Tanks, Chauffeurs and Backseat Drivers: Competence in MMORPGs

In this paper learning and competence in the MMORPG World of Warcraft are explored. In order to facilitate movement between in-game and the real-world contexts of play, data was collected from couples who play the game together while sharing real space. Through the collection and analysis of interview data the authors develop a framework for the examination of learning practices. The ways in which players acquire and assess skills, balance different skill levels, and accommodate different play preferences, are discussed. It is argued that competence in MMORPGs is complex, variously constituted and assessed by players in diverse ways.    

Through analysis of interview data we arrived at the notion of the 'management of resources' as a framework through which to consider learning practices, and we identified three categories of resource: ludic, social and material.It became apparent that learning in online games involves multiple competencies, and that competence is assessed in a variety of ways.Furthermore it became clear that managing both relative expertise and increasing competence within social contexts calls for tact and negotiation.
This research into learning in MMORPGs was one element of a small project titled 'Learning from Online Games; Teaching in Second Life' 1 .The aims of the project involved investigating the learning that happens informally in online worlds; developing an improved understanding of online worlds as a context for learning, and exploring the implications of such learning for education.
In this instance, we were interested in World of Warcraft (in contrast to social worlds such as Second Life, for example), because the MMORPG's rules and generic structures mean that there is a 'curriculum' of sorts -an incrementally complex array of skills, competencies, jargon and conventions that players need to acquire in order to progress.Play involves working with existing structures (such as the rules), but there are also less determined pleasures, including exploration, experimentation, humour and chat.In addition to questing, group missions and 'grinding' (i.e.repetitive collecting) there are various styles of game-play on offer: from dramatic in-character role-playing, to player-v-player contest in 'capture the flag' style battlegrounds, as well as highly coordinated, large scale 'raids'.Each of these forms of participation involves the acquisition and performance of particular skills.Eladhari and Lindley, 2003;Krzywinska, 2006;Mortensen, 2006).Online player culture has been described through in-depth ethnographic research (Taylor, 2006).Game-play has been examined through detailed, quantitative research (see Ducheneaut, Yee, Nickell, and Moore 2006).Our research is informed by these studies, but in this instance our particular focus is learning and competence.World of Warcraft has a training level and the game incorporates effective tutoring.As such it recalls 'games and education' literature more generally (such as Gee 2003).Additionally, it has been demonstrated that social aspects of MMORPGs, especially guilds, are important for mentoring and support (Steinkuehler 2005, Galarneau 2005).
We recognise the importance of guilds, but we were keen to devise a research exercise that would, from inception, involve movement between online and offline realms.In this we were motivated by a methodological dilemma recognised and debated by the digital games research community.In such debates, research that produces analysis of 'the game itself' tends to be framed as antithetical to that which focuses on players and the contexts of play 2 .We wished to address play and players while also recognising structural features of the game.We wanted to accommodate shifts between game, play and context as (or if) they emerged, and to collect data in such a way as to support such shifts.For this reason, we decided to focus on couples who play World of Warcraft together, while sharing real space.

Methods and strategies
In order to gain an insight into what players learn, and how, we chose to use interviews as a means of data collection.The global player population of this online game is around nine and a half million (http://www.mmogchart.com/Chart11.html).The player population is split across different servers.Servers are differentiated by time zones, language, and play preference (role-playing, player-v-player, player-venvironment).For the sake of practicality, we collected data on various English language European servers.We created new characters in order to meet interviewees at a location of their choice (server, territory, etc.).We interviewed, for example, a pair of undead warlocks in a hotel in the village of Brill, on a PVP/RP server.In the course of logged, semi-structured interviews, lasting between 60 and 90 minutes, we collected material relating to preferred play styles, the gaining and sharing of expertise, in-world social life and guilds.In the quotations included in this paper, both players and avatars have been anonymised, and in some cases minor alterations to spelling or grammar have been made for the sake of clarity.In the sections that follow, short descriptions of game-play and game terminology are mixed with analysis.
