The Good, the Bad and the Ugly: Short Stories in Short Game Play

What are the minimal requirements to tell a story by means of some digital game? What are minimal short stories worth to be told? How to establish the minimal preliminaries of interactive storytelling? What do we expect of computerized counterparts that are worth to be mentioned in a story humans experience when playing a digital game? The authors have designed, implemented and experimentally used some particularly simple game in which the personalities of digitalized agents– the good, the bad and the ugly–substantially contribute to stories that may be worth being reported. . . . 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. (Andrew Stern in [1] p. 3) 1 Conceptualization of Stories in Digital Game Playing Does it make any sense to speak about the story that evolves when playing the conventional board game of Nine Men’s Morris? Do you experience anything such as a story when playing Tetris? Conceptualization is a matter of taste and, in particular, a matter of purpose and necessity [2]. What does it buy us to talk about stories of falling blocks in Tetris? In particular, how does the storytelling perspective contribute to the design of a digital game? How does it contribute to the understanding of affect and social impact? What is seen as a story and what is called a story depends on the focus of investigation, on the interest in game design or analysis. The present research aims at an identification of, so to speak, atomic building blocks of storytelling. What are minimal ingredients to make stories emerging? Particular emphasis is put on the design of non-player characters’ profiles. How much “personality” is necessary to play some role in an interesting story? The authors’ original story space concept [3] will be exploited to set the stage for non-player characters strong enough to drive an interesting story evolution. I.A. Iurgel, N. Zagalo, and P. Petta (Eds.): ICIDS 2009, LNCS 5915, pp. 127–133, 2009. c © Springer-Verlag Berlin Heidelberg 2009 128 S. Gaudl, K.P. Jantke, and C. Woelfert 2 The Impact of NPC Characters on Alternative Stories The authors central hypothesis is: Character is crucial to interactive storytelling. In the most simple approaches, stories are seen as linear sequences of events, as expressed by David Thue et al., recently: Fundamentally, stories are sequences of events, each of which involves some form of action. ([4], p. 231) Other authors such as Marie-Laure Ryan, e.g., strive to abandon the linearity of stories: The trademark of the epistemic plot is the superposition of two stories: one constituted by the events that took place in the past, and the other by the investigation that leads to their discovery. ([5], p. 7) Her epistemic plot is leaving linearity behind, but the story concept itself still sticks with the principle of sequentiality. In contrast, the present authors propose to abandon strict linearity of events. Designing a storytelling game means the anticipation of varying experiences including events which possibly exclude each other. This very much resembles planning in dynamic conditions [6]. The complete design may be presented as a storyboard [7] potentially summarizing a number of mutually exclusive stories. When human game playing happens, it unfolds one of the possibly many stories. Due to the lack of space, further details have to be suppressed. The varying decisions of a human player drive the emergence of possibly different stories. But how can purposeful game design–seen as dramaturgical design–predetermine and direct the players’ decisions, thus, leading to a variety of stories foreseen by the designers? It is the authors’ intention to answer this question with an approach stripped to the essentials. Just varying NPC profiles lead to experiencing different stories. Fig. 1. Gorge Agent Settings The case study outlined in the following section reports the authors’ design and implementation of a minimalist approach in which easy to control NPC characteristics lead to quite different interactions between humans and NPCs experienced accordingly. Game play is short, and so are the stories. In the game Gorge under consideration, human as well as computerized players have to move along a track to reach a certain goal area. The track is interrupted by gorges. Facing a gorge, you may learn about the other players’ peculiarities of cooperating for passing the gorge together or defecting. By means of slide controls, the NPCs’ preferences can be adjusted. Players may turn an NPC either into a good or a bad one. The short stories players are going to experience depend on the characters of the NPC they meet and approach during game playing intentionally or by chance. The Good, the Bad and the Ugly 129 There may arise a short story of fighting, tricking, fooling, and the like or, alternatively, variants of consideration, help, care, and waiting for each other. It’s all a question of character. The adjustment of characters directs the story. 3 A Case of Meeting Digital Adversaries and Friends In 2006, for the purpose of scientific investigations, the second author of this paper has designed some particularly simple game named Gorge which has been implemented in dozens of variants within several lectures on digital games. 3.1 The Gorge Game Concept Gorge resembles a board game. A typical Gorge board can be seen on figure 2. Up to four players play in turns, trying to get their four pawns from the start fields in the upper left corner to the final fields at the bottom center of the board. Fig. 2. The Gorge Board in an Early Students’ Design and Implementation A player’s turn begins using the dice to see how far a pawn can be moved on the board. Since it is very important to reach the final fields faster than the competitors, larger numbers from dicing mean better results. A player chooses from one of its pawns and moves it using up the range indicated by the dice result. It is allowed to move a pawn to a field that is already occupied by another pawn, especially those of the other players. That pawn is ordered to retreat to the next vacant position. This action is called “to jostle”. Since jostling other players’ pawns puts a pawn ahead jostling can be utilized as an aggressive tactic that helps winning the game. But it might be also a bit time-consuming. Paths on the Gorge board are intersected by gorges, here denoted by the brown color. A gorge cannot be passed by a single pawn but only by building roped teams, requiring the two fields preceeding the gorge field must not be 130 S. Gaudl, K.P. Jantke, and C. Woelfert vacant. Once a team is built, one of that team’s members can move into the gorge. As long as there is a pawn, other pawns can pass that gorge. When a pawn is passing a gorge field it rescues the pawn from the gorge field, but for that rescue to take place there needs to be a vacant field after the gorge field and before the destination field of the passing pawn. The rescue action is omitted if there are no such vacancies. Gorges are the crucial places where players learn about their competitors of being the good, the bad or, perhaps, even the ugly. Characters becomes decisive. 