Taxonomic Contributions to Digital Games Science

2010 2nd International IEEE Consumer Electronics Society's Games Innovations Conference

Taxonomic Contributions to Digital Games Science Klaus P. Jantke & Swen Gaudl Fraunhofer IDMT Children’s Media Dept. Erfurt, Germany Email: [email protected]

Abstract—Pondering digital games, in general, and discussing details such as, e.g., innovations of NPC behavior and AI, of interface technologies, of serious games application areas, and the like requires a language of scientific discourse. Contributions to a taxonomy of digital games and of digital game playing are essential preliminaries to a digital games science. There is introduced a canonical taxonomy of digital games which is later on extended toward a comprehensive taxonomy of digital games. Practical work shows the usefulness of the approach whereby emphasis is put on serious games development and applications. It is not only for lack of trying that a good vocabulary for describing game experiences does not exist. It is downright hard to describe video games and experience of playing them. (Bruce Philips [1], p. 22)

The GIC 2010 Call for Papers1 is listing a large number of areas of interest usually refined and explained by a larger number of sub-areas. Interestingly, the science of games area of the CfP does occur without any refinement or explanation. Does there really exist anything that might be possibly called a science of games or, to narrow it down to video, computer and console games, a digital games science? Since 2006, the first author is running a special interest group on “digital games science” within the community platform XING2 . These four years of experience reveal enormous deficiencies of terminology when discussing issues of digital games scientifically. Indeed, Bruce Philips [1] is putting it right claiming “that a good vocabulary [ . . . ] does not exist”. At least it does not exist yet. Ever since, science needed a scientific discourse. Every scientific discourse needs an appropriate language. At least, one should be able to ask questions and to formulate hypotheses. The digital games science or what some day might be called the digital games science is nowadays still lacking its language of discourse. It is not yet a science. The present submission is intended to contribute to the future language of discourse on digital games and digital game playing. The approach chosen is a taxonomic one. Terminology is developed by very systematically studying the essentials of game playing. An application to a few problems of serious games and game-based learning demonstrates the usefulness of the efforts. 1 2

http://ice-gic.ieee-cesoc.org/2010/pages/cfp22.htm http://www.xing.com/net/DigitalGamesScience/

978-1-4244-7179-9/10/$26.00 ©2010 IEEE

27

I. I NTRODUCTION To approach a much better terminology than currently available, the authors are focusing on serious games. Opinions about serious games are numerous and largely controversial. For illustration, in their introduction to [2], the authors characterize serious games as “. . . any form of interactive computerbased game software for one or multiple players to be used on any platform and that has been developed with the intention to be more than entertainment.” (ibid., p. 6) When stripping this definition to the essentials by removing tautological utterances such as “for one or multiple players” and “to be used on any platform”, it remains the simple concept of any form of interactive computer-based game software developed with the intention to be more than entertainment. In other words, serious games are determined by some developers intention–a highly esoteric and impractical conceptualization. Developers’ intentions are rarely accessible. The present authors prefer the more pragmatic approach to define a serious game as any game that may be used for more than entertainment. Although nice at a first glance, this approach has the obvious disadvantage that literally every digital game may be seen as a serious game. This coincides with Sawyer’s and Smith’s claim that “all games are serious” ([3], slide 3). However, when every game is a serious one, different games may be, so to speak, differently serious. As already pointed out in [4], there remain several questions such as the following. • How good or how much does the game work seriously? • What is the game useful for? • Whom does the game serve? • In which conditions and how does the game support the serious purpose? Those questions are spanning dimensions of a space in which games may be classified in respect to their “seriousness”. Some believe in the idea of a continuum of seriousness, a view which is a bit oversimplified. To be more precise, it is just a single dimension. When considering digital games, many more aspects have to be taken into account. Some of them interfere with the dimension of seriousness. Understanding such a multidimensional interplay– in some cases, at least–is essential to a better understanding of a game’s psychological and/or social impact. Taxonomies will introduce dimensions of digital games description.


II. P RELIMINARIES TO TAXONOMIC R EASONING Speaking and writing about anything of interest beyond the limits of game play such as game-based learning, e.g., requires necessarily to have an idea of those limits. It turns out that several attempts to determine game play, such as in [5], are surely of historical and cultural interest and may be used for numerous social science investigations, but are–unfortunately–far too uncertain to base upon in-depth discussions of phenomena which surpass the limits.

Fig. 1.

