Game Studies

Automatic-Play and Player Deskilling in MMORPGs

Abstract

The goal of this essay is to investigate some of the consequences related to the substitution of human-play with automatic-play in Massively Multiplayer Online Role-Playing Games (MMORPGs). The concept of automatic-play refers to the use of game bots, macros and other software that allow a total or partial automation of gameplay and in particular avatar levelling. The paper theorizes a key aspect of the automation of gameplay: the deskilling of players with the transfer of human skills to automatic-play software. The paper also theorizes that the deskilling of players implies that bots and automatic-play technologies obtain skills, with the result meaning an enskilling of machines. The substitution of human play with automatic-play is investigated in the longitudinal and qualitative case study of the MMORPG Tibia which was conducted during the period January 2009 - June 2010. In the conclusion general reflections are provided on the links between the automation of play in MMORPGs and the contemporary debate on the automation of work. The conclusions advance the idea that MMORPGs play could be undergoing some of the same consequences affecting contemporary work.

Keywords: automatic-play , avatar levelling, MMOPRGs, deskilling, bots, work automation.

Introduction

Massively Multiplayer Online Role-Playing Games (MMORPGs) are a genre of computer games in which thousands of players participate in a persistent Virtual World. MMORPGs are also platforms in which players can entertain complex social interactions and explore the Virtual World (Nardi, 2010; Castronova, 2005), while still advancing in the game. Socializing and exploring however are not the only things players can do and a relevant part of the experience relates to levelling one’s avatar/character. While I acknowledge that levelling might not be the only aspect for players in participating in an MMOPRG, for the specific goal of this paper I will concentrate mainly on this aspect.

In an MMOPRG an avatar is born weak- usually at level 1 - and a key task for players is to increase their level and make the avatar more powerful. Levelling is pursued by acquiring experience points which are obtained by killing monsters or completing game quests. Collecting virtual gold is also important, because with it avatars can acquire special items. Levels and gold are, however, usually obtained via a number of repetitive, tedious and fragmented tasks ��" an activity often referred to as “grinding”. This entails repetitively killing the same monsters hundreds of times in order to achieve appreciable results.

Some parts of the academic debate[1] see a parallelism between MMORPG levelling and the scientific organization of labor (Taylor, 1911), the so called Taylorism, which is also composed of repetitive, tedious and fragmented tasks, with little skills or no intellectual ability required for workers. While this parallelism may seem an attractive conceptualization, I think that only a specific aspect of it is really relevant: the idea that levelling is indeed often repetitive, tedious and composed of fragmented tasks. Different from Taylorism however, levelling is an activity that requires players to develop a number of skills to perform certain tasks (Nardi, 2010, Schrader and McCreery, 2008). Therefore, levelling could be better compared to the manufacturing production described by Marx (1976, Chapter 14) rather than to Taylorism: repetition and fragmentation of work indeed with, however, a wide presence of workers’ skills that are necessary for production.

Starting from this consideration, I will investigate the relations between players and the machines that are sometimes devised to automate MMORPG levelling. While some forms of automation such as simple macros come from the design choices of game companies, most of the existing automatic-players are Bots: software programs that by the means of Artificial Intelligence routines, can automate some aspects of the game (for an introduction see Hoglund and McGraw, 2008), often at the expense of violating game End User License Agreements[2].

Here we can leverage the parallelism between levelling and manufacture production. In Capital (1976, Chapter 16) Marx described that large scale industrial machines were devised by capitalists in order to automate the repetitive tasks of manufacturing workers. The goal for capitalists was to increase both productivity and surplus extraction. What is relevant for my work is, however, that in this process of automation the skills and strengths of workers get transferred into machines via a process that reduces workers to appendices of machines. With this paper I would like to investigate a similar process of skills transfer taking place with automation of MMORPG levelling.

The key contribution of this essay is then to introduce the concept of automatic-play in order to characterize the use of game bots and other software that allow a total or partial automation of levelling in MMORPGs. I will show that, similarly to the automation of manufacturing, automatic-play in MMORPGs triggers a deskilling process (Braverman, 1974; Zuboff, 1988) of players with many of their levelling skills getting incorporated into software programs. However, the deskilling of players/workers should not be seen as an unilateral process. Indeed, the deskilling of human actors leads to an enskilling of non-human artifacts (Latour, 1988a) with bots and macros that are subsequently able to better perform tasks in place of the player and that display skills usually belonging to humans. In this manuscript I am therefore more concerned with the socio-technical dynamics of the circulation of skills between human (players) and non-human actors (automatic players) and a little less with the problem of capitalist production and surplus extraction. The substitution of human-play with automatic-play is investigated in the longitudinal and qualitative case study of the MMORPG Tibia (http://www.tibia.com) which was conducted during the period January 2009 - June 2010.

In the conclusion I generalize my findings by linking the concept of automatic-play with the current debate on the impacts that workplace automation has on human labor. This is a subject that has recently (re)gained worldwide attention, thanks to the publication of the book Race Against the Machine by Brynjolfsson and McAfee (2011), in which the authors claim that human labor will become obsolete and largely objectified because of an uncontrolled process of automation of workplaces. My conclusion is a provocation for the game studies field for starting reflecting on whether the automation of levelling in MMORPGs could display similar characteristic and consequences of current automation of workplaces.

The paper is organized as follows: firstly I introduce the debate on the convergence between work and play; secondly I introduce the concepts of skill and deskilling and link them with game studies; thirdly I analyze the automation of gameplay in the Tibia case study. Finally, I draw the general conclusions.

