Eskelinen about explorations in game ecology

If we were to apply probably the most typical intertextual slogan to games and try to model our approach to them highlighting the idea that a game is a mosaic of other games, we are on the verge of seeing what's missing both from video game theories and the actual scholarly practice not to mention methodology. In short, how should we handle the complex co-presences, imitations, implementations, adaptations and transformations between and among different rule systems, goals, quest structures, tasks, challenges and obstacles, player representations, resources, non-playing characters, game worlds, game engines and platforms etc.? This is the conceptual territory this paper tries to begin to chart by crafting and exploring a general game ecology and its multiple relations to other transmedial modes and entertainment ecologies that also shape and affect individual games.

Introduction

This paper has two simple points of departure. Video games are seen as remediated games, that is, as one species among several other and older ones (such as team sports, board games, card games and toys) within a general and transmedial game ecology. Secondly, gaming is seen as configurative practice - in other words, in art we might have to configure in order to be able to interpret, whereas in games we have to interpret in order to be able to configure, and proceed from the beginning to the winning or some other situation. If and when developed a few steps further these points of departure will have unpleasant or unexpected consequences for the beliefs of the moral majority of game scholars. The existence of a general game ecology (as well as several other ecologies connected to games and gaming) seriously limits the scope and importance of transmedia storytelling (cross-media circulation of narrative content) and it also necessitates the move towards comparative game studies. The position of games relative to other configurative practices indicates that content enters games not only through watered-down story formats but also through other avenues.

While there’s much need and room for academic game studies from multiple traditions (most of them were already in good use in multidisciplinary studies of non-digital games), the actual scholarly practices of today seem to focus on the easiest possible targets [1] (with not so well-hidden agendas). By these I mean the seemingly endless discussions about narrativity on the one hand, and on the other hand, studies tracking down the tiniest visual similarities between games and other forms of audiovisual entertainment as if it is a novelty to figure out how games (or films or novels) use ready-made expectations from other cultural genres and everyday life or to compare film and game representations generated from the same IPR (Intellectual Property Right). These minor overlap and surface studies are of course well in keeping with the usual depth and range of academic work, where the mere continuation of ready-made traditions is good enough.

Ludology vs. overlap studies

Against common, but sadly uneducated, belief the scholars either taken or mistaken to be ludologists (Aarseth; Frasca; Juul; Eskelinen) made an ontological argument about games being a formally different transmedial mode and cultural genre of expression and communication than stories (and then continued their work on more important, challenging, urgent and rewarding games research issues). This is different from saying there can’t be hybrids (still a matter of mere speculation as it has been for the last 20 years or so) or that narrativity can’t enter games, but for reasons beyond me this is still not very well understood.

In my own case I believe one of the most misquoted syntagms from The Gaming Situation (Eskelinnen 2001) is this half-sentence taken out of its context: »stories are just uninteresting ornaments and gift-wrappings to games«. Most commentators seem to have read only this cut sentence and jumped to unwarranted conclusions. However, if one manages to read two whole sentences instead of almost one half, my actual argument goes like this:

The old and new game components, their dynamic combination and distribution, the registers, the necessary manipulation of temporal, causal, spatial and functional relations and properties not to mention the rules and the goals and the lack of audience should suffice to set games and the gaming situation apart from narrative and drama, and to annihilate for good the discussion of games as stories, narratives or cinema. In this scenario stories are just uninteresting ornaments or gift-wrappings to games, and laying any emphasis on studying these kinds of marketing tools is just a waste of time and energy.

Let’s read it: in the second sentence I make a reference to a scenario that is explained in the first sentence. And if one reads that very sentence until its bitter end I’m summarising there my paper-length ontological argument against defining games »as stories, narratives, or cinema«. So: I don’t say there can’t (or shouldn’t) be narrative elements in games, I just say they are not central or interesting in any scholarly sense, and I also give a list of key things that are not explained or even taken up by any sophisticated narrative theory.

