The first episode of my «RTS Galore!» series got some feedback that I wish to address. Commenters on Reddit and Raddle (I’m not registered on the latter and can’t answer directly) pointed out two main problematic aspects in the piece that, if not dealt with, can become systematic in my other writing on the subject. I haven’t intended this urge to play and write about as many RTS games as possible to be serious academic research. All in all, it’s just for pure fun with a bit of frivolous textual analysis to spice things up.
Still, I feel the need to establish a more solid framework to avoid misunderstandings and make the series less arbitrary and more focused. Sorry that I haven’t thought about that earlier, well, alas.
System Justification
One commenter thought that I’m critiquing RTS games for not delivering a specific political message (that appropriate to me) and devalue them because they don’t provide enough justification for the depictions of violence. Despite clearly stating in the text that I do enjoy RTS games, the message apparently was not so clear. So, let’s clarify.
What I meant under «justification» when saying «that absence or scarcity of justification makes most RTS games fun in the digital fictional spaces» is the need to rationalize and defend the status quo. Jost and Andrews define system justification as follows:
«System justification is a social psychology term of art that designates any motivational tendency to defend, bolster, or rationalize existing social, economic, and political arrangements. It is conceptualized as a response tendency possessed by many, or perhaps most, members of society to see aspects of the overarching social system as good, fair, and legitimate». (Jost & Andrews, 2011)
Since video games are made dominantly by people, it’s logical to assume that they represent system justification in their textual or mechanical aspects. We live in a capitalist system with neoliberalism as its justifying ideology, so digital games (as well as other forms of media) tend to rationalize the current status quo while enforcing certain myths and stereotypes. It’s important to emphasize, just in case, that there is nothing wrong with enjoying games despite these facts—acknowledging ideological constructs in entertainment media doesn’t inherently make it less enjoyable.
Andrew Baerg argues that digital games may entrench subjects within neoliberal ideology. Relating to «Persuasive game» by Ian Bogost, Baerg writes:
«Whether the purpose of a digital game might be to amuse, train, simulate, advertise, publicize, or motivate, its procedural rhetoric always produces some type of engagement with processes that can influence and potentially persuade.» (Baerg, 2009)
Baerg explores two elements linking neoliberalism to digital games:
«1) the progressive expansion of free markets characterized by choice; and 2) an accompanying market-based calculative rationality grounded in risk-management. By employing these two elements, the digital game, and its processes of computational representation, enacts rules driving a symbolic system that represents processes at work in neoliberalism.» (Baerg, 2009)
In my previous article, I suggested that RTS games are less intricate in their justification of the status quo than the other genres of strategy games due to their mechanical constrictions. On the second thought, I see this assumption as silly and meaningless. How in the world can I factually measure whether Civilization justifies capitalistic expansion more than Age of Empires? So let’s change the focus of the series.
Souvik Mukherjee analyses how digital games construct conceptions of politics and ethics and how postcolonial myths inspire them. Taking Empire: Total War as the leading example, Mukherjee argues that RTS games enforce imperialist and capitalist logic at the core:
«The map of ETW is a map but also a world where the player and the AI agents are constantly moving and their movement in space determines the action on the political level. One could argue, therefore, that the gameplay of RTS games primarily follows the expansionist logic of empire and its Great Games.» (Mukherjee, 2015)
«The struggle for Lebensraum then becomes equated with the struggle for resources. Resource gathering is an important aspect of RTS games: to have a production-based capitalist economy (summed up earlier as ‘trade and taxes’) is to win the game. Such a resource-hungry geopolitics also creates the binarism of centre and its peripheries.» (Mukherjee, 2015)
D. I. Waddington sees the problem in the real-timeness itself, warning about potential impacts of RTS games on education:
«If it were true that certain game experiences help us fetishize speed and acceleration, it would be a worrisome outcome, particularly for those who oppose neoliberal capitalism. Just as we thought of the early progressive schools as ‘laboratories for democracy,’ we might think of certain types of video games as ‘laboratories for dromocracy.’ In other words, we can see them as places that teach us to see everything as a resource, to be relentlessly active, and to maximize efficiency at every turn. In Virilio’s terms, they are a concentrated time war experience: a place in which we can locate the disappearance of war into everyday life.» (Waddington, 2020)
Now, I don’t intend to agree explicitly or disagree relentlessly with these scholars—they seem to know what they’re talking about. Again, I don’t want to write academic essays. My goal is to make the fun at least somewhat focused, and that’s why I prefer not merely to play RTS games for the sake of it but complement the play with some sort of reflection.