The Approach to Game Play Underlying Taxonomic Concepts

The authors’ approach to game play has been adopted and adapted3 from [6]. It has been investigated in detail elsewhere (see [7], p. 8, e.g.). Those who play willingly engage in an activity different from daily life4 ; they perform an act of framing. Whenever you play, you are facing a certain balance which is expressed in the figure above by means of the dichotomy “indetermination – self-determination”. Indetermination may be caused by built-in randomness, by other human beings, or simply by complexity [8]. To gain control over indetermination, players continuously– and widely unconsciously–learn. The balance of indetermination and self-determination is decisive for learning. So, for letting players experience this balance, we have to set up some appropriate game mechanics. When dealing with digital games scientifically, which side should be emphasized? Shall we focus the intended experience? Or shall we concentrate on the rule-system perspective? Can we do both at once? A taxonomic approach means • to separate issues by disciplinary topics allowing for indepth disciplinary investigations and • to integrate reasoning by lucidly relating issues of rather different disciplines inter- and transdisciplinary. The next two sections sketch the novel taxonomic approach followed by sections demonstrating its application. 3 The author does not agree with J¨ urgen Fritz’ frequently published overemphasis of risk. 4 There are games named pervasive or augmented reality games trying to blur the line between game playing and reality. The movie T HE G AME, David Fincher, 1997, is an impressive cinematic attempt to do the same. Nevertheless, those who play such a game as well as the main protagonist of T HE G AME engage knowingly into play.

978-1-4244-7179-9/10/$26.00 ©2010 IEEE

28

III. P RINCIPAL D IMENSIONS : T HE I LMENAU TAXONOMY Taxonomies are seen as multi-dimensional spaces in which every object of interest–in the present investigation always a game–can be characterized by its values according to the dimensions of the space. In the most simple case, a game is characterized by a point in the space. But in contrast to conventional mathematics, dimensions cannot be assumed to be orthogonal. Consequently, certain games are better imagined as clouds sitting somewhere in the space. The Ilmenau Taxonomy [9] is introducing three dimensions that are canonical. The Erfurt Taxonomy [10] is raising the question whether or not we might need, perhaps, infinitely many dimensions. We may characterize games according to so diverse dimensions like gender specifics, skill requirements, cultural localization, role of music, role of knowledge, time, humour, control feedback, gratification feedback, . . . and the need to exit the game from time to time: extra game play. The present section deals with the principal dimension that establish the so-called Ilmenau Taxonomy5 . The key question of this taxonomy is for those disciplinary aspects that are relevant to every digital game. To every digital game, it applies that • the game is a computer program, • the game is entertainment media, • both in computer science and in entertainment media, there is nothing more interactive than games. Consequently, digital games are investigated with respect to such a 3-dimensional space. D1 The game is seen as a computer program and all its related properties such as the admissible number of users, networking features and the like are considered. D2 The game as a piece of artwork and media has a genre. D3 Through intense interaction, players act and experience engagement such as building, fighting, or trading, e.g. Interaction determines the psychological/social impact. Disciplinary in-depth investigations are setting the stage for well-founded systematic studies of interferences between these three canonical dimensions D1, D2, and D3. The remaining part of the present section about the so-called Ilmenau Taxonomy is devoted, first, to a separate discussion of every canonical dimension and, second, to a brief discussion of mutual interferences of these dimensions. First of all, the introduced dimensions are apparently of quite distinct qualities. D1 This dimensions is totally objective, because it refers to properties of a computer program. D2 This dimension is partially objective, because it refers to genre properties that can be identified by having a closer look at the content of a game. However, the way in which the genre is appreciated is largely subjective. D3 This dimension is mostly subjective. It refers to the individual experiences perceived when playing a game. 5 The name refers to the Thuringian town Ilmenau, where the basics of this taxonomy have been published as a technical report of the Ilmenau University of Technology in 2006 [9].