So, the real issue with the minor topic of narrativity is how and to what degree narrativity or narrative elements may enter games, as from a purely technological standpoint it would be trivially easy to, let’s say, turn or slice up the whole Godfather trilogy into a series of cutscenes for the next Max Payne game. Neither ludologists nor narrativists ever claimed that games can’t contain narrativity in the form of gradually revealed backstories or that they can’t have narrative sequences in the form of cutscenes. Some games have these, and others don’t. In contrast to these occurrences, all games have rules, variable and quantifiable outcomes and it is also necessary for the player to manipulate and produce events instead of merely interpreting them. Therefore it is natural to think these are and should be the primary features to be studied (especially if you wish to set up and legitimize a new field). Secondly, it is trivially easy to study cutscenes and backstories, which may help to explain why narrativists are not focusing on them and their functional role in digital games (and thus found another new field, that of cutscenology).

Thirdly, if the realm of narrativity is extended much beyond backstories and cutscenes, conceptual difficulties start to rise to the detriment of the narrativist approach (or ideology). As an extreme but illuminating example of the latter we can quote Marie-Laure Ryan from her introduction to the last section of essays in Narrative Across Media: »In my own view a retrospective availability of meaning is sufficient to ascribe narrativity to games.«[2] A retrospective availability of meaning as a criterion? So I guess aphorisms, anagrams, puns, concrete poetry, Gertrude Stein’s a rose is a rose is a rose, t-shirt with logos, shopping and laundry lists, statistics, you name it, all contain narrativity according to Marie-Laure Ryan. For me presenting these kinds of criteria is not an act of scholarship, but a simple and counterproductive act of renaming, which doesn’t solve any of the demanding research issues.[3]

Almost equally absurd are the four ways Henry Jenkins found for narrativity to enter games on a more local level:

Environmental storytelling creates the preconditions for an immersive narrative experience in at least one of four ways: spatial stories can evoke pre-existing narrative associations; they can provide a staging ground where narrative events are enacted; they may embed narrative information within their mise-en-scene; or they provide resources for emergent narratives.[4]

Depending on one’s point of view the first one either confuses narrative associations with useful prior knowledge or prefers narrative associations to all the other possible associations designers can play with; the second one confuses similar events in films and in games without paying any attention to effects of the change of context; the third one either confuses narrative information with all the other types of information or prefers narrative information to them; and the fourth one either confuses games with one possible use of games or prefers it to all the other possible uses.

The contradictions and weaknesses in Jenkins’ essay are to some degree counterbalanced by the orientation he later explained in his response:

I see myself involved in a rather different exercise, attempting not to construct an academic discipline around games, but to intervene in a public debate among game designers, game critics, and game players -- as well as policy makers and other media producers and consumers -- about the current state and future development of an emergent and hybrid form of »interactive entertainment«[5]

Most game designers know a great deal more about the theory of play and games than they know about narrative. Most of the game design books currently on the market tell little or nothing about character and plot. Yet, these practitioners consistently express a hunger to know more about traditional storytelling; sessions on games and narrative have been among the most highly attended at industry conferences. My essay’s arguments came out of that dialogue with the game design community rather than within a more academic context.[6]

Although educating theoretically ignorant game designers as an interactive entertainment lobbyist is a formidable task or vocation in itself, it should not be confused with making a valid scholarly argument in a theoretical debate with game scholars who know their narratologies. Perhaps even more importantly, the ideology of narrativism is at its most visible in Jenkins’ reply: never mind games and game studies, the only important thing is to add more narrativity to »interactive entertainment« as dictated by the dogma and practice of transmedia storytelling, perhaps less misleadingly described as branded content moving between platforms.

Ecologies and economies

It is clear that from the entertainment industry viewpoint everything that could be used for marketing and delivering branded content ultimately based on the IPRs owned by the current corporate giants is a medium. Hence the need (and funding) for comparative media studies as market oriented studies of the use value of both the new and the old commercial channels alike. So far, so good, as this has or should have nothing to do with serious academic scholarship.