Prior to conducting interviews, we played World of Warcraft on a regular basis for approximately 18 months.This engendered a first-hand familiarity with the game world, the various forms of play on offer, the pleasures of the game, and aspects of its player cultures.This familiarity informed the research design and the data collection process.To recruit interviewees, we drew on our own in-game and real world social ties, recruiting couples we had met through our guilds, and couples who play that we knew in real life.Two initial interviews within the project team and their families provided a starting point for analysis.These were gradually augmented by interviews with 'contrasting' couples in order to incorporate a degree of breadth and variety in terms of the nature of the relationship, player preference and access issues.We eventually interviewed 10 players over a 4 month period, by which point patterns were emerging from the data.
Our approach to data collection, data analysis and emerging models was informed to some extent by the 'grounded theory' of Glaser & Strauss (1967).However our practice is much closer to the constructivist version of grounded theory, such as that proposed by Charmaz (2003).Charmaz describes 'a revised, more open-ended practice of grounded theory that stresses its emergent, constructivist elements' (p 251).Within this approach it is acknowledged that 'the categories, concepts, and theoretical level of an analysis emerge from the researcher's interactions within the field' (p 271).Constructivist grounded theory, as described by Charmaz, involves a reflexive approach to data collection (it may include a return to the field over time, for example), and recognition of the central role of the researcher in shaping the data collected and making sense of it.
The resulting data was reviewed, which involved independently reading transcripts and coding coherent excerpts (either statements or short exchanges that formed selfcontained points or claims).These independent codes were then jointly reviewed, and particular themes were collated under headings.
Through an engagement with and analysis of this material we developed a framework through which to consider learning practices.This involved looking at increasing competence in relation to the management of resources, where 'management' involved the recognising, negotiating, accessing and applying of 'resources', which might be categorised as material, ludic or social.In actuality these three classes of resource constantly intermingle, yet these provisional distinctions allow us to gesture to specific examples of learning practice, and to the complexity of competence in this context.

Material resources
Material resources involve the physical, situated and embodied aspects of play.The material or real world resources to be managed during play included practical and logistical issues, from 'who gets the best computer' through to who is seated in the more comfortable chair.It also might include a couple negotiating childcare.Gender did emerge as a factor here, as it was generally our female interviewees that expressed consciousness of these issues; who mentioned playing on the laptop rather than the 'best computer', or playing while juggling parenting responsibilities.
In one instance a lack of resources (financial) led to a couple sharing an account, which meant that they were unable to log-in at the same time.This impacted on their game-play, as the role of helper-spectator was expanded to the point that the line between playing together and playing solo was blurred.As this suggests, the game has rules but it remains to some extent customizable.Of course, in addition to communicating within the game world via text and/or voice, couples who play while sharing a real space have the option of speaking to one another directly.The interviewees described flitting between private and public modes, and between ingame text chat, in-world voice chat, and real world conversation.Languages might also be switched: one couple tended to use English in the game, German and Czech in real space.
The two 'worlds' occupied by a player continually blend and mingle.Interviewees moved between references to real world factors and in-game issues when describing, for example, the life-cycles of their guilds.Real life friends became guildmates, who might cease playing due to real life commitments, to be replaced by online friends, who eventually met physically.Time was another resource to be managed, from couples' negotiations as to what constitutes an appropriate investment of time in the game, to players' recognition of the time-consuming nature of guild commitments.Interviewees referred to play interfering with life, and to life interfering with play.

Ludic resources
As is generally the case with role-playing games, players in World of Warcraft begin by creating a character while choosing from various categories such as class (warlock, hunter, shaman, warrior, etc.), 'race', (troll, orc, human, etc.) and affiliation (Alliance or Horde).Players also make cosmetic choices (colour of hair, size of tusks, etc.).Becoming skilled entails gaining the ability to successfully develop and deploy the various capabilities of an avatar, from levelling and 'builds' (class-based specialisation) to professions (herbalist, skinner, etc.).
It is in the management of these game-play factors that the trajectory from 'noob' to established player is most obvious.At the time of data collection, the highest level that a character could attain was 70.However, a high level is not necessarily regarded as a guarantee of competence.Expertise is measured by a range of markers, including the capacity to adequately prepare for particular activities (bring the appropriate key, wear the appropriate gear, etc.), the holding of informed expectations, the fluent use of jargon or abbreviations, and the understanding of the strategic role of a character class within a group.