This briefly sketched distinction of the three canonical dimensions shall be considered more in-depth. Dimension D1 Whether or not a digital game may be played online, for instance, is a typical property of the corresponding computer program. This is an objective fact which does not depend on any subjective interpretation. Similarly, the number of admissible players is a fact depending on properties of the given program and, thus, is objective. For illustration, there does exist the quite prominent term “Massively Multi-Player Online Role Playing Game” (MMORPG, for short) rather frequently misjudged as a genre. What a big confusion. In this term, “MMO” is clearly a D1 characteristics of the particular computer program, whereas “RPG” is a certain D3 characteristics (see below). In “MMORPG”, there is no D2 characteristic–the game’s genre in its proper sense– mentioned. Dimension D2 Given a game, one may ask for virtual time and virtual location of the game play and, thus, for peculiarities of the atmosphere human players may expect. For illustration, the game may play in a fantasy world with elves, magicians and the like. In contrast, the action may take place in some future science fiction environment or, more specifically, in the near future. Time and location of the game play may be an antique world such as the Roman empire or some romantic world which appears like some “good old days”. The games of the AGATHA C HRISTIE series and of the S HERLOCK H OLMES series are of the latter appearance, but also particular games such as BLACK M IRROR and N IBIRU. The game genre in this narrow sense is widely objective, but the way in which it is perceived and possibly appreciated by a human player is highly subjective. Dimension D3 What particular human players experience when playing a particular game is highly subjective. It depends on personal history and preferences, on current conditions, and possibly on the player’s mood. Consider, for instance, the controversially discussed game D EAD R ISING. For completing the game, you have to kill not only hundreds, but thousands of virtual zombies. For some players, this may be experienced as an orgy of slaughter. Other players, e.g., the authors, who have been very much interested in the underlying story, experienced something quite different. In other games, players may experience a different behavior such as, problem solving or trading. But if the problems are too tough or the trading is too boring, you don’t experience the flow and, thus, you are not touched by the game. The key balance of indetermination and self-determination discussed in section II may be one of the reasons for the largely different subjective experiences of game playing. As discussed so far in this section, the three canonical dimensions of the Ilmenau Taxonomy are widely independent. This makes them appear as three orthogonal dimensions of a conventional 3-dimensional vector space. But media science is not mathematics; the assumption of really independent dimensions is incorrect. Orthogonality of the dimension of a taxonomy may be seen as follows. Take a particular game which has certain charac-

978-1-4244-7179-9/10/$26.00 ©2010 IEEE

29

teristics along every dimension. Modify the game such that its characteristics along one and only one of these dimensions slightly changes. This should not have any impact on the game’s description along the other two dimensions. Let us investigate a few illustrative cases of interference. Case 1 Consider a particular role playing game. Role playing is a feature of dimension D3. Many human players like to identify themselves with a virtual character and really experience the development of their character in a game as their own success–they live the role. Interestingly, there are different types of those who like digital role playing. Some may be completely satisfied by successful role playing in single-player games such as the classic BALDUR ’ S G ATE. Others, however, enjoy role playing only if they are aware of the fact that other humans are watching and recognizing their achievements. Consequently, if you slightly change a particular role playing game along dimension D1 from a multi-player game to a single player game, for some players this might change the game along dimension D3 in the sense that they no longer experience the pleasure of role playing. Case 2 There is a rather intriguing relationship between dimension D1 and dimension D2 in the area of pervasive games. Pervasive games6 are those that need, to some extent, the real world for game playing. When playing a pervasive game, human players are equipped with some digital technology. They may use their mobile phones or PDAs, or they may be loaded with special hardware such as head-mounted cameras and glasses. In the game E PIDEMIC M ENACE, e.g., heavily loaded players run through a real environment to fight virtual viruses displayed to them. The game has been played only a few times. It seems that the genre, i.e. the game’s D2 characteristics, may require enormous technical installations, i.e. D1 effort. Case 3 Many human players, are very sensitive to genre characteristics (dimension D2) of a game. Thus, their experience when playing a digital game (dimension D3) is highly affected. Humans who consider magics complete nonsense have no proper opportunity of immersion in games like T HE W ITCHER or W ORLD OF WARCRAFT. Those who dislike erotics in digital games have no chance to enjoy problem solving in L ULA 3D or to playfully engage in fighting in S OUL C ALIBUR IV. Instead of S OUL C ALIBUR, C ASTLEVANIA may be more acceptable. To sum up, although the three dimensions of the Ilmenau Taxonomy are canonical, i.e. relevant to every digital game, and basically different, there is a variety of potential interferences. On the one hand, it is methodologically justified to consider these dimensions separately. On the other hand, one should always carefully look for possible interdependencies. 6 The potentials of pervasive games have been badly overestimated. On the Web page http://idw-online.de/pages/de/news125867, a self-proclaimed game specialist forecasted that by 2008 pervasive games will be as popular as visiting a cinema.