As there are different modal contexts and origins (films and games for starters) for cross-media franchises such as Star Wars or Tomb Raider, it is old-fashioned and inaccurate to rename or indiscriminately label these franchise economies and strategies as mere or pure storytelling ecologies or transmedial storytelling; these lazy habits are reminiscent of the bygone era when the film industry was by all accounts bigger than the game industry. It’d probably be wiser to focus on the complex co-existence and multi-centricity of different »ecologies« that are only partly overlapping such as franchise (or content), game, storytelling, and media ecologies. The first three are clearly transmedial whereas the last one is the quite banal home ground of comparative media studies. The basic differences and variations among and within these and some other ecologies are shown in the charts 1a-e below.

Chart 1a shows one possible sample of the most important ecologies and economies that surround, include, and shape digital games. It shows how limited the scope and concept of transmedia storytelling actually is and how insufficient comparative media studies are in handling the multiple connections games have to other ecologies. There’s no reason why independent academic game research should prefer the same connections (or family resemblances) as the industry, focus on the circulation of adapted or co-created narrative content, and be equally surprised every time a major new design idea comes from somewhere else (such as Will Wright’s bestselling software toys up to The Sims and beyond that are the most successful hybrids ever, not of games and stories, but of games and toys). Chart 1b situates games and transmedial game ecology in an alternative branch of economies foregrounding human activities rather than products, processes, modes or media. The rest of the charts show the most common transformations between and within modes on global and local levels with some examples. They show something that is too often either ignored or misunderstood in narrativist similarity studies [7]: when bits and pieces of content move across »media« they change context, function and position which may affect and usually also affects their modal status. In novels and films characters are not the tokens for interaction that they usually become in video games. Likewise, the other stock source for cross-media franchises, the diegesis or the spatiotemporal setting of novels and fiction films changes its status from fiction to simulation when it is transformed into the game world. And in both cases the consumer’s role changes as well, which may be a good enough reason to foreground differences between ergodic and non-ergodic entertainment (and to understand that only within the latter narrative content is central and can move smoothly).

Chart 1a Multiple and multi-centric ecologies

Chart 1b Situating game ecology

Chart 1c Macro level transformations between and within different modes

Chart 1d Micro level transformations between and within different modes

Chart 1e Examples of intermodal transformations from narrative/story elements to game elements

However, if we choose to follow the implicit and underdeveloped thread of intertextuality here and take it seriously, we’ll quickly find ourselves in an uncharted territory. If we were to apply probably the most typical intertextual slogan to games and try to model our approach to games highlighting the idea that a game is a mosaic of other games, we are on the verge of seeing what’s missing both from video game theories and the actual scholarly practice not to mention methodology. In short, how should we handle the relationships, co-presences, imitations, implementations, adaptations and transformations between and among different rule systems, goals, quest structures, tasks, challenges and obstacles, player representations, resources, non-playing characters, game worlds, game engines and platforms etc.? If one takes game studies seriously, these topics should not be avoided, as hard as they may be as research objects.

Compared to sophisticated theories of intertextuality and transtextuality in literary and film studies, we are a long way (probably decades) from establishing even the basics for similar inter- or transludic networks in game studies. One possible place to begin the quest for one of the cornerstones of emerging comparative game studies is the supposed dividing line between digital and non-digital games. This time we’ll go forward by going backwards in time, to 1971 and 1982 especially.

Game ecology and remediated games

In this section we’ll take a quick glance at some finer points of game ecology. First, we’ll try to figure out the most potent differences between digital and non-digital games. Then these findings will be calibrated or balanced with Jesper Juul’s classic game model. This in turn may give us a glimpse of a general game ecology, and perhaps an understanding of how much hard work there is to be done within this vast area before game studies are even close to reaching scholarly maturity.

