant to be employed with every single NPC interaction, but instead used intermittently to seed prosocial behavior.

- Align the NPCs appearance with the character’s appearance. Match their hair color, their clothing style, or the weapon they are carrying.

- Match the first letter of the NPCs first name with the first letter of the character’s first name.

- Store the user’s style of greeting other users by matching with a small database of known greeting words, such as ‘hi’, ‘hey’, ‘what’s up’, and so on, and have the NPC greet the user with the appropriate words.

- Store the verbosity of users’ exchanges with other users. Laconic users prefer laconic NPCs and verbose users prefer verbose NPCs - it functions as an approximation for personality differences.

The often-assumed freedom that comes with virtual worlds is a double-edged sword. In the real world, laws constrain behavior, but in virtual worlds, code dictates behavior. If shouting is not allowed in virtual worlds, then you cannot shout in public spaces. You can communicate with other users only through tools provided by the virtual world. In a strange way, relationships in virtual worlds are not created as much as engineered by the mechanics of the world. As these environments evolve, they might - for better or worse - become tools of social engineering that were never imagined even possible in the real world.

====

References

Cappella, J., & Panalp, A. (1981). Talk and Silence Sequences in Informal Conversations. Interspeaker Influence. Human Communication Research, 7, 117-132.

Chartrand, T. L., & Bargh, J. A. (1999). The chameleon effect: The perception-behavior link and social interaction. Journal of Personality & Social Psychology, 76(6), 893-910.

Kendon, A. (1970). Movement Coordination in social interactions. Acta Psychologica, 32(2), 101-125.

LaFrance, M. (1982). Posture Mirroring and Rapport. In M. Davis (Ed.), Interaction Rhythms: Periodicity in Communicative Behavior (pp. 279-298).

LaFrance, M., & Broadbent, M. (1976). Group Rapport: Posture Sharing as a Nonverbal Indicator. Group and Organizational Studies, 1, 328-333.

Lakin, J. L., & Chartrand, T. L. (2003). Using nonconscious behavioral mimicry to create affiliation and rapport. Psychological Science, 14(4), 334-339.

Lakin, J. L., Jefferis, V. E., Cheng, C. M., & Chartrand, T. L. (2003). The chameleon effect as social glue: Evidence for the evolutionary significance of nonconscious mimicry. Journal of Nonverbal Behavior, 27(3), 145-162.

Levelt, W., & Kelter, S. (1982). Surface form and memory in question answering. Cognitive Psychology, 14(78-106).

Meltzoff, A., & Moore, A. (1977). Imitation of facial and manual gestures by human neonates. Science, 198(75-78).

Neumann, R., & Strack, F. (2000). Mood Contagion: The automatic transfer of mood between persons. Journal of Personality & Social Psychology, 79, 211-223.

Provine, R. (1986). Yawning as a stereotyped action pattern and releasing stimulus. Ethology, 72, 109-122.

Provine, R. (1992). Contagious Laughter: Laughter is sufficient stimulus for laughs and smiles. Bulletin of the Psychonomic Society, 30(1-4).

van Baaren, R. B., Holland, R. W., Kawakami, K., & van Knippenberg, A. (2004). Mimicry and Prosocial Behavior. Psychological Science, 15(1), 71-74.

van Baaren, R. B., Holland, R. W., Steenaert, B., & van Knippenberg, A. (2003). Mimicry for money: Behavioral consequences of imitation. Journal of Experimental Social Psychology, 39(4), 393-398.

In Search of Fan Videos

One of the easiest ways to explain the complexity of what occurs in MMORPGs is through fan videos (especially to non-gamers). They offer compact, visually engaging distillations of what would take many words to explain. To that end, I am trying to compile a set of high quality fan videos (in terms of content and video resolution) that document the many interesting social phenomena that emerge in MMORPG environments.

Please email me links to videos with the subject "MMORPG Fan Videos". Submitted links can be in the following categories or any particularly meaningful videos you have come across. I will then include them on this page for other readers and for reference.

Submitted video links may be raw footage or edited footage, but in either case should have decent resolution. Submitted links should be to videos with both good content and visual appeal. The goal is not to create an exhaustive list, but to have a selection of great fan videos.

1) Memorials / Tributes. Footage of memorials or tributes to real life events or real people.

- Teletha's Memorial (DAoC)
- Remember (DAoC)

2) Combat/ Raid / RvR. Footage that showcases the complexity of the organizational and leadership skills needed in a large group combat or large scale raid.

- Drakulv Executioners (DAoC)

3) Funny Moments. Footage of hilarious moments or occurrences in the world.

4) Scripted Play / Music Video. Edited footage that uses the MMORPG as a stage for a music parody or scripted play.

- Ice Ice Baby (SWG)
- Go Beyond (E&B) - use DivX player to open
- Justice (CoH)
- Come As You Are (DAoC)
- Drunken Scottsman(DAoC)
- Has Anyone Seen My Corpse (EQ)

5) The Complexity of Play. Footage of how game play involves a lot of work. Crafting professions (such as in SWG) would be great examples.