The Convergence/Parallelism Between Work and Play

In the modern world, work and play are often seen as two separate spheres of human activity (Ortoleva, 2011). For instance, the influential idea of the “Magic Circle” of play, originally coined by Huizinga (1938/1950) and brought to a worldwide scholarly attention by Salen and Zimmerman (2004), is often understood as a transcendental space in which play takes places separately from other human activities, and especially work. A convincing explanation as to why this supposed separation between work and play seems often taken for granted comes from Dibbel (2006): this is grounded in the puritan-protestant way of conceiving the religious-economic life (Weber, 1920), which in its secularized form ended in rejecting activities that were non rational, unproductive and non efficient. Play falls outside this essentially modern way of life.

With the advent of digital games the “puritan” separation between work and play seems less clear. While fully conflating work and play could be problematic in some cases (i.e. work is something people must do for a living while play is not), there is clear evidence that some forms of play display properties that are typical of work conditions (e.g. avatar levelling) and vice-versa (e.g. gamification of workplaces).

To capture the new convergence between work and play several concepts have been proposed. Ruggil et al. (2004), for instance, propose the neologism gamework in order to describe various kinds of work involved in computer game development, play and analysis. Pearce (2006) proposed the notion of “productive play” in order to describe how the creative productions of players become an active part of the digital game experience. A similar notion was introduced by Humphreys (2005) who talks about productive players and their investment in the game. Consalvo et al. (2010) defined levelling in MMOPRGs as “hard work” describing the relations between levelling and the protestant ethic and capitalistic culture. Another area where scholars see a convergence between work and play is “modding”, that involves players actively modifying the game content (Sotamaa, 2007; Postigo, 2003; Kücklich, 2005).

A number of studies have related MMORPG levelling with repetitive work and some even with the scientific organization of labor. This aspect of the convergence between work and play is the most interesting for us. Levelling is an activity composed of a number of repetitive, tedious and fragmented tasks[3]. Castronova (2005) proposed the analogy with the Greek myth of Sysiphus in order to characterize the situation of endless repetition and boredom embedded in avatar levelling. Nardi (2010) in her ethnography of World of Warcraft (Blizzard Entertainment, 2004), partially supports the idea that repetitive levelling resembles industrial work.

Nardi, however, criticizes too simplistic comparisons between repetitiveness in MMORPGs and industrial work, an idea sustained instead by Yee (2006), Kent (2008) or Rettberg et al. (2008). Yee (2006) for example says that MMORPGs are work platforms that condition us to work harder, faster and more efficiently. MMORPGs therefore prepare players to take a position in the industrial organization of labor. Kent (2008) argues that levelling is a production process in which each unit produced has no marginal production cost. Rettberg et al. (2008, p. 30) went even further by directly comparing the scientific organization of labor and MMORPG levelling:

“the majority of the play involved in advancing a World of Warcraft character is mindless and repetitive to the extent that it verges on Taylorism. There is an assembly-line mentality involved in many of the quests, many of which involve killing a staggering number of a certain type of beast or enemy (grinding), over and over again. There is little more novelty involved in grinding than there would be in welding two sections of a fender together, over and over again, all day long”.

This quotation is perhaps the clearest link between the scientific organization of labor and levelling in MMORPGs based on a fragmentation of work in which the productive process is decomposed in a number of repetitive, fragmented and tedious tasks, with no skills or intellectual abilities involved.

Game Studies and Skills

In this paper I follow the position that could be productive to consider a parallelism between work and levelling (Nardi, 2010). However I do not second the simplistic conceptualizations that see a convergence between MMOPRGs levelling and the scientific organization of labor, in particular the idea that levelling requires no skill. It is indeed widely acknowledged that playing digital games requires players to develop various skills. Also levelling in MMORPGs, even if repetitive and boring, requires players to develop certain skills.

In this work, I use the term “skill” consistently with a definition by Adler (2004) for which a skill has two basic components: the mastery of the complexity of the tasks required of workers in their jobs, and mastery of the relations that coordinate activity across these tasks. As we will see in the empirical part of the paper, this definition of “skill” is consistent with the MMORPG levelling in which a player needs to master several tasks and coordinate her skills across these tasks. It should be further noted that skills belonging to players are “real skills” something that the player has (e.g. dexterity in clicking the mouse or an arrow key of the keyboard), opposed to “virtual skills” that the player pretends to have by the mediation of the avatar (e.g. sword fighting inside the game) (for this distinction see Schell, 2008)[4].

For instance, one of the players studied by Nardi (2010, p. 55) refers to real skill development as the “mastery of the game”: “the need to attend to the development of skilled performance as critical to player experience”. Nardi (2010, p. 55) further argued that “players wanted the loot and the experience points. But mere acquisition was not the sole source of satisfaction; the loot and points accumulated, over time, toward the player’s object of becoming a better player- of “improving yourself” ”. The acquisition of real skills is a crucial aspect of playing MMOPRGs. Castronova (2005) sees real skills as a core component of what he calls the MMORPG “production function”[5] which accounts, for instance, for the total output of gold or levels accumulated by a player.

Other studies on the relevance of real skills in digital games (Wang et al. 2011; Kennedy, 2006) have pointed out issues related to play performance and development of experience or the ability to take decisions, often in risky situations, while playing the game. Levelling in MMORPGs also requires (real) cognitive skills and ability in decision making (Schrader and McCreery, 2008). For example, an MMORPG player might need to choose whether to attack certain monsters or not. If the monster will not provide a substantial amount of experience points (i.e. it is a low level monster) then the player might decide to skip targeting such a monster. On the other hand, if a monster is considered too strong for the current level of the avatar (i.e. high risk of being killed), then the player might also decide to skip targeting that monster. Furthermore, at any stage of the levelling the player might decide to heal the avatar. This also implies a cognitive skill as the player needs to evaluate whether the current level of health is enough to continue playing.