As we all know Chris Crawford took the differences between games and computer games seriously back in 1982. From his game designer’s point of view Crawford found six major advantages in computer games as compared to games based on other game technologies (Crawford 1982). First, there’s a greater responsiveness to the player’s wishes. This is because »the computer is dynamic; it imposes little consistency on any element of the game.« Second, the computer can serve as game referee. Thirdly, there’s an advantage in real-time play, as »the computer is so fast that it can handle the administrative matters faster than the humans can play the game.« The computer is also able to provide an intelligent opponent and to limit the information given to the player in a purposeful way. And finally, »the use of telecommunications for game play makes possible game structures that are out of reach of other game technologies. It allows us to create games with huge numbers of players.« These advantages were to some degree offset by several disadvantages: the poor I/O rate, single user orientation in the hardware design and the harsh requirement of programmability.

All this is common knowledge now, a history that already happened with advanced single-player games and massively multiplayer online games all around us. Still, Crawford’s list merits another look in order to follow certain trajectories and dimensions it opened and also because it would be arrogant and short-sighted to assume these advantages are already fully exhausted. In hindsight the vast potential of Crawford’s first advantage (the little need for constancy) seems to be the most underdeveloped one while having always been taken for granted by both game scholars and game designers. There’s much more to it than simply catering to the player’s wishes in changing the speed, length, difficulty or rules of the game, but we’ll get back to this point a little later.

Crawford’s artificial referee points in the direction of greater complexity of rules and of game causality in general and therefore also to the possibility that several (over)ambitious non-digital games (such as many social simulation games [12]), that were perhaps felt to be too complex and hard to play because the players and the referees had to upheld the rules, could be more successful if that responsibility could be delegated to the computer. Needless to say telecommunication networks around us have multiplied during the last two decades, and if we were to follow W.J. Mitchell’s insights on networks scaling up and receivers and transmitters scaling down, we could easily see that the advantages this process offers already go far beyond enabling games with huge numbers of players. In addition to mobile gaming and trans-reality games that are capable of connecting, combining, embedding and intertwining virtual and physical playgrounds in ever more complex ways, there are other networks working at closer range that could, among other things, easily transmit the information about the state of the player’s body (pulse, blood pressure etc.) to the game system – thus adding an extra element to be played with that has been lacking from most if not all kinds and types of commercial games and sports so far. Advanced communication networks could also be used to disrupt the magic circle from the outside by letting real-time real-life information change and affect the behaviour of the game elements (rules, obstacles, resources, player positions etc.) in a crucial way while the game is being played [13].

It is also well understood that the nature of hidden and purposefully limited information in computer games is much more dynamic than it could possibly be in traditional games of imperfect information. If we play card games with a standard deck, we always know the possible range of information that is not available to us, but we usually can’t have that same luxury or primitiveness when trying to figure out the quantity, quality and dynamics of what we need to but don’t yet know in computer games. If we compare video games to sports then the non-physical nature of most video games is seen as a major barrier or limitation. As compensation video games can model events that would be plainly impossible in sports. Compared to most complex board games the obvious advantages of video games are based on their automated nature: the player can control an increased number of units while the computer takes care of upholding the rules and calculating the consequences of the players’ decisions.

Two decades after Crawford Katie Salen and Eric Zimmerman [14] saw the differences between digital and non-digital games in less absolute terms. They discerned four special qualities of digital games: immediate but narrow interactivity, manipulation of information, automated complex systems and networked communication. These features are also present in non-digital games, but digital games usually embody them more robustly. Salen and Zimmerman’s criteria are in many ways similar to Crawford’s and in principle they just confirm and repeat his findings. The capability of manipulating, hiding and revealing information overlaps with one of Crawford’s advantages and digital games as automated complex systems contain much of what Crawford says about responsiveness and artificial opponents and referees. Salen and Zimmerman’s fourth point, networked communication, is compatible with Crawford’s last criterion.

This leaves us with immediate and narrow interactivity to focus on (although it mirrors Crawford’s first disadvantage). It foregrounds easily forgotten limitations in the power of a computer to mimic other media and provide a wide range of experiences. Within a general game ecology this dimension could be made a bit clearer by applying Jesper Juul’s (2003) distinction between game adaptations and game implementations. Many physical games and sports can be only roughly adapted to a digital game [15] while most if not all classic board and card games can be digitally implemented without causing significant changes in their rules and possible game states.