IV. G AMES WITHOUT L IMITS : T HE E RFURT TAXONOMY The idea of the Erfurt Taxonomy7 as layed out in [10] is to systematize–beyond the canonical dimensions of the Ilmenau Taxonomy–all those aspects that may be relevant for the one game, but completely irrelevant to another one. For illustration, music is playing a crucial role in several interesting digital games. In contrast, there are numerous other games that need no sound at all. Thus, the role of music or, simply, music is a typical dimension of the Erfurt Taxonomy. From the Erfurt Taxonomy, two dimensions are selected for a slightly more in-depth discussion in the present brief section: extra game play and meta game play. A. Extra Game Play The idea of extra game play concepts has been introduced in [10] using notations and concepts such as “Extra-Game Communication” (ibid., p. 9). It applies to the majority of digital games that playing the game is somehow dovetailed with certain activities that are clearly not game play. For illustrating the authors’ ideas, let us start with a simple real-world example. Assume some little boy playing football and being watched by his parents. When the little boy is looking for his parents to see whether they pay attention to his challenge, this is an extra-game activity. Nevertheless, the feedback he receives may encourage him and, effect his actions, thus, game play. Apparently, what is considered to be the step from game play to extra game play depends on the one hand, on the line between game play and real life8 determined by framing and, on the other hand, on the investigator’s decision of what to take into account when discussing some player’s activities9 . B. Meta Game Play Interactive storytelling is one of the challenging sources of the meta game play concept. A high ambition has been summarized by Andrew Stern [12] claiming that “the first wish that most players, developers and researchers originally feel when first encountering and considering interactive story, is the implicit promise to the player to be able to directly affect the plot of the story, taking it in whatever direction they wish.” (ibid., p. 20) When you play a particular game and when you make your hopefully influential decisions, how can you ever be sure that you really affect the plot of the story? You can not! There is a variety of phenomena which make only sense with reference to different game plays. The most simple and, very likely, most popular phenomenon of meta game play is the high score concept. One of the most involved meta game play problems relates to interactive storytelling. There is the 7 The name refers to the Thuringian capital Erfurt, where the basics of this taxonomy have been published as a technical report of the Erfurt Children’s Media Department in 2009 [10]. 8 There is the game category of Pervasive and Augmented Reality Games which, on purpose, blurs the borderline. 9 This is an issues of Layered Languages of Ludology [11].

978-1-4244-7179-9/10/$26.00 ©2010 IEEE

30

extremely simple question whether or not two game players experience the same story. The question only makes sense when both players play somehow differently. The not so complex conventional point & click adventure S HADOW OF D ESTINY, for illustration, has six quite different endings. Some of them mean mutually very different stories. A correct in-depth discussion of those meta game play phenomena requires a terminology that allows for the comparison of varying game playing experiences and that supports expressions to describe both common features and substantial differences, respectively. C. Extra Game play vs. Meta Game Play Formalizations sometimes help to make fine distinctions explicit. Through the remaining part of this section, they are introduced to clearly separate the two taxonomic dimensions under consideration from each other. All the subsequent conceptualizations refer to a particular digital game G. On a certain level of description, there is the set of all potential actions and interactions taken into account. The set is called M . Unfolding game play is formally represented by sequences of M . The set of all those sequences is M ∗ . For simplicity, we assume that there is some concept such as a completed game play. Some sequences of M ∗ represent completed game play whereas others do not. All completed game plays are collected in a set Π(G) ⊆ M ∗ . These are the minimal formalizations10 which may be exploited in rather different ways (see, e.g. [14]). For the purpose of the present submission, the terminology is used to have a look at the extra game play and the meta game play concept from another point of view. As we have briefly mentioned, it depends on the perspective what is considered to belong to game play and what goes beyond. Assume a particular view, i.e. assume a certain collection of relevant (inter-)actions M . Let us consider a certain game play π ∈ Π(G). π may be seen as μ1 μ2 μ3 . . . μn (every μi ∈ M ). Extra game play occurs in case some player, while playing π, interrupts the sequence of (inter-) actions by performing some “foreign” action a. This may be written as μ1 . . . μk a μk+1 . . . μn for some index k (1 ≤ k < n). It holds a ∈ M . There is an obvious problem of conceptualization. One might extend M to cover a as well. This is particularly challenging in case of augmented reality games. When players need to experience more than only one game play, we are apparently facing meta game play issues. In formal terms, we deal with π1 , π2 ∈ Π(G), at least. The terminology points to an enormous amount of clearly addressable problems, e.g., the question of common substrings of π1 and π2 . Differences are similarly interesting. Very different formalizations relate to extra game play and meta game play, resp., leaving no doubt that these dimensions are substantially different. 10 The set Π(G) is a formal language in accordance with classical computer science conceptualizations [13]. This insight has a variety of implications.