The Deskilling Thesis: Braverman, Latour and Marx

The central hypothesis of this essay is that the use of machines, such as bots or macros, that replace human-play with what I define as automatic-play, bring with themselves a deskilling process (Braverman, 1974) involving a “computerization” of skills (Zuboff, 1988), with players skills getting incorporated into software programs.

The original deskilling thesis is that management control and degradation of workers are the basic preconditions for productive efficiency. Deskilling is then the process of reducing the skill level of a work force with the replacement of humans with technologies. This replacement has some consequences such as the dissociation of the labor process from the skill of workers, the separation of work conception from execution and the concentration of knowledge in the hands of management (Heisig, 2009). My intention is not to follow entirely the original deskilling thesis (Braverman, 1974), in particular in its harsh critique to capitalistic production process. However I am intrigued by the problem of investigating how skills and abilities circulate between humans (i.e. players) and non-humans (i.e. bots, macros) and some consequences that this phenomenon entails, especially the problem of humans becoming appendices of machines.

Indeed, several critiques have been developed on the deskilling thesis. A position that I second comes from Latour, (1988a, 1988b, 2005) that remarked the existence of a social reductionism, as the deskilling concept does not take into account the role of non-human actors, such as technologies, in shaping social relations. For Latour the deskilling of humans implies a symmetrical process of non-human enskilling. Latour makes the example of the door-closer, a mechanism common in public spaces that automatically closes doors.

Figure 1. Door-closer mechanism[6]

The door-closer creates an enskilled door which is able to close autonomously itself, at the same time a hypothetical human porter whose job is to close the door behind passersby (for instance in a Hotel) is deskilled as she loses her working skill. Skills for Latour circulate symmetrically among humans and non-humans. Therefore the second central hypothesis of this essay is that the deskilling of players leads to an enskilling of bots and automatic players more generally.

The process I wish to investigate with the deskilling of players and the consequent en-skilling of bots has some similarities with the automation of manufacturing described by Marx in Capital (1976, Chapters 14 and 15). In manufacturing, workers perform the same tasks over and over, with the production process fragmented in repetitive tasks carried out with tools (e.g. hammer). Manufacturing however retains the character of a handicraft with the strength and skills of workers having a relevant role.

In the description of large scale industry Marx (1976 and 1939-41) adopts a similar perspective to that of Latour. The shift from manufacture to industrial production with large automatic machinery brought a replacement of craftsmen with unskilled workers. Marx (1939-41, pp. 693-694) however remarked that when the tasks of manufacturing got embedded in industrial machines, then:

“it is the machine which possesses skill and strength in place of the worker, is itself the virtuoso, with a soul of its own in the mechanical laws acting through it; and it consumes coal, oil etc. (matières instrumentales), just as the worker consumes food, to keep up its perpetual motion. The worker's activity, reduced to a mere abstraction of activity, is determined and regulated on all sides by the movement of the machinery, and not the opposite”.

For Marx the deskilling of manufacturing workers did imply an enskilling of industrial machines. This brought, on one side, the reduction of workers to mere appendices of industrial machinery and on the other side machines taking a “virtuoso” role in the productive process, with the machine acquiring an independent form, emancipated from the human operator. This is what Marx (1976, p. 503) calls an “automatic system of machinery”: a machine that executes without man's direct help, in a situation in which human labour becomes complementary and largely obsolete.

The Case Study of Tibia

In order to study automatic-play in MMORPGs (the substitution of human-play with machines) I propose an analysis coming from the longitudinal and qualitative case study research of the MMORPG Tibia (http://www.tibia.com).

Tibia is developed by the German company CipSoft (1997-2013). Tibia has a player base of around 300,000 with a total number of accounts around 1.3 million and played on approximately 70 servers located in Germany and the United States. Tibia, as with many other MMOPRGs, has in recent years been affected by the diffusion of bots. Using a bot is a violation of the Tibia game rules and a form of cheating. Bots in Tibia are mainly developed by two independent companies who sell their bots over the internet.

In January 2009, Cipsoft launched an anti-cheating campaign whose goal is to fight the use of bots in Tibia. This involved the introduction of bot detection software and the temporary banning of thousands of accounts. This anti-cheating campaign was the main reason for choosing Tibia as the case study for this research.

I collected data related to the anti-cheating campaign for one and a half years (from January 2009 till June 2010) and this came from a number of sources, using participant observations and archival research. The core of the data comes from the Online Forums of Tibia (I collected 1485 forum pages composed of 20 posts each), where players and the company discuss the game experience. The data collection further included game documents such as the Tibia game licenses, the manuals and several articles written by Cipsoft prior to and during the anti-cheating campaign. I played Tibia with two different characters/avatars (at least 6 hours a week for the first 6 months of research) in order to understand the gameplay and, in particular, the avatar levelling. In addition I collected forum threads (1724 in total) coming from the websites of the software companies developing bots for Tibia.

The Tibia case study has been designed with an inductive, bottom-up approach. The data collected have all been analysed using a Grounded Theory approach (see Charmaz, 2006) in which the theory (i.e. the relations between bots and player deskilling) is built via a set of recursive relationships between the data and the concepts composing the theory.