While a few of the interviewees had one primary character (or 'main'), most had a more promiscuous selection of avatars, designed to support different play styles (avatars for particular player-v-player battlegrounds, for example).Couple members that invested different amounts of time in the game would use 'alts' (alternative characters) to manage their rate of levelling, in order to ensure that they had a character progressing at the same rate as partners and/or guild-mates.
If an interviewee had been playing for a while when his or her partner joined the game, it was common practice for the first partner to create a new character to level up with the new arrival -for the sake of company, but also for the sake of in-game support, in that the more experienced player would usually adopt a specific role -as a healer, for example.As the term suggests, a healer character could support a new player by repairing damage and injury.
While most of our interviewees had a level 70 character, not all prioritised the end game.One of the players interviewed, for example, had reached level 70 only to become bored of repeating the end-game instances (although she expected to rejoin the game once new content was released).Another player had reached level 70 with an ice mage, only to find that levelling was the part of the game he actually enjoyed.Players have different preferences and being 'good' at playing World of Warcraft meant different things to our interviewees.This was particularly evident when players were interested in role-playing, which involves creating and performing a particular personality for a character.
Interviewer: "Do you socialise in the same way with each of the characters?"Pru: "I try not to -but I think your personality shines through anyway.I try and play Pru as a deranged sociopath for instance.But I also like to be polite." R: "I always considered myself a beginner though at Role Playing.I've always felt that I just play myself.I'm not that great at coming up with backgrounds etc. […my partner is 'better' at Role Playing] definitely.He thinks quicker than me…and is better at bullshitting."Skilled role-playing, the limitations of the game in terms of role-play, and the difficulties involved in motivating members of role-playing guilds, were issues mentioned by interviewees.Role-playing, of course, is not to everyone's taste ("I have an aversion to role playing…actually 'aversion' is too strong.But I've never really seen the point of role playing in wow.There just isn't enough scope to do anything interesting").Those of our interviewees who do enjoy role-playing also spoke about the difficulties of staging satisfying role-play in World of Warcraft.There was reference, for example, to the inevitability of certain themes, especially romantic entanglement.One of the interviewee's mentioned how 'weird' she found this initially, given her partner's character's 'engagement' to a troll.She adds, in "my first guild that was all anyone was ever doing.They were all getting married all over the place and having love triangles and that".Another interviewee express discomfort around similar issues and their potential interface with real life: "I can get touchy about the RP side that some players extend to flirting to a degree of explicitness…I don't like it in real life and that, stupid as it may sound, also extends to the game world."Yet, for at least one of the couples, role-playing is more expressive, and hence involves a more 'together' mode of play: Interviewer: "Is player together, spending time together?"R: "I think it is when we're playing these two characters together.But it's not necessarily 'quality' time together."Interviewer: "Is RP 'more together' than grinding?[i.e. point scoring in order to level up] " R: "Yes, because we're talking more and paying each other attention.Talking and listening and yes, playful and creative.Whereas a lot of questing is more mechanical." For these interviewees, the pleasures of playing together included the sharing of a specialist language, sharing an understanding of the game, undertaking shared inworld experiences and developing joint friendships.

Social resources
Learning practices relating to social aspects of the game were evident.Players referred to coaching within a guild, for instance.In most cases, when players were asked about how and why they had started playing World of Warcraft, the answers involved real-world relationships (friendship group, work colleagues, for instance).For three of the couples interviewed it was a male partner who first played World of Warcraft, but in the other couples a female partner was the initiator.In several cases the question of 'who started', or who was more experienced, was complicated by earlier relationships (where World of Warcraft had also been played) or by a shared history of playing table-top role playing games, or by reference to experience gained in another MMORPG such as Guildwars.
Interviewees also made reference to peer assessment.G, in the above quote, is working as a 'tank'.To perform this role effectively, he must generate sufficient 'aggro' to attract and keep the attention of the group's foes.G's team-mates, meanwhile, must assist by providing effective 'crowd control'; by immobilising the enemy in various ways by dealing damage, and by managing the amount of 'threat' that they generate in the process.Factors such as 'aggro' and 'threat' are aspects of game-play that indicate the importance of effective collaboration.