V. TAXONOMIC C ONCEPTS AND R EASONING AT W ORK Novel concepts shall be practically validated. ”The proof of the pudding is the eating of the pudding.” This section aims at a demonstration of the usefulness of the introduced taxonomies. Three subsequent subsections A, B, and C put emphasis on different aspects and demonstrate the usefulness of taxonomic concepts and taxonomic reasoning in different ways. The key digital games used for illustration are G ORGE, S ABOTEUR, and S HADOW OF D ESTINY.

ellipses. The robots of this team appear somehow altruistic and help others to proceed in the game. The way in which different NPC teams behave can be controlled by the human players (figure 2).

A. Learning Artificial Intelligence in and with Gorge G ORGE is a game which has not been developed as a game, but as a research tool. Used as a game, G ORGE tends to propagate a certain technology competence [15] [16]. Meta game play is an appropriate concept for investigating the way in which G ORGE works.

Fig. 3. Altruistic NPC Behavior in G ORGE–Look at the Encircled Characters

Fig. 2.

The Player’s Control of the NPCs’ Artificial Intelligence in G ORGE

In [17] it has been demonstrated how human players can determine some behavioral characteristics of their computerized adversaries in G ORGE. Those Non-Player Characters (NPCs) can be equipped–on purpose–with a certain degree of Artificial Intelligence (AI). The value of AI for an interesting game play can easily be recognized when playing the game several times with varying NPC settings–clearly an issue of meta game play. Let us briefly discuss a few features of the game G ORGE to illuminate the relevance of AI and to explain the importance of playing several times. In G ORGE, all players move their little robots along different paths from the starting area in the left upper part of the screen to the target area next to the rocket in the right lower corner (see figure 3). There are deep gorges interrupting the paths. A gorge can only be crossed when one of the robots steps down into the gorge voluntarily. In figure 3, two robots of the same team who stepped down into a gorge are indicated by surrounding

978-1-4244-7179-9/10/$26.00 ©2010 IEEE

31

When human players set the behavior of NPC adversaries on purpose, they usually have certain expectations concerning the effects of their set-up. For illustration, if you set up two of the other robot teams to be particularly aggressive aiming at jostling each other whenever possible, you may expect to literally see how the aggressive behavior arises during game play. To check this, you need to play the game, perhaps, several times. If the unfolding game play does not meet your expectations, you may change the set-up and try it again. Meta game play is obviously a key feature. As a result of meta game play, young subjects who have not had any previous knowledge about or experience with Artificial Intelligence gained, to some extent, specific technology competence. A related qualitative study [16] revealed an understanding of basic principles, some pride of being in control of technology, and a certain appreciation for the value of AI in applications. In contrast to extra game play, it is not sufficient to play the game only one or two times and to exchange ideas with others. When playing GORGE, for an effect of competence gain in the field of AI, players need to play the game a larger number of times. In the experiments reported in [16], subjects have been playing the game about 10 to 20 times or even more. Digital games, in general, and serious games, e.g., G ORGE, in particular, may be classified along the taxonomic dimension of meta game play. This dimension is apparently not linear. The investigation may be deepened by asking for purpose and didactic principles underlying the meta game play. It is an independent and particularly interesting area of investigation to analyse so-called pervasive games from the view point of meta game play (or, similarly, extra game play). Unfortunately, this goes beyond the limits of the present paper.


B. Control Feedback vs. Gratification Feedback vs. Humour in Different Digital Games ranging from S ABOTEUR to A NKH The high degree of interactivity of digital games is bringing with it a particularly large amount of feedback for the player. At a first glance, the form of feedback is overwhelmingly diverse and seems to evade any systematization. The Erfurt Taxonomy proposes a classification of feedback with respect to the role it is playing to the human player. There are two basic variants of feedback given by the game: control feedback and gratification feedback. Every digital game responds to the human player’s actions in many ways. Game response which is useful to advise the player in how to play successfully is called control feedback. This type of feedback is widely objective, because it refers to the game mechanics. There does exist an enormous manifold of appearances of control feedback ranging from heads-up displays that show relevant data to virtual blood splatters on the screen that inform the player about some recent harm. Much digital game response aims at the human player’s satisfaction. In response to successful actions and to solving quests, players may score points, get objects such as weapons, healing potions, treasures, or information. Players may be awarded prestige, ranks, skills, and the like. Particular awards have the form of music, pictures11 , or videos. According to the Erfurt Taxonomy, these phenomena are summarized under the dimension of gratification feedback. Apparently, such gratification feedback is highly subjective. It may easily be that the one player enjoys certain gratification feedback very much, whereas others do not. At a first glance, it seems that these two forms of feedback are clearly discriminated by the feature of being either largely objective or essentially subjective. But a closer look reveals that this is not the case; the dimensions of game taxonomies are rarely mutually orthogonal (in contrast to vector spaces in mathematics we are used to since our school days).