Avatar Levelling in Tibia: Skills and Tasks

Avatar/character levelling in Tibia is similar to other MMORPGs: players need to kill monsters which provide characters with experience points whose accumulation leads to increases in levels.

Tibia levelling can ideal-typically[7] be broken down in a number of tasks - carried on by the use of specific tools/game controllers such as the mouse or the keyboard - requiring specific skills that are summarized in Figure 2 and in Table 1. Levelling involves the player moving the character around in the game map (Task 1), following and killing (Task 2) the monsters by using the computer mouse and the arrow keys of the keyboard (skill). Once the monster is killed, the player is required to “open” the monster (Task 3) with a mouse right-click operation (skill) to see what the monster was carrying. Then the player will loot the content (Task 4): still with the mouse, the player is required to drag and drop the content (skill) of the monster (such as the gold) into the character's back-pack or bag on the right side of the screen. During or after these actions, the player might also be required to heal the character (Task 5). Healing can be done either with a magic potion or by casting a spell. Healing requires the player to click a hotkey to which such an action has been assigned or, alternatively, to drag and drop a potion from the character bag onto the character itself located at the centre of the screen (skill). Hotkeys allow players to cast spells, use runes or other items by pressing one keyboard button. Players can assign spells and other actions to Function keys (F1, F2 etc.) (hotkey) at the top of keyboard.

Figure 2. Ideal-type of the levelling cycle in Tibia with tasks that need to be repeated several times

Table 1. Tasks and Skills of the levelling cycle.

Tibia and Automatic-Play

Levelling in MMORPGs often entails a problem with repetitiveness and boredom. The problem is known by Tibia players as well, as the following posts well remark:

“There are many other frustrating factors in hunts[8] other than healing, for example looting monsters. It isn't hard to do, but when you're killing sometimes thousands of monsters on a hunt, click dragging starts to become a tedious task[9].”
[Tibia Forum, Character[10] Iven Tepes, 05 January 2009, Post #21753805]

“It can also get very boring to hunt. I challenge you to find anyone who actually enjoys hunting rotworms, or larvas[11]. The repetitive click, wait, click, drag cycle can become frustrating and tiresome “
[Tibia Forum, Pumpkin Pie, 06 June 2008, Post #19022044 ]

Boredom and repetitiveness are identified as a direct outcome of fragmented and trivial tasks of levelling that are repeated in the same mechanical form for an endless number of times. Given this problem, players look around for alternative solutions. One solution would be to make changes to the game design. For instance with an update/addition to the game done by Cipsoft in 2010, a message is displayed at the interface level whenever a monster is killed. This message tells the player what is the “content” of the killed monster before opening it: ‘loot of a monster: 2 gold pieces’; ‘loot of a monster: nothing’ and so on.

Figure 3. GUI Message that replaces a levelling task

This Graphical User Interface (GUI) message sometimes replaces the task of opening the monster (T3 in Figure 2) as well as the related “right-click-open” skill which is necessary for looking inside the killed monster. This GUI message allows the player to skip (sometimes) repetitive and unproductive tasks such as opening and finding the monster void (Figure 3).

Another solution to mild down repetitive and boring aspects of the game would be to adopt machines that can carry on autonomously the tasks in place of players. There are at least three types of machines for automating play: (1) “machines” that players think should be implemented directly by CipSoft (such as official macros); (2) bots that are effectively implemented and whose use is a violation of the game rules (unofficial software); (3) official design changes that automate some aspects of the gameplay. The following message is an example of how a player requests (to the game company) the first type of machines:

“Please implement some sort of automatic loot function because that's the most tedious thing in the world for me to do. Say I'd be hunting some rotworms on a lower level character and I simply won't loot the corpses because it's so boring having to constantly click-this-now-drag-whatever-sum-of-gold-into-this-backpack.“
[Tibia Forum, Skurth, 11 January 2009, Post #21819785 ]

Indeed, an official loot function will allow the player/character to automatically collect the loot, therefore skipping the repetitiveness of the action of right-click-open / drag-and-drop. An official looting macro however does not exist for Tibia. Automating the whole levelling is instead a key feature of bots.

I conceptualize the adoption of a bot as a form of replacement of real skills with computer skills embedded/translated in programs that do the same tasks. A bot is an automatic machine that replaces players and makes (human) play a complementary aspect of the game. If the manual levelling is fully replaced by machines what we have is a replacement of real skills with automatic skills: a process of players/cheaters deskilling, in which the bot becomes enskilled and can carry on the production-levelling in almost full autonomy. Fair players and bot users clearly acknowledged this replacement of human skills with machines:

“What are bots? Bot is another word for 'macro', an automated program which, in Tibia's case, controls your character “
[From http://tibia.wikia.com/wiki/Bot]

“Exactly, PEOPLE play... What we're doing isn't playing the game ourselves, we're letting a program play it for us”
[Tibiabot Forum, secre555, 02 September 2009].

With bots we have the dissociation of the levelling process from the real skills of players (“we're letting a program play it for us”) and the separation from the game conception (what I as a player/cheater would like to achieve in terms of levelling, e.g. my goal is to reach level 100 in 1 month) and the mere execution of this (what actions my avatar does, when controlled by the bot, in reaching my goal). The cheater using a bot becomes a supervisor of a machine that does largely in full autonomy the levelling.

Automatic-Play: Replacement of Real Skills with Software

For Marx (1939-41) an Industrial machine could autonomously produce thanks to “mechanical laws acting through it”. Physical processes such as leveraging the combustion of coal for producing energy allows industrial machines to become autonomous. With bots we do not have a mechanical law, rather the bot becomes an “autonomous agent” because it is composed of artificial intelligence routines.