In the quote that follows, Z makes reference to performance anxiety, collaborative play, and to social aspects of competence.An 'instance' is a dungeon that is accessed by a group -generally it involves slaying a variety of monsters and a final 'boss' while penetrating a maze of some kind.Instances involve better quality 'loot' and a more intense form of play than regular questing.Also referred to here is classspecific error.Z indicates her responsibility for a 'wipe' (when an entire team is killed during a particularly challenging encounter) and the unease this evokes.The class or character type that Z refers to is that of 'hunter'.Hunters work with a companion beast, or 'pet'.
Z: "[by then] I had a pretty good idea what I was doing.Though instances were still scary."Interviewer: "Why?" Z: "Not being sure if I was doing the right thing.They're not scary now of course : ) There are quite a few ways to mess things up as a hunter : ) nothing too disastrous.Though every hunter has had the odd 'pet moment'.I misdirected on to the wrong person in [high level dungeon] not that long ago and caused a wipe.I felt really bad about that." Interviewer: "How did folks respond?"Z: "I think that's one of the things that makes it hard is that it's difficult to know, even when we've got teamspeak on.You can't get a feel for people's reactions like you can in real life."A: "Most people didn't notice and if they did they wouldn't care [we] don't worry about the odd wipe."Z: "It happened twice the same night though : ( " Interviewer: "But other people weren't bothered?"Z: "I don't think so really…though people do complain about people who play badly."Z's partner, A, sees tactfully addressing incompetent play as one of his responsibilities as guild leader, along with coordinating large-scale raids, and devising and hosting special events (such as 'games within a game' treasure hunts, for example).Another interviewee referred to the role of guild leader in terms of her pleasure in helping 'people feel comfortable and like a family'.
Clearly, 'playing together' can mean different things.The couple who play together yet share a single account (which means that they are unable to log in at the same time) are an example of this.Playing together might mean playing in the same room, but in different places in the game-world.One interviewee made it apparent that having a shared knowledge of the game was a form of 'togetherness' that transcended game-play.These couples worked to accommodate different preferences in play style and shifting levels of competence.Difference and accommodation were an issue when interviewees described their experiences within casual 'pick-up groups', which are temporary groupings or teams assembled for a single session or mission.As G explained "When we have had to do casual parties […] results are very mixed.And frustrating.We had one recently…two people talking about drugs, their pet dogs, and so on, and practicing jumping…but not actually helping much." There are schisms relating to expectation, performance and interpretation running throughout the player population.It is interesting to consider these matters might be managed within the context of different relationships (couples, friendship groups, guilds, casual groups).It is certainly the case that guilds can disintegrate in the face of such tensions.Here, for example, a player refers to the demise of her first guild: S: "I was trying to get them to roleplay…that was the point for me […] but they wanted to do instances instead and so on.Now I [can't] take guilds seriously anymore, maybe I should try again…don't know." As this suggests, it is important to recognise the social and collaborative nature of online, multiplayer gaming, but it is also important to avoid presenting a sanitised account of in-game social life.There is evidence of supportive mentoring, but playing World of Warcraft can also involve being reprimanded by fellow players who emphatically disapprove of a playing style, strategy or error.Players do not rate skills, interpret the game or actualise the game in a uniform manner.Consider, for example, the following comments.The interviewee, M, refers to a strategy that is not actually cheating, in that is does not contravene the rules or involve manipulating the software.It is practiced by some players, tolerated by some, and yet completely objectionable to others.
"Are you looking at how the trade structure in-game affects people?For instance, how some people focus all their time and attention on making money, be it through trade skills or through by cheap and selling high?[…It's] annoying when people take advantage of noobs -selling moderate items and claiming they are outstanding and charging sky-high prices.I suppose it's a question of morality.It's taking advantage of people's ignorance […] and some people are so unethical that it can actually ruin the game for me at times." Other, similarly controversial in-game practices include 'twinking' and 'ganking'.To 'twink' is to use either money or a high level benefactor to equip a character with disproportionately powerful weaponry.Twinking is interesting in relation to competence -a more skilled player might lose against a less skilled player with a more twinked character.To complicate matters, twinking itself might be regarded as a form of expert play (for a discussion of twinking, see Glas 2005).'Ganking' is a term used by players to refer to the hunting and ambushing of lower level characters by higher-level assailants in those areas that support player-v-player combat.As these contentious forms of play indicate, participants hold conflicting interpretations of fair play -along with different play preferences, and multiple measures of competence.