The central questions addressed in the present section are: How may control feedback and gratification feedback interfere? How may both a separation and a consideration of the mutual impact of these two dimensions lead to a better understanding of game features? May the consideration of these dimensions lead to new insights? As a case study, let us consider the World War II adventure game S ABOTEUR. You begin playing the game in a largely black and white virtual world. When you have resolved the first problem of blowing up your enemies’ fuel depot, color returns slightly to the game. With more and more tasks completed, the virtual game environment becomes increasingly more colorful, bright, and friendly such as the right screenshot in figure 4. In such a way, the game system responds to the human player’s actions. How would you name this type of feedback? At a first glance, this is clearly some gratification feedback. During game play, human players will experience that in brighter and more colorful parts of the virtual environment, supportive non-player characters are around ready for certain cooperation with the player. The coloring of the environment turns out to be quite useful control feedback as well. As to possible new insights, e.g., imagine any variant of a well-designed and well-implemented gratification feedback. Due to the principal subjectivity of perceiving gratifications, there will definitely be those players who appreciate this type of gratification feedback, whereas several others do not. One may classify players with respect to their reception of gratification feedback. This may be one more access point to the understanding of game playing impact and to the success of digital games on the market.

Fig. 5.

Fig. 4. S ABOTEUR –At the beginning, virtual Paris is largely black and white (left screenshot) and turns step by step colorful when enemies are defeated. 11 In T HE W ITCHER , those interested in may “win the heart of a woman” and receive an erotic picture of her, whereas in the perhaps a bit irritating S UPER C OLUMBINE M ASSACRE ROLE P LAYING G AME (in which you replay the massacre at the Columbine High School, Littleton, CO, USA, in 1999), the completion of the game is rewarded by an original photograph of the bodies of the two dead crazed gunmen–somehow a matter of taste.

978-1-4244-7179-9/10/$26.00 ©2010 IEEE

32

A NKH – Fake Energy Indicator only seen as a Joke

Finally, let us have a brief look at another form of feedback as shown in figure 5. In this scene of the game A NKH, there suddenly appears a certain life energy indicator (right upper corner). This, however, is no control feedback, but simply a joke. In the game A NKH, your avatar can never die. The game designers and developers are kidding. To sum it up, understanding the difference and the potential mutual impact of control feedback and gratification feedback is relevant to understand digital games. These taxonomic dimensions lead to exciting new research question such as, perception-based player categorization.


C. Extra Game Play, Meta Game Play, and Manipulation of Time in the Point & Click Adventure S HADOW OF D ESTINY When we introduced the taxonomic dimension of meta game play above, we have been citing Andrew Stern [12] who sketched the dream of interactive digital storytelling aiming at “the player to be able to directly affect the plot of the story”. Current digital games practice is far from meeting Stern’s expectations. However, a few games allow, at least, for a few different stories to unfold during game play. With a clear story concept at hand [18], it is not difficult to describe alternative stories in sufficient detail. The authors have chosen a rather simple point & click adventure of the genres “good old days” (see section III above), mystery, and survival/horror named S HADOW OF D ESTINY. The game S HADOW OF D ESTINY is of a particular interest, because it has six substantially different endings12 such that from a certain point of game time and in accordance to a player’s behavior unfold substantially different stories. It is a matter of dramaturgy that certain essentials of a story become only obvious when close to the end of the game. The central question addressed in the present section is: How do human players experience the freedom of driving the game play into different directions? Which game features are appropriate and in use to enjoy the manifold of storytelling? Taxonomic concepts and reasoning shall be invoked. Essentials of the unfolding story are (i) that the player (in fact, his avatar) frequently falls victim to a murder (see figure 6, left screenshot) and (ii) has to time-travel backwards (figure 6, right) to take preventive measures for the next time.

However, there are some apparent limits of experiencing alternative stories exclusively by means of virtual time travel.

Fig. 7.

In a large number of cases, the player’s decisions have some decisive impact on the unfolding story and on the remaining reachable endings of the game. For illustration, on some of your time travels back to the middle ages, you meet a person named Margarete (see center of figure 8). When playing the game, the suspicion arises that– in this virtual world–Margarete is your (avatar’s) ancestress. In dependence on whether or not you tell Margarete this guess, the game story branches and you can reach either the one half or the other half of the endings, respectively.