In bots the replacement of real skills with automatic skills (a deskilling process of players with the enskilling of non-humans) relates, therefore, to computer code in the form of scripts and GUIs representing these scripts for the user. Tibia bots possess a number of standard features that the user can control from the GUI. When a player/cheater runs the bot she will see the GUI composed of several options (see Figures 4 and 6). Actions such as “Auto-attack” a monster (replacement of tasks T1 and T2 of the levelling cycle), “Auto-loot” the corpse (replacement of tasks T3 and T4) or “Auto-Heal” the avatar (task T5) are standard features of the bot and can easily be configured from the GUI. I will now analyse the replacement of real skills with software using a reproduction of some aspects of the GUI of a well known Tibia bot[12].

I analyse the example of the Auto-Healing[13] feature of the bot (which replaces the task T5 and associated skills). That of “health” is a key component of an MMORPG avatar and it is a numerical indicator and/or a progress bar (see Figure 5) found on the game interface, that decreases when the avatar is hit by monsters or other players and that can be increased usually with healing potions. If health reaches the value of zero, the avatar dies and looses most of her gear. Using a potion or a spell to heal the avatar and avoid death requires real skills: for instance the player needs to be quite fast to heal the character for example with the drag and drop, otherwise there is the risk (with the character low on health) of being killed by nearby monsters which will suddenly attack. While I was playing Tibia as part of my participant observation I was killed several times by monsters, because I was not fast enough to drag and drop the potion on my character. The bot does not have this drawback as the healing is quickly and autonomously done once the health of the character is below a specified threshold. The process is as follows.

Once the bot program is launched, the user will find the Auto-Healing option on the GUI (Figure 4 reproduces this feature[14]). The user can set the minimum level of health for healing: below the specified threshold/value (245 in the example) the bot will automatically heal the character, for example using an healing spell, again specified by the player[15] in the field “Healing Spell” of the GUI. In this way the healing task becomes setting the parameters upon which the bot will take its decisions whether to heal or not. After this the bot can autonomously heal the character and human intervention becomes only a supervision to the operations of the bot.

Figure 4. Reproduction of the Auto-Heal function on the bot GUI

Figure 5. Reproduction of the Tibia health bar

We can further see how the player’s healing skill is translated in the bot as computer code. The following is a script that Auto-Heals a character by the means of a spell (spellname), when the health is below the specified value (245 in Figure 6).

Figure 6. Healing Script. From http://scriptstibia.webnode.com/scripts/health-potion/

Figure 4 is the GUI representation of this script. This portion of code contains at least two player skills that have been translated in the automatic player. One is a cognitive skill as the bot can decide autonomously to heal the character when the health value is below the specified threshold. The second human skill translated in this script relates to the use of the healing spell. In order to use a spell the player is required to type the name of the spell in the game chat and press enter. Another (and quicker) option for casting a spell is to click a hotkey to which the command of the spell has been previously assigned by the player. These actions rest on the player's skills. The bot replaces these skills with the command Self.Say('spellname’). Another way for the player to heal the avatar is to drink a healing potion. In this case the command Self.Say('spellname’) in the script should have been replaced by the command Self.Containers.UseItemWithSelf(266), where 266 is a number that identifies the item “healing potion” for the game client. This command would again replace the player’s skill of dragging and dropping the healing potion from the character bag to the character itself (UseItemWithSelf) (see Figure 2/Task 5).

A further key feature of bots is controlling the walking of the avatar. The bot is able to move (without human intervention) the avatar along a number of waypoints: specific coordinates of the game map (see Figure 7). The bot is able learn these waypoints (after the user inputs) and follow them.

Figure 7. The concept of Waypoints

Once a set of waypoints has been defined, the bot user is in the position to allow the bot to autonomously move the avatar. If we look at Figure 8, we see where this skill is embedded in the bot GUI. Sets of Waypoints can be entered in a textual form of the GUI (“Current Waypoints”) and when the player ticks the option “Follow Waypoints”, the bot will automatically move the avatar around the game map following the defined waypoints (waypoint1, waypoint2 and so on). The “Follow Waypoints” command replaces the task T1 and the related skills (Figure 2 and Table 1): the player’s skill of pointing with the mouse and clicking the map-location she wishes the avatar to move on.

Figure 8. Reproduction of some aspects of a bot interface

It is not easy to reproduce the scripts of this functionality, due to the length of the code [16]. The principle at work is the same as in the case of the healing script: computer code replaces human skills and makes the bot an autonomous player.

For completeness, it is interesting to look at the bot interface reproduced in Figure 8 observing further player skills that are translated into the bot. The bot can automatically loot the dead monsters and it can also decide what to loot. In the reproduction of the GUI (Figure 8) the bot will only loot gold pieces. Other looting items can easily be added by the user by simply typing in the form “Loot Items” the name of the item (e.g. food). With this feature we have the replacement of a number of player's skills: (1) the human skill which is necessary to open the monster once killed (right-click with the mouse) and (2) the skill for dragging and dropping the loot from the corpses to the character’s back-pack. Further we have a player’s cognitive skill which is replaced as the bot decides (thanks to the prior user definition of the loot) if it is worth looting or not the content of the dead monster.

Automatic-Play and Official Deskilling

Also official changes to the game design can replace real skills. The message at the interface level in Figure 3 is an example. A comparison between cases of “official deskilling” and the one triggered instead by the use of illegal bots will allow us to better understand the consequences of the automation of gameplay.