If we take narrow but immediate interactivity as a point of departure, we may want to ask what other options there are. How about delayed and broad interactivity and what about the combinations of the four basic possibilities this move opens? Even if the immediate response is deemed as necessary, this doesn’t exclude the question of precisely determined and non-negotiable delayed consequences. By these I don’t mean the usual cumulative long-term consequences, which are ultimately dependent on the player’s strategic skills (or lack of skills) and her adversaries’ actions and therefore negotiable, reversible and uncertain at least to a certain degree. We could call such consequences conditional. In contrast to them, it is the relative novelty of several irreversible and delayed consequences I’m interested in. While it may possible to design a non-digital game that has such consequences, it would be very hard for a human or humans to implement and keep track of them while keeping the game playable at the same time.

According to Salen and Zimmerman’s common sense account (2004: 122-125) the rules that limit the player’s action are supposed to be explicit and unambiguous, fixed, binding, repeatable, and to be shared by all players. There are both designed and non-designed exceptions to these qualities in already existing serious games and critical social simulations, and lots of untapped potential for interesting design diversions. One of the obvious new possibilities inherent in the automated complex systems is that some routine decisions and actions of the player could affect the rules and change them without the player’s knowledge. This is a form of responsiveness that goes against Crawford’s recommended use of it, but it could be very useful in serious games. As digital games are superb in hiding their rules from the players compared to non-digital games, serious digital games could continue seamlessly and without breaks for negotiation even if one team could at some point gain the position from which to secretly alter rules for its benefit.

Perhaps even more importantly the possibility of changing rules in the middle of the game could be used for specifying the differences and similarities between games and other social and cultural configurative practices that are (at least to some degree) only metaphorically games. The more we know about the rules, goals and player effort in the context of games, the more informed and accurate our comparisons of games and gaming to other constraints and limitations of human action: laws, agreements, etiquette, conventions, instructions, orders, obligations, commands, rituals, and obsessive and compulsive behaviour to name but a few. Many of our most important social relations and bonds are based on these conventions that are usually more implicit, multivalent and ambiguous but also less fair, shared and binding than traditional game rules. By allowing the default settings (fixed, binding etc.) of the rules to change based on evaluation of the players’ actions in the course of the game, it may be possible for the multiplayer games to simulate the functioning of the social fabric much better than they do today.

Jesper Juul’s classic game model (2003; 2004) is first and foremost an attempt to build a universal understanding of pre-digital games. In that sense it has two heuristic edges, the second one being its power to show how some digital games differ from the model that is supposed to have held good at least for the last 45 centuries. This is obviously very helpful in our search for the main connections and dividing lines within a general game ecology. After comparing several relatively well-established game definitions Juul discerns their essences and reduces their differences to a system of the following six dimensions:

The rules, the outcome and player effort clearly characterize games as formal systems. The other three are less formal and as such slightly more ambiguous. At first negotiable consequences seem to form a necessary criterion, as it helps to highlight the optional nature of real-life consequences of games. However, it is much less universal than has been believed ever since the pioneering studies and speculations of Huizinga and Caillois. In many ritual and some political settings, the harsh real-time consequences were to be accepted by the players regardless of their willingness to do so. Another small problem with Juul’s model has to do with the player’s attachment to the outcome, as it runs the risk of reducing the range of the player’s personal attachments, motivations, styles, and reasons for playing into the mere outcome or preferring one type of attachment to all the others. Sometimes people are also playing out of boredom and not attaching much or any value to the outcome of the game, and it is equally common that virtuoso players might be more attached to putting their brilliance at display and paying much more attention to the way or the style of their playing than to the outcome (sometimes to the detriment of team play). Still, this is just a minor issue, easily remedied by either adding, removing or modifying one dimension.