3.2 The Gorge Agents’ Implementation The Gorge board is modelled using a linked sequence of fields with predecessors and successors. The ownership relations are represented by the current GameState. The linked structure is invariant in a game but the GameState changes. Artificial agents can be set up as players with characters favouring different strategies by introducing likes and dislikes for events defined by the game rules. public double evaluateNode(int[] playerConfig) { double value = jostling * playerConfig[GorgeKI.JOSTLING] + ropeTeaming * playerConfig[GorgeKI.ROPE_TEAMING] + rescueing * playerConfig[GorgeKI.RESCUEING] + gorgeEntering * playerConfig[GorgeKI.GORGE_ENTERING]; return value; } Fig. 3. Getting a move’s score value A code snippet calculating the weighted sum is shown in figure 3. To enable agents acting on plans and intentions the scoring algorithm is extended with a hypothetical GameStates storing the field ownership relations after a hypothetical move. Now agents are able to choose a path at junctions to reach the roped team or a gorge pawn to rescue. Figure 4 displays a game situation that requires the support for future hypotheses. There is a pawn (the yellow one) trapped in the gorge. It is the other player’s (the red one’s) turn. After using the dice this other (red) agent must decide about a moving range of three fields. Being set with a positive weight for rescueing other pawns, the red player picks the upper path at the junction, so it is possible to rescue the yellow pawn in the future as can be seen at the upper right. At the lower right the agent is configured with a negative weight for rescueing other pawns. To avoid possibly rescueing the yellow pawn from the gorge he evades this path. The Good, the Bad and the Ugly 131 Fig. 4. Preparing the Rescue . . . or Deserting the Helpless 4 Interactive Storytelling by Character From Aristotle to the present, there have been developed concepts and ideas of dramaturgy both in the performing arts and in the visual arts galore ([8,9], e.g.). Many of them might be carried over to digital games. However, digital games bring with them the enormous intensity of interaction which may potentially result in an unprecedented manifold of varying experience. Dramaturgy of interactive storytelling deals with the anticipation, the planning, the implementation, and the control of this manifold. The present paper advocates the principal perspective that, in general, the character of protagonists is decisive to real life communication and cooperation and to the unfolding of stories in theater, in motion picture, and in game playing. The authors’–so to speak–dramaturgical technology proposed is to design and implement characters of NPCs such that the peculiarities of the NPCs have a substantial impact on the unfolding of stories when humans play some game. If this works truly successfully, tuning the parameters of NPC characters may become an approach to modification and control of the emergence of stories. This is a very particular, but hopefully original approach which, doubtlessly, has to be seen within the overall endeavor of creative ideas to advance interactive storytelling. The present rather particular perspective leads to the explication of a few rather particular phenomena and related questions such as, e.g., the following. – How much does a large variety of NPC characters contribute to the potential range of emerging stories? – How important are conflicting characters to the attractivity of the story? 132 S. Gaudl, K.P. Jantke, and C. Woelfert – Given a particular game and a particular set of characters, how to adavance the story space by substituting or adding characters? – Are there prototypical configurations of dramaturgically useful sets of more or less complementary characters such as the good, the bad, and the ugly? Sample implementations as surveyed in section 3 demonstrate a way towards experimental investigations into those problems. Obviously, even partial answers to the questions above depend on the special digital game under consideration. Even more importantly, they depend on the human subjects playing the game. In certain application scenarios, the dependence of the effects (and affects) under consideration on peculiarities of the audience are of a particular interest. – Are certain phenomena gender-sensitive? Are–seen statistically–boys more dependent on or tolerant to aggressive characters than girls? – How important is the average age of the audience? – What about the impact of social status? In many of the conventional role playing games such as, e.g., Baldur’s Gate or Morrowind, you meet NPCs in large numbers and either talk to them, trade with them, or fight them. The authors advocate the approach to equip NPCs with character such that different human players in different conditions deal with them differently such that the evolving story may turn into different directions. 5 Summary and Conclusions The gist of the present paper may be circumscribed by the following surely oversimplified illustration. Assume a digital game in which you have designed and implemented some NPC such that there are those players who find the NPC interesting and somehow nice to talk to, whereas other players avoid getting in contact with him. Assume furthermore that cooperating or not cooperating with this particular NPC drives the game story into different directions. In this way you got what we call an (extremly simple) instance of interactive storytelling by character. In addition to other ideas, concepts and models of interactive storytelling, the authors’ approach offers a new opportunity to drive the unfolding of stories intensionally. The approach is modular, because it refers to the characteristics of individual NPCs. It supports modification and experimentation, because you may play–and, perhaps, have fun–with changing profiles of an individual NPC.