With an upgrade brought by the Tibia 7.8 version of the game client (September 2008), Cipsoft introduced two relevant changes to the gameplay: “hotkeys” and a “macro for auto-aiming” at opponents. These two changes affected both the levelling process as well as the Player Vs Player (PvP) part of the game. We did not touch upon this second aspect, but most Tibia servers allow players to engage in battles with other players. Killing another character does not provide experience points, but the winner can loot the gear of the dead opponent. Furthermore, battles among players are often carried out in opposing guilds. Winning guilds can dominate the politics of a server.

Going back to the changes introduced with the 7.8 update, hotkeys replaced the commands that players had to type inside the chat with a simple button-click, and in particular spells, runes or other communications (e.g. phrases used to interact with Non Player Characters). Prior to the introduction of hotkeys, casting a spell like the “exura” (a spell that heals characters) required the player to type the text in the game chat and then press enter (left side of Figure 9). This is still possible, however players can now also assign spells or other commands directly to hotkeys (F1, F2, etc.). Hotkeys speed up the process of casting a spell: the player no longer needs to type the text inside the chat and press enter, she just needs to click the hotkey (Figure 9).

Figure 9. Typing the spell in the chat and press enter, opposed to clicking a hotkey

The second change introduced with the update is what players call “the official aimbot”: a macro that automatically aims at monsters or player opponents once they appear on the screen. Before this change, the player had to manually aim at each opponent with the mouse pointer and click the left button of the mouse.

According to players both these new features were introduced because players could not keep up with other players using unofficial bots and the changes were meant to facilitate the competition. It is a common request on forums that these two changes are reverted by the game designers, especially now that the anti-cheating campaign is meant to eradicate illegal bots from the game:

“How good is the person at healing, how good is a team at performing combos manually? Who is the dedicated person who puts the most hours in and hunts the most skillfully to level the quickest? These are just some of the great aspects of Tibia that were once present and I hope we will see again, of course that would have to include the removing of the built in Aimbot, which I hope will be carried out, but that is another matter.”
[Tibia Forum, Ariser, 31 January 2008, Post #22046912]

We see that this player considers how the design changes negatively affected the gameplay. What I found interesting of these changes is the relation between players' real skills and the automation of gameplay[17]. The following post well describes this aspect:

“7.8 - cipbot introduced, pvp was gone, the most fun part of this game, a superb quality of Tibia where if you had no skills, you wouldn't survive, you had to trust your own skills, otherwise you would get killed or you wouldn't kill.”
[Tibia Fourm, Ruffles of Hell, 21 December 2007, Post #17035771]

A number of Tibia players (many posters on forums did express this concern) were clearly not happy about these changes. What happened here was an official deskilling process of players. Both casting spells by typing the command in the game chat and manually aiming at monsters or other players were considered advanced gameplay skills. The faster you could cast a spell or aim at your opponent would make you a better player, with more chances to survive and win in PvP battles or more chances to kill high level monsters. Furthermore, if a player could easily do combined actions (combos) such as casting a spell (in the chat) while in the mean time aiming (with the mouse pointer) at the next monster, then this player possessed great playing skills. With hotkeys and the aiming macro, casting a spell, aiming at opponents and these two tasks combined have been reduced to an automation and related real skills have been taken away from players and translated in the game design. Only with the old setting could one demonstrate her skills, whereas the new changes flattened the skill level of players. Now with the new automations everything is reduced to clicking a hotkey and supervising the auto-aiming of opponents.

It is clear that what took place with the official macros is a deskilling process with players feeling that their skills were taken away from them by the new design. Being good at fighting and in particular at aiming, was removed as part of player skills with the introduction of the “aimbot”:

“old tibia was fun. no doubt about that. now its just about lvls and it requires no skills. if u r hunted by most ppl of the server u cant stand up for urself... in the old days even if it was u against a lotta people u could still make difference...”
[Tibia Fourm, Vicenti board, 31 January 2009, Post #22358502]

For precision, it should be noted that some Tibia players still consider the changes of the 7.8 update acceptable and in particular those who began to play after the 7.8 update would not even know how the game was before.

It is clear that for a number of players there are problems related to the automation of gameplay and the replacement of human skills with automatic skills. There is however a substantial difference between the case of illegal bots: in this case players are not voluntarily giving away their skills, the decision is imposed onto them by the new design.

Discussion and Conclusion

The convergence/parallelism between work and play in digital games is a key problem for game studies. The concept of automatic-play as theorized in this manuscript advances our knowledge on a fundamental aspect of the future evolution of this: a possible parallelism between automation in MMORPGs and workplaces. The contribution of this paper therefore acquires relevance if we link its results with the current debate on automation of workplaces. In particular what I would like to achieve with this conclusion is to provoke the game studies field in thinking about the possible connections between these two phenomena.

There are indeed recent contributions (Brynjolfsson and McAfee, 2011; Ford, 2009) that have renewed the debate about the possible negative impact that work automation could have on the economic system as a whole. These contributions extend a discussion that goes back at least to the influential book by Rifkin (1994) The End of Work, which foresaw a decline of work in manufacturing and agricultural sectors, because of the wide diffusion of information technologies. These authors describe in particular the ever increasing diffusion of smart (i.e. artificial intelligence) computerized artefacts and the consequent replacement of workers (including skilled ones) with technologies. For them, work automation brings the risk of a decline of the number of jobs available on the market, and the obsolescence of human labor more generally. We do not need to go as far as these authors in discussing the issue of work automation, also because their goal is clearly to devise a general critique of the contemporary economical systems. However, I believe it is legitimate to ask, following them, if there is the risk that with the automation of gameplay, manual play (and in particular MMORPGs levelling) could become obsolete.