The real power of Juul’s classic game model lays in its comparative value and its ability to show and contain exceptions to, modifications of, and advantages beyond the model (including its implicit assumptions). There are plenty of those: games with automated rules leading to more complex rules and decreasing the necessity of the players having to know the rules from the outset; games with no final outcome and with several outcomes from the same saved session; outcomes and consequences of the player’s actions and decisions can be based on data that would not be discernible to human players or referees; no hierarchy among outcomes; increased number of player-controlled units and increased number of players compared to non-digital games; advanced single player games; games with other attachments than to the outcome; games with no specific duration and no specific location.

We shouldn’t forget to take the disadvantages into account as well. There seems to be at least two adaptive barriers within the game ecology. The first one is the limited and narrow physicality one encounters in adapting sports to sport simulations. The second one manifests itself in card games such as poker: psychological manipulation and bluffing can’t be translated into the repertoire of an artificial or a mediated human opponent to a degree that matches the intricacies of the unmediated presence of another human player. While the former barrier may gradually vanish through ubiquitous computing and all the networked sensors and tracking devices that come with it, the latter is less easily bypassed, as it is so deeply rooted in the inherent problems and dead ends of AI research.

In chart 2 several exceptions (derived mainly from Juul and Crawford) to Juul’s classic game model are grouped under the model’s main features into dimensions [16] along which the comparisons between games can be made. There are more relations than just adaptations and implementations, as this distinction seems to concern only whether it is possible to fully copy (or imitate) one game in another game medium (or technology). There are many other possibilities [17]: deliberately partial implementation (leaving out game states that could be implemented), amplified implementation (full implementation and additional game states) and also reductive adaptation, amplified adaptation and augmented adaptation depending on whether the amplified game contains only the game states that could be faithfully mapped, makes qualitative changes in order to increase the relative importance of some core mechanic existent in the hypogame [18] or adds entirely new mechanics to compensate what has to be left out.

Chart 2

rules (manipulation rules, ludus rules, metarules [19]):

variable and quantifiable outcome:

values attached to the outcome:

player attachments:

player effort:

networked communication (y/n; type and range of network; deliberate distortions)

»interactivity« (immediate/delayed and narrow/broad, and different combinations of these)

negotiable consequences:

Conclusion

If we’d want to go any further into the unknown or the poorly explored, our modest model could be expanded by integrating into it dimensions from the well-known model of non-digital games by Redl, Gump and Sutton-Smith (1971). The latter should in principle be compatible with Juul’s model, as they both target non-digital games although from different perspectives. While Juul focuses mainly on formal features, Redl, Gump and Sutton-Smith concentrate on dimensions that »are thought relevant to the behavior that games may provoke.«
[to be continued]

Markku Eskelinen (Helsinki)

Markku Eskelinen
Provosoft
PL 276
00531 Helsinki

(27. August 2005)

References

Aarseth, Espen/Smestad, Solveig-Marie/Sunnanå, Lisa

2004 A Multi-Dimensional Typology of Games. In: Marinka Copier/Joost Raessens (eds.): Level Up – Digital Games Research Conference Proceedings. Utrecht: Utrecht University Press, p. 48-53.

Frasca, Gonzalo

2004 Videogames of the Oppressed: Critical Thinking, Education, Tolerance, and Other Trivial Things. In: Noah Wardrip-Fruin/Pat Harrigan (eds.): First Person – New Media as Story, Game and Performance. Cambridge, Massachusetts: The MIT Press, p. 85-94.

Genette, Gerard

1997 Palimpsests. Literature in the Second Degree. Lincoln and London: University of Nebraska Press.

Greenblat, Cathy / Duke, Richard

1971 (Eds.): Gaming-Simulation. New York: Wiley.

Jenkins, Henry

2004a Game Design as Narrative Architecture. In: Noah Wardrip-Fruin/Pat Harrigan (eds.): First Person – New Media as Story, Game and Performance. Cambridge, Massachusetts: The MIT Press, p. 118-130.

Juul, Jesper

2003 Half-Real. Video Games between Real Rules and Fictional Worlds. PhD dissertation, IT-University of Copenhagen.