Fig. 8.

Fig. 6. S HADOW OF D ESTINY –Playing the game, players travel back to some time point prior to their own virtual death to avoid the disaster the next time.

In this game, the ability to manipulate game time and, in particular, time travel back and forth are the key to experience the player’s ability to drive the story into different directions. In this way, players may get some impression of alternatives. The game offers opportunities to consult a fortune teller (depicted in figure 7) for orientation in the virtual space of time; this advice may help the human player to act successfully. 12 In fact, if you play the game several times and you complete all endings, you will get access to more content and–in some sense–to even more endings. We confine ourselves to the six ‘regular’ endings. There is no space to go into further details.

978-1-4244-7179-9/10/$26.00 ©2010 IEEE

33

S HADOW OF D ESTINY – Taking some Advice from a Fortune Teller

S HADOW OF D ESTINY – Meeting Margarete back in Virtual 1580

But the importance of those and several other branching points is not revealed during game play. Human players have no chance during regular game play to recognize when and how they are effectively driving the story into the one or the other direction and, thus, cannot enjoy the freedom they have. As to those alternatives, when the game is over, you have no idea of what you missed. In practice, extra game communication substitutes for what the game cannot offer. Players discuss in much detail–largely based on the Internet–where branching points are hiding and how to explore the manifold of possibly unfolding stories as a whole. The game design enforces meta game play, i.e. completely playing the game several times. There is no other way to access the additional “bonus” game mode which, again, has more than one outcome. To sum up, S HADOW OF D ESTINY is surely a great game, but the in-game feature of time travel is not sufficient for experiencing the impressive space of potential stories. Features of extra game play and meta game play are inevitable.


VI. S UMMARY AND C ONCLUSIONS On a conference dealing with Games Innovations one might expect to find innovative ideas of game mechanics and game play, innovative interface devices, innovative physics engines and visual effects and, perhaps, revolutionary ways to apply and to exploit game playing. But terminology, just words . . . ? A. Summary Whatever appears to be innovative or even revolutionary, on a scientific conference, we need to talk about the issue. More precisely, • • •

authors need to present their ideas, approaches, implementations, and experiences, the audience needs to be able to ask questions and to express guesses, doubts, and hypotheses, and all of them need an in-depth debate.

The digital games science needs some language of discourse. As Bruce Philips put it critically [1] (citation above), there is a severe deficiency in the field of digital games terminology. From this perspective, even if the present contributions to digital games taxonomies do not appear too much innovative, they are, at least, a necessary prerequisite to every scientific treatment of games innovations. And the taxonomic concepts introduced work in practice. Consider, just for illustration, the essentials of changing colors in the World War II point & click adventure S ABOTEUR which is investigated in section V-B. Most game magazines reporting about S ABOTEUR have not been able to characterize this game feature under discussion. They use descriptions such as, e.g., “an interesting stylistic device”. What a misinterpretation. B. Outlook The authors admit that his present approach to digital games taxonomies as developed in [9] and [10] leaves several interesting questions open. Digital games taxonomies are still in their infancy. Consider, just for illustration, the varying pattern concepts ranging from [19] to [20]. In an extreme, every individual pattern (see [21] for a logically precise pattern concept) may be seen as determining a particular taxonomic dimension. Let us consider a very specific pattern. In games such as C HESS, there exist the concept of zugzwang13 . Both in classical C HESS and in modern digital games, zugzwang means a situation in which a player in enforced to perform a very particular unique action. This concept is very useful to distinguish certain games. Whereas in T HE S ECRET F ILES : T UNGUSKA zugzwang is essential to the mastery of the game, it is the key reason for frustration when playing W INTERFEST. There opens a wide field of future research and development. An interference of patterns research and taxonomic research will surely lead to new perspectives and new insights–a slight indication of this taxonomic approach’s innovative potentials. 13 Some

English topical literature on C HESS is adopting this German term.