I believe that this paper clarifies that the automatic-play is part of the same process of work automation described by Brynjolfsson and McAfee (2011): an increasing tendency toward the replacement of human activities with artificial intelligence technologies. One of the consequences of automatic-play is the deskilling of players with playing software demonstrating skills that would usually belong to human players: bots and macros create what Marx would call an “automatic system of machinery” for MMORPGs avatar levelling.

In particular, in the Tibia case we had the opportunity to compare two different forms of deskilling of players related with automatic-play. These two forms of deskilling have differences but also similarities. Clearly there are some differences that must be acknowledged between to automation of MMORPGs and automation in workplaces, especially those described in the early models I refer to (i.e. Braverman, Marx). We analyzed the unofficial deskilling with players/cheaters who eventually give away their skills. Deskilling here is a voluntary process that seems sometimes to improve a player’s playing conditions (i.e it offers a solution, even if illegal, to the grinding). In this case the player is the one who gains from technological efficiency. So in this case it is as if automation positions the player more as an owner of the means of production than a worker. This is different from the traditional deskilling models.

The situation is partly different to the official automation (i.e. hotkeys, aimbot). In this case the decision to automate was taken by the game company. For several players, the consequence of the introduction of gameplay automations has been a deskilling of the sort similar to the traditional deskilling thesis (Braverman, 1974): we have a degradation of play as players feel that some aspects of the gameplay have been reduced to mere automations: clicking hotkeys and supervising machines. The link with technological efficiency is however less clear in this case as neither the player nor the company seem to gain in productivity.

The discrepancies existing between the models I use in this paper (Marx, Braverman) and the case study are not limitations per se however, as I explicitly used these models because they provide a solid background for studying the issue of deskilling and work automation. The discrepancies with the models might indeed also be an opportunity and open the space for further research to improve the models as well as for better addressing the peculiarities of the automation processes taking place in MMORPGs.

Despite the clear differences both these deskilling processes (bots and official macros) display a similar outcome. Indeed, a key aspect of the automation of gameplay is the objectification of MMORPG levelling. In Capital Volume 1, Marx (1976, p. 508) argues that in Manufacturing, production is “subjective”: it is based on workers abilities to master the labor process. Instead, in large scale industry, with the massive use of automatic machines, production becomes “objective” with the workers being only a “pre-existing material of production”: workers are mere appendices of machines who carry on the production process in full autonomy. This outcome is a common denominator in automation of workplaces and MMORPG levelling.

The official deskilling in the Tibia case which comes from a decision of the designer clearly shows dynamics of objectification. Those aspects of play that are automated (casting a spell, aiming at the opponent) are not based any more on players’ real skills. The old skill of performing combined actions or the skill of being able to fast type a spell in the chat have been removed and once play becomes automated it is also objectified for many players: for those aspects players become the appendices of automations. An objectification also happens with bots. Players/cheaters are the supervisors of machines that play in their place. Players/cheaters just need to set standard parameters on the bot interface and after this players become the appendices of automations. It is clear then that in both cases the automation of gameplay can bring an ever increasing objectification and a tendency toward obsolescence of MMORPG levelling. The question-provocation that this paper leaves open for future critical investigations on automatic-play is then: could play in MMORPGs really becoming obsolete as much as work?

Endnotes

[1] Literature review is presented in the next section.

[2] Bots are also a form of cheating.

[3] See empirical sections for a further discussion.

[4] Castronova (2005) differentiates between “personal” and “avatar skills”.

[5] Production Function is a concept from Economics, which specifies the output of, for example, a company or a nation state.

[6] Creative-commons image from http://en.wikipedia.org/wiki/File:Door_closer_manual_dscn1666a.jpg

[7] What is presented is an ideal-type in the weberian sense of the term as an approximations to the reality, with emphasis on selected important aspects.

[8] Hunting is synonym of levelling for Tibia players

[9] Most forum messages are authored by non English mother language speakers. Posts are presented as they are with spelling and language mistakes.

[10] “Character” will not be repeated in the other presented posts.

[11] Rotworms and larvas are types of monsters.

[12] For advanced users, bot allows the creation of tailored scripts.

[13] The Healing Feature is relatively easy to analyse, that is why is taken here as an example.

[14] This is just one component of the whole interface.

[15] Using a spell requires a certain level of Mana. The same script can be used for an Auto-Mana recharger.

[16] A list of scripts, including this one can be found here: http://scriptstibia.webnode.com/scripts/

[17] The official “aimbot” is a macro which automates the gameplay.

Acknowledgements.

This research received the initial support of the Irish Higher Education Authority under the PRTLI 4 programme and their partners on the ’Serving Society: Future Communications Networks and Services’ project (2008-2010).

I would like to thank Dr. Aphra Kerr for her support and Dr. Giacomo Poderi and Prof. David Hakken for reading early versions of this manuscript. I also wish to thank the ahref Foundation for supporting the initial writing of this manuscript.

References

Adler, P. (2004). Skill Trends Under Capitalism and the Socialisation of Production. In C. Warhurst, I. Grugulis & E. Keep (Eds.), The Skills that Matter (pp. 242-60). London: Palgrave-Macmillan.

Blizzard Entertainment. (2004). World of Warcraft. Irvine, CA: United States.

Braverman, H. (1974). Labor and Monopoly Capital: the degradation of work in the twentieth century. London & New York: Monthly Review Press.