- Building a Pyramid (A Tale in the Desert)

Addiction

About 40% of users would consider themselves addicted to the game. This percentage is quite striking given the social desirability bias built into the question. There were no gender differences, and age differences were minimal - about 2 years between “definitely” and “definitely not”. In other words, male and female users of all ages are about equally likely to consider themselves addicted to the MMORPG they are playing.

The average hours played per week in each of the five responses offer perhaps a rough guideline as to how much game play is too much game play. As a comparison, typical Americans watch about 27 hours of TV per week.

There were interesting correlations between this self-reporting of addiction and motivations. In particular, self-reported addiction was positively correlated with scores on Achievement (r = .25) and Escapism (r = .25). This finding dovetails with an earlier discussion of addiction. It appears that both internal (skinner box model) and external factors (compensation model) contribute to addictive behavior.

Note (12/02/2006): This data is complicated and can be misleading for several reasons. First of all, there is no accepted definition of "online gaming addiction" to begin with, so it's not clear what players meant by that (or why we would trust people to diagnose themselves even if that were the case). Secondly, it's not clear they didn't just take the statement to mean "the game is very fun" as the word is used commonly in the game industry. And finally, we have no comparison data with other hobbies - i.e., If we asked golfers that question, how many would have said "yes"? In other words, could it be that 40% of users of all hobbies would say they are "addicted" to their hobby (whether that is golf, knitting, or watching the TV show "Lost").

See Also (listed in chronological order):

- Understanding MMORPG Addiction
- The Seduction of Achievement
- Problematic Usage
- A New Disorder is Born

The Blurring of Work and Play

The boundaries between work and play have become increasingly blurred due to digital media technologies. Take web cookies for example. Companies such as DoubleClick.com use hidden cookies to build in-depth profiles of individual users as they surf the web:

Over a period of time DoubleClick compiles a list of which member sites the user has visited and revisited, using this information to create a profile of the user's tastes and interests. With this profile in hand the DoubleClick server can select advertising that is likely to be of interest to the user. (from http://www.w3.org/Security/Faq/wwwsf2.html)

In other words, as individuals browse the web, they become both content producers and consumers. More importantly, users are performing this work under the guise of entertainment, producing information with economic value but which they derive no profit from. Users are in fact performing free labor.

The same occurs with TiVo. As Andrejevic (2002) has pointed out, as TV watchers use TiVo, they reveal their tastes to the system which in turn caters more and more to them - a “commercial paradigm of interactive media as a means of inducing viewers to view more efficiently”. This form of work benefits media companies at the expense of individuals. After all, in a strange way, consumers are paying TiVo while performing free labor.

The blurring of work and play in MMORPGs has three primary forces - economic profiteering from the sales of virtual goods, the growing resemblance of play to real work, and the embedding of real work into these environments.

====

Sale of Virtual Goods

As Dibbell (2003) describes, virtual items and property in online environments such as Ultima Online have economic value, and users accumulate these items to sell for real money on Internet auction sites such as eBay. The transformation of gaming activity into economic activity is most striking in the presence of companies that pay teenagers in developing countries to “play” these games for 40 hours a week and derive a profit from selling these virtual goods for real money (The Walrus, 2004). This transformation of play into work is driven by entrepreneurial users who see the opportunity for bringing economic activity into the game, but what is more intriguing is how the increasing complexity of play in these worlds is coming to resemble real work.

The Growing Resemblance of Play to Real Work

Most MMORPGs base their core mechanics on operant conditioning, a system of rewards that increases frequency of a behavior. The rewards cycle is a random ratio schedule (exactly like casino slot machines) where a reward is given every x times an action is performed, where x increases exponentially as the user progresses. Simple actions are gradually replaced by complex, time-consuming actions and typical users are essentially trained to spend on average 22 hours per week in these environments. This is striking given that the average MMORPG user is 26 years old and about 50% of MMORPG users work full-time. As some users note,

But for every user who has burned out from the treadmill, there are thousands of users who are still “playing” industriously. One user articulates the seduction of achievement in these environments.

====

In other words, these online environments are structured such that they reward and seduce you to perform complex, tedious tasks.

While early MMORPGs focused on combat-oriented advancement, recent MMORPGs have provided non-combat advancement options, so users can now choose to become chefs, hair-stylists, architects, pharmaceutical manufacturers, and yes, even politicians in certain MMORPGs. The example of pharmaceutical manufacturers (PMs) will be used to illustrate the seriousness of play in the MMORPG Star Wars Galaxies.

To become a PM, a user must locate raw chemicals and minerals on the planet using geological surveying tools. Different resources have different attributes which contribute to the final quality of the product as described on each product schematic. To produce high-quality products, users must therefore find the best combination of resources through a time-consuming planet-hopping process. Resource gathering is only meaningfully performed using industrial harvesters bought from architects (other users). Even then, resource gathering is slow and users must check their harvesters on a daily basis. Raw resources are combined into sub-components in factories (also bought from architects), and then combined into the final product. Because each factory run take a few hours of real time to complete and because most products require several sub-components, users have to plan out their production chain fairly well to ensure their machinery doesn’t idle and drain their capital. To complicate matters, resources are randomly replaced on a weekly basis, so users have to constantly survey for new resources as well. Finally, users must market and sell these goods on the open market, which means competing with other users who are selling similar products.