978-1-4244-7179-9/10/$26.00 ©2010 IEEE

34

ACKNOWLEDGMENT The present research and development has been partially supported by the Thuringian Ministry for Education, Science, and Culture (TMBWK) within the project iCycle under code PE-004-2-1. R EFERENCES [1] B. Philips, “Talking about games experiences – A view from the trenches,” interaction, pp. 22–23, September/October 2006. [2] U. Ritterfeld, M. J. Cody, and P. Vorderer, Eds., Serious Games. Mechanisms and Effects. New York: Routledge, 2009. [3] B. Sawyer and P. Smith. (2008, February) Serious games taxonomy. http://www.dmill.com. [Online]. Available: www.dmill.com [4] K. P. Jantke, “Toward a taxonomy of game based learning,” in Proc. Intl. Conference on Progress in Informatics and Computing, Shanghai, China, Dec. 10-12, 2010 (in print), 2010. [5] J. Huizinga, Homo ludens. A Study of the Play Element in Culture. London: Routledge & Keagan Paul, 1949. [6] J. Fritz, Das Spiel verstehen. Eine Einführung in Theorie und Bedeutung. Weinheim & München: Juventa, 2004. [7] K. P. Jantke, “Serious Games – eine kritische Analyse,” in 11. Workshop Multimedia in Bildung und Wirtschaft, Ilmenau, Sept. 20-21, 2007, R. of TU Ilmenau, Ed., 2007, pp. 7–14. [8] C.-P. Schnorr, “A unified approach to the definition of random sequences,” Mathematical Systems Theory, vol. 5, no. 3, pp. 246–258, 1971. [9] K. P. Jantke, “Eine Taxonomie für Digitale Spiele,” TU Ilmenau, IfMK, Diskussionsbeiträge 26, Dezember 2006. [10] ——, “Taxonomien für Digitale Spiele: Von Ilmenau nach Erfurt,” Fraunhofer IDMT, Abtlg. Kindermedien, Report KiMeRe-2009-02, Dezember 2009. [11] C. Lenerz, “Layered languages of ludology – Eine Fallstudie,” in Digitale Spiele – Herausforderung und Chance, A. Beyer and G. Kreuzberger, Eds. Verlag Werner Hülsbusch, 2009, pp. 35–64. [12] A. Stern, “Embracing the combinatorial explosion: A brief prescription for interactive story R&D,” in Proceedings of the 1st International Conference on Interactive Digital Storytelling, ser. LNCS, U. Spierling and N. Szilas, Eds., vol. 5334. Springer-Verlag Berlin Heidelberg, 2008, pp. 1–5. [13] J. E. Hopcroft and J. D. Ullman, Introduction to Automata Theory, Languages, and Computation. Addison-Wesley, 1979. [14] K. P. Jantke, “Dramaturgical design of the narrative in digital games: AI planning of conflicts in non-linear spaces of time,” in IEEE Symposium on Computational Intelligence and Games. IEEE Press, 2009. [15] ——, “The Gorge approach. digital game control and play for playfully developing technology competence,” in CSEDU 2010. 2nd International Conference on computer Supported Education, Proc., Vol. 1, Valencia, Spain, April 7-10, 2010, J. Cordeiro, B. Shishkov, A. Verbraeck, and M. Helfert, Eds. INSTICC, 2010, pp. 411–414. [16] K. P. Jantke, I. Hoppe, D. Lengyel, and A. Neumann, “Time to play Gorge – Time to learn AI: A qualitative study,” in eLearning Baltics 2010, Proc. 3rd Intl. eLBa Science Conference, S. Hambach, A. Martens, D. Tavangarian, and U. Urban, Eds. Fraunhofer Verlag, 2010, pp. 99– 110. [17] S. Gaudl, K. P. Jantke, and C. Woelfert, “The good, the bad and the ugly: Short stories in short game play,” in Proceedings of the 2nd International Conference on Digital Storytelling, ser. LNCS, I. Iurgel, N. Zagalo, and P. Petta, Eds., no. 5915. Springer-Verlag Berlin Heidelberg, 2009, pp. 127–133. [18] K. P. Jantke, “The evolution of story spaces of digital games beyond the limits of linearity and monotonicity,” in Proceedings of the 2nd International Conference on Digital Storytelling, ser. LNCS, I. Iurgel, N. Zagalo, and P. Petta, Eds., no. 5915. Springer-Verlag Berlin Heidelberg 2009, 2009, pp. 308–311. [19] S. Björk and J. Holopainen, Patterns in Game Design. Hingham, MA, USA: Charles River Media, 2004. [20] K. P. Jantke, “Patterns in Digital Game Playing Experience Revisited: Beiträge zum tieferen Verständnis des Begriffs Pattern,” TU Ilmenau, Inst. Media & Comm.-Sci., Diskussionsbeiträge 33, January 2008. [21] ——, “Logical formalization and reasoning for computerized interactive storytelling,” in Proc. Intl. Conference on Progress in Informatics and Computing, Shanghai, China, Dec. 10-12, 2010 (in print), 2010.