Brynjolfsson, E. & McAfee, A. (2011). Race Against The Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy. [kindle edition]. Lexington MA: Digital Frontier Press.

Castronova, E. (2005). Synthetic Worlds: The Business and Pleasure of Gaming. Chicago: Chicago University Press.

Charmaz, K. (2006). Constructing grounded theory: A practical guide through qualitative analysis. London: Sage.

CipSoft, (1997-2013). Tibia. [Online Game], Regensburg: Germany: CipSoft, played January 2009-June 2010, http://www.tibia.com

Consalvo, M., Alley, T. D., Dutton, N., Falk, M., Fisher, H., Harper, T. & Yulish, A. (2010). Where Is My Montage? The Performance of Hard Work and Its Reward in Film, Television, and MMOGs. Games and Culture, 5(4), pp. 381-402.

Dibbell, J. (2006). Play Money: or, How I Quit My Day Job and Made Millions Trading Virtual Loot. New York: Basic Books.

Ford, M. (2009). The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future. Acculant Publishing.

Heisig, U. (2009). The deskilling and upskilling debate. In R. Maclean, & D. Wilson (Eds.), International handbook of education for the changing world of work (pp. 1639-1651). Dordrecht: Springer

Huizinga, J. (1938/1950). Homo Ludens. Boston, MA: Beacon Press.

Humphreys, S. (2005). Productive Players: Online Computer Games’ Challenge to Conventional Media Forms. Journal of Communication and Critical/Cultural Studies 2(1), 36-50.

Hoglund, G. & McGraw, G. (2008). Exploiting Online Games: Cheating Massively Distributed Systems. First: Addison-Wesley Professional.

Kennedy, H. (2006). Illegitimate, monstrous and out there: Female Quake players and inappropriate pleasures. In J. Hollows, & R. Pearson (Eds.) Feminism in popular culture (pp. 183-201). London: Berg.

Kent, M. (2008). Massive Multi-player Online Games and the developing political Economy of Cyberspace. Fastcapitalism 4(1). Retrieved 10 October 2011 from http://www.fastcapitalism.com/

Kücklich, J. (2005). Precarious playbour: Modders and the digital games industry. Fibreculture 5. Retrieved 25 July 2010 from http://journal.fibreculture.org/issue5/kucklich.html

Latour, B. (as Jim Johnson) (1988b). Mixing humans and nonhumans together: The sociology of a door-closer. Social Problems 35(3), pp. 298 - 310.

Latour, B. (2005). Reassembling the Social: An Introduction to Actor-Network Theory. Oxford: Oxford University Press.

Marx, K. (1976). The Capital: A Critique to Political Economy, Volume 1. trans. B. Fowkes. London: Penguin.

Marx, K. (1939-41., trans. 1973). Fragment on machines. In Outlines of The Critique of Political Economy, pp. 690-712. Retrieved 18 January 2012 fromhttp://www.marxists.org/archive/marx/works/1857/grundrisse/

Nardi, B. (2010). My life as a Night Elf priest: An Anthropological account of World of Warcraft. Ann Arbor: University of Michigan Press.

Ortoleva, P. (2011). Homo ludicus. The ubiquity of play and its roles in present society. GAME, the Italian Journal of Game Studies, 1/2012. Retrieved 18 October 2012 from http://www.gamejournal.it/homo-ludicus-the-ubiquity-and-roles-of-play-in-present-society

Pearce, C. (2006). Productive play: Game culture From the bottom up. Games and Culture 1(1), pp. 17-24.

Postigo, H. (2003). From Pong to Planet Quake: Post industrial transitions from leisure to work. Information, Communication and Society 6(4), pp. 593-607.

Rettberg S., Corneliussen H. G. & Rettberg J. W. (2008). Corporate Ideology in World of Warcraft. In H. G. Corneliussen & J. W. Rettberg (Eds.) Digital Culture, Play, and Identity: A World of Warcraft Reader Cambridge, (pp . 19-38). Cambridge MA: MIT Press.

Rifkin J. (1994). The end of work. New York: Putnam.

Ruggill, J. E., McAllister, K. S. & Menchaca, D. (2004). The gamework. Communication and Critical/Cultural Studies 1(4), pp. 297-312.

Salen, K., & Zimmerman, E. (2004). The rules of play. Cambridge, MA: MIT Press

Schell, J. (2008). The Art of Game Design: A Book of Lenses. Burlington, MA: Morgan Kaufmann Publishers.

Schrader, P.G. & McCreery, M. (2008). The acquisition of skill and expertise in massively multiplayer online games. Educational Technology Research and Development 56( 5-6), pp. 557-574.

Sotamaa, O. (2007). On modder labour, commodification of play, and mod competitions. First Monday 12(9). Retrieved 10 September 2011 from http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/2006/1881

Taylor F. W. (1911). The Principles of Scientific Management. New York: Harper & Row.

Yee, N. (2006). The labor of fun: How video games blur the boundaries of work and play. Games and Culture 1(1), pp. 68-71.

Wang, J., Huffaker, D. A., Treem, J. W. , Fullerton L., Ahmad, M. A., Williams, D., Poole, M. S. & Contractor, N. (2011). Focused on the prize: Characteristics of experts in massive multiplayer online games, FirstMonday 16(8). Retrieved 18 January 2012 from http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/3672/3028

Weber, M. (1920). The Protestant Ethic and the Spirit of Capitalism, translated by Talcott Parsons, 1930. London: Unwin University Books.

Zuboff, S. (1988). In the Age of the Smart Machine. New York: Basic Books.