So let’s forget about the quantitative comparison of system justification between genres. Let’s focus on qualitative system justification within strict mechanical constrictions. We’ll see how different RTS games make horrible things fun. Again, the point is not to devalue the whole genre of games but to provide critical thematic and mechanical analysis while unapologetically having fun while conquering imaginary digital worlds.
Typological Framework
Another problem that people pointed out is that I haven’t strictly defined which games I’ll cover in the series. RTS is an extensive term, and lots of games with entirely different mechanics may be considered real-time strategies. In other words, I need to establish a typological framework and better explain the logic I base my choices of games upon.
Agreed.
Let’s pick a game that is widely considered as typical RTS in it’s purest form. The game that its title is associatively adjacent to the term real-time strategy. This game is going to be an example against which we’ll adjust our criteria.
Christian Elverdam and Espen Aarseth proposed modifications to the games classification model presented in «A Multi Dimensional Typology of Games» by Aarseth et al. in 2003. This model consists of dimensions that describe certain game elements and provides a more systematic and accurate classification than an arbitrary and contradictory genre-based one. It’s also open-ended, which means that we can modify or reject individual dimensions without compromising the model’s integrity.
Elverdam and Aarseth established a pretty comfortable for our needs typological model, so all we need to do is merely “tune” it to our chosen game. That game is Starcraft, by the way—I don’t think there will be objections to the notion of it being considered as an RTS in its purest form.
For ease of use, dimensions make up eight metacategories, the first of which is Virtual Space. There are three dimensions here: perspective, positioning, and environment.
«Perspective describes whether the player has a complete overall view of the game space (omnipresent) or if the avatar (or game tokens) must be moved strategically (vagrant). Positioning describes whether the player can discern his or her position exactly as the game rules dictate it (absolute) or if he or she must relate to other objects to decide his or her position (relative). Environment dynamics describes whether the player is allowed to make additions or alterations to the game space (free) or if such alterations only alter the status of predetermined locations (fixed) or finally if no changes to the game space are possible (none).» (Elverdam & Aarseth, 2007)
Starcraft‘s perspective is omnipresent because the camera is free and can reach any place on the map with a mere click of a mouse. Note that it is irrelevant whether there is a «fog of war» or not—our perspective isn’t bounded to the player’s avatar.
Positioning in Starcraft is relative—there are no coordinates or tile-based system for the placement of units.
Environment dynamics are free—we build base structures in Starcraft. For the widening of the scope, let’s also include fixed as passable criteria—for the games where we can’t choose the exact location of buildings.
The second metacategory is Physical Space. In our case, it’s irrelevant, for we focus only on purely digital games.
The third metacategory is External Time. Its dimensions are teleology and representation.
«Teleology describes if the game ends at a given time (finite) or if it in principle could go on forever (infinite). Representation describes they [sic] way time is represented in the game, either reflecting the way time would pass in our physical world (mimetic) or disjointed from reality (arbitrary).» (Elverdam & Aarseth, 2007)
Starcraft‘s teleology is finite. There is an actual win\lose state at the end of each match or mission.
Time representation is, of course, arbitrary. Structure building, I suppose, takes a bit more time in real-life than in Starcraft.
The next metacategory is Internal Time. Its dimensions are haste, synchronicity, and interval control.
«Haste describes whether the mere passing of real time alters the game state (present) or not (absent). Synchronicity describes whether game agents can act at the same time (present) or if they take turns (absent). Interval control describes whether the players decide when the next game cycle will commence (present) or if such control is denied (absent).» (Elverdam & Aarseth, 2007)
In Starcraft‘s case, both haste and synchronicity must be present, or else the game would not be considered typical real-time. Interval control, though, has to be absent because it suspends haste and compromises synchronicity.
Additional metacategory is Player Composition. It is a one-dimensional category, which describes how players in the game are organized: single-player, single team, two-player, two teams, multiplayer, or multiteam. Everything except single-player is relevant to Starcraft. But it wouldn’t be wise to reject an RTS game’s single-player aspect—they could be some campaign missions that avoid including the player’s adversary and focus on the purely environmental challenges. Still, for the sake of the series’ focus, such instances have to be atypical relative to the questioned game’s normal flow.