Managing resources in combination, and in context
All of the interviewees made reference to what might be described as social learning practices.What also became clear during the analysis of the interviews is that the relationship between the 'donor' and 'recipient', and the terms of such exchanges and practices, called for careful and continual negotiation: D: "I didn't really mind helping out…but at times that made her feel irritated when I watched her and kept giving advice all the time."M: "If she feels she is having a bad run, or she is getting annoyed I can jump on the computer for her and take over or suggest ways for her to play differently.Although the latter can cause arguments at times." In the following excerpt, Mel expresses exasperation connected to the role of learner.
Mel: "[I had] an alliance character for a while, to be in unexplored territory…[for access to new areas] but also because Pru had been everywhere ahead of me…like a walking spoiler."Interviewer: "Pru, is that a fair description?"Pru: "Well I try and keep my trap shut.Ain't easy though."Interviewer: "Have you had people more experienced do that to you as well Pru?" Mel: "Yeah he has!The guild babysat him thru some instances and he thought it was annoying!"Pru."I just ran after the leader and killed things.But I was in [a high level dungeon] a few days ago and no-one knew where to go -it was really fun." While it was noted that the point of playing together, is being together (rather than tutoring) all the interviewees expressed an awareness of how useful various forms of support can be.The role of expert was occasionally thankless, but there were certainly positive accounts of the leveraging of comparative expertise.Learner support took various forms.These included one interviewee's references to 'chauffeuring' -which referred to simply following another character to a particular quest location (there is a function called 'follow' that will make one character trail another).References were also made to 'boosting', which involves a low level character being aided by a much higher-level character through a particular dungeon or quest.Interviewees referred to boosting as advantageous and fun, but also as potentially dull as it strips an event of unpredictability and risk (see Pru and Mel's comments, above).Other forms of support included cases where one player in a couple was more investigative and more liable to seek out and then share specialist information on gear, weapons or effective class builds.
Existing alongside an appreciation of the benefits of mentorship was sensitivity to its potentially unpleasant or overbearing aspects.We would hesitate to make assertions about gendered roles on the basis of this study, yet (as the pronouns in the above quotes suggest) particular patterns were at least suggested by the data.There was recognition of how annoying 'back-seat drivers' can be, from both victims and perpetrators.Clearly there is a fine line between supportive mentoring and intrusive backseat driving, and its co-ordinates are vague.Crossing this line involves the compromising of a player's own sense of investment, connection and initiative.These factors could be regarded as aspects of agency, and this indicates that agency, as affect, is supported or undermined by social factors as well as game structures.
Each of our interviewees expressed an awareness of the need for tact and flexibility when managing different skills, preferences and competences.The couples learned to leverage comparative expertise while accommodating the shifting roles of learner and expert.It would be expected that similar tensions would be played out across groups and guilds, potentially contributing to both their formation and their disintegration.

Conclusion
During analysis it became clear that 'playing together' meant a number of things in addition to exploring areas or sharing a quest.Togetherness might involve illicitly sharing an account, or the shared ability to discussing the game knowledgeably over a meal.Learning in World of Warcraft is visible in players' increasingly sophisticated management of real world, ludic and social resources.The intersections of these categories are particularly interesting.As this intermingling suggests, competence in this context is complex and multi-faceted; it is variously constituted, and variously assessed.
In the process of undertaking this analysis, we developed our understanding of learning practices and social play in this MMORPG.Additionally, by focusing on the 'couple as unit' we found ourselves crossing between the game and its ludic attributes, players and their preferences, and the real-world contexts of play.This has provided a richer insight into the processes of learning in MMORPGs, revealing an ongoing process that involves sharing, specialization and negotiation, as well as a constant blurring of the boundaries between play and other areas of life.
G: "Lenny [a guildmate] thinks I'm reckless."R: "Well it's him that thinks I'm chaotic and also reckless.He's wrong about G though, I'd say."G: "Well he thinks I pull too much at once.I think I manage group aggro.It's a matter of perspective, I guess."R: "After all it takes Lenny 5 minutes to decide who to attack when he's in charge"