The time required to acquire the expertise and capital to become a PM in Star Wars Galaxies is in the order of 4-6 weeks of normal game play, and thereafter requires sustained daily time investment to maintain the business. The irony is that users are paying to perform what increasingly resembles serious, complex work in these environments.

====

Embedding Real Work into MMORPGs

Given that MMORPGs are creating environments where complex work is becoming seductively fun, how difficult would it be for MMORPG developers to embed real work into these environments? In fact, this is already occurring in There.com – a virtual world in a contemporary setting. Fashion companies pay There.com to embed test products into the environment where they can track user purchases and how often they wear the garment. More importantly, because the social network of every user can be mapped, it is also easy to spot who the trendsetters are. The information of how likely trendsetters purchase and wear a test product provides highly valuable information to the fashion companies. Of course, the irony is that There.com users are paying to work for a third-party company, with both sources of profit going to There.com.

This can be taken one step further because we know that users are willing to perform complex, tedious tasks in these environments. In fact, they have been trained to have fun performing these tasks. Consider the fact that cancer screening is routinely out-sourced to India because it is relatively cheap to train a lay-person to identify suspicious patterns on a diagnostic scan, and it is cheaper for several dozen of these workers to look at a single scan than it is to have a doctor in the US look at the same scan. Moreover, the accuracy rates are actually better because the redundancy lowers the rate of misclassifications. MMORPG environments could easily tap into their free labor pool of dedicated users by embedding real world tasks into the “game”. What is clear is that there are many different ways in which real work can be embedded into MMORPGs – different ways in which game developers can seduce users to pay to perform free labor.

====

Conclusion

Considering all the things that happen in these worlds - weddings, political elections, sales of virtual real estate for real money, genocides, and teenage mafia gangs and prostitution rings - it seems strange that some people, both gamers and non-gamers, still say that “it’s just a game”. Ironically, the most appropriate reply might be “No, it’s just work”.

References

Andrejevic, M (2002). The work of being watched: interactive media and the exploitation of self-disclosure. Critical Studies in Media Communication.

Dibbell, J. (2003). The Unreal Estate Boom. Wired 11.01, January, 2003. Available at: http://www.wired.com/wired/archive/11.01/gaming.html

The Walrus. (2004). Game Theories, from http://www.walrusmagazine.com/04/05/06/1929205.shtml

It's A Matter of Perspective

In a sample of 2200 respondents across several MMORPGs (29% EQ, 23% SWG, 12% DAOC), the following gender difference emerged as to preference for 1st or 3rd person perspective. The answer choice "prefer both equally" is excluded from this brief presentation for clarity.

While there were differences between games (i.e. more EQ players preferred 1PP, and more DAOC players preferred 3PP), the gender difference was always present in every game. Exploring the data by age groups also revealed a similar pattern.

Because female players tend to be older than male players, it is possible that the above two graphs may be showing the same underlying factor. To show that age and gender are in fact impacting preference for perspective independently, the 1PP case is shown below split by gender and age groups. Women always prefer 1PP across all age groups.

====

Past data had suggested that gender differences are driven by different motivations for participation. In very broad strokes, female players are more drawn to relationship-oriented activities while male players are more drawn to achievement-oriented activities.

The perception and use of an avatar - as the primary means of agency in online environments - might be expected to be shaped by the motivations for participating in the environment. In particular, goal-oriented users may be more likely to treat avatars as tools/pawns to achieve goals, thereby encouraging a preference for 3PP that objectifies and externalizes the avatar, whereas relationship-oriented users may be more likely to treat avatars as representations of themselves in a social environment, thereby encouraging identification and treating the avatar as the self through 1PP. This would also be supported by the age differences given that younger players tend to be more achievement-driven. In other words, I argue that more fundamental motivational differences are driving the gender and age differences.

To test this line of reasoning more directly, users who preferred 1PP vs 3PP were compared on their motivations for playing based on an assessment derived from a previous study. Users who preferred 3PP scored higher on Achievement (t = 5.5, p < .001) and Grief (t = 8.5, p < .001), and lower on Relationship (t = -8.0, p < .001) than users who preferred 1PP, which supports the hypothesis.

To tease apart the relative importance of age, gender and the motivations, a logistic regression was performed using 1PP/3PP as the categorical predicted variable. The Relationship motivation emerged as the most significant predictor (t = 7.7, p < .001), followed by age (t = 6.2, p < .001), Grief (t = -5.0, p < .001) and then gender (t = 2.38, p = .002).

Thus, it appears that the observed gender difference is being driven by underlying motivational differences between users who play to form and sustain relationships and users who objectify the environment and other users for personal gain. In either way, what is clear is that motivational differences are linked with preferences for perspectives in these environments.

While causality can't be directly inferred from this data set, the opposite claim that default (or fixed) perspective shapes motivations for playing doesn't easily explain the observed gender differences.