Additionally, it is essential to note that defining the game’s dimension as a multiplayer does not necessarily imply the real multiplayer mechanic. Player’s adversaries may be computer regulated. The important thing is for adversaries to be in equal (or reasonably close) conditions as the player and not becoming the environmental hazard (like monster waves in a tower defense game).
Player Relation is next metacategory. Its dimensions are bond and evaluation.
«Bond describes whether the relation between players can change during play (dynamic) or not (static). Evaluation describes how the players or the outcome of the game is quantified. The individual player can be evaluated (individual), the players can be evaluated as a team (team), or they can be evaluated both as a team and as individual players (both).» (Elverdam & Aarseth, 2007)
Although I consider this category not critical for my series and won’t enforce its weight in a game’s characterization, I hesitate to reject it altogether. Evaluation is the least essential dimension—for typical RTS games, it is individual even when players bond into teams. Nevertheless, I will still cover any game which fits every other criterion except this. So, let’s ignore it, what the hell!
Bond is a little bit more problematic. There is a sort of basic diplomacy option in Starcraft—you can ally your opponents, but only if playing regular melee multiplayer mode. That option is blacked out in other play modes and shouldn’t be considered as the core gameplay mechanic. Comprehensive diplomatic options are not typical for RTS games, so I should constrict the bond dimension as static with an opportunity for an exception on a singular case basis.
Finally, the last partly relevant metacategory is Struggle, containing the dimensions of challenge and goals.
«Challenge describes three principal ways a game can provide opposition. It can come in the form of predefined challenges, which are exactly the same each time the game is played (identical). It can come from a predefined framework that is varied by mathematical randomness (instance). Finally, opposition can come from game agents whose actions are autonomous (agent). Goals describe if the game has an exact and unchanging victory conditions (absolute) or if the goals are subjective to the unique occurrences in a specific game or the players’ interpretations (relative).» (Elverdam & Aarseth, 2007)
In Starcraft, challenges may vary—in campaign mode, they are often identical, but in online play, they always come from agents (other players). Thus, we don’t care about how a game challenges the player. We do care about goals, though. They have to be absolute—there is always an exact win\lose situation at the end of each Starcraft match or mission.
There is another metacategory in the presented typology—Game State. It contains dimensions of mutability and savability that describe how game state changes affect players (temporal, finite, infinite) and whether the game state can be saved and restored, respectively. Due to the various gameplay modes in the typical RTS game (campaign, skirmish, multiplayer), those dimensions are irrelevant.
To sum it up, I will cover games that meet all of these criteria:
— Include free or fixed environment dynamics, relative positioning, and omnipresent perspective.
— Have arbitrary external time representation and finite teleology.
— Game’s internal time mechanic has haste stress factor and synchronicity but denies the option of interval control.
— Sets absolute goals for the player.
— Dominantly multiple player composition. The adversary is not environmental.
Additional criteria for consideration:
— Preferably unchangeable relation between players.
For example, Populous fits perfectly to all of the criteria.
Pharaoh fails the criterion of player composition, and the adversary (when present) is environmental.
Those Are Billions also fails the same criterion but additionally has the mechanic of active pause. That’s a big no-no because it is a type of interval control.
Crusader Kings II doesn’t set absolute goals and fails the haste criterion—the player has unlimited time for decision making during a pause.
So now, when the framework is established, it’s time to play some RTSes!
P.S. Although this article is in no way academic, I am still going to cite all of the articles quoted and the book referenced in the text.
Jost, J. T., & Andrews, R. (2011). System Justification Theory. The Encyclopedia of Peace Psychology.
Andrew Baerg. (2009). Governmentality, Neoliberalism, and the Digital Game. Symploke, 17(1-2), 115–127.
Mukherjee, S. (2015). The playing fields of Empire: Empire and spatiality in video games. Journal of Gaming & Virtual Worlds, 7(3), 299–315.
Waddington, D. I. (2020). Time War: Paul Virilio and the Potential Educational Impacts of Real-Time Strategy Videogames. Philosophical Inquiry in Education, 27 (1), 46–61.
Elverdam, C., & Aarseth, E. (2007). Game Classification and Game Design. Games and Culture, 2(1), 3–22.
Bogost, I. (2010). Persuasive games: The expressive power of videogames. Mit Press.
Leave a Reply