DATAPLAY: Mapping Game Mechanics to Traditional Data Visualization


Macklin Colleen Wargaski Julia Edwards Michael Li Kan Yang
2009 DiGRA '09 - Proceedings of the 2009 DiGRA International Conference: Breaking New Ground: Innovation in Games, Play, Practice and Theory

In William Playfair's 1786 book, The Commercial and Political Atlas, he states that information, “imperfectly acquired, is generally as imperfectly retained.” [6] Playfair is commenting on the failure of tables to represent comparative data in way that was useful to the reader. Since Playfair, many different forms of media have arisen beyond ink and paper. Yet printed charts (or their digital representations) remain, by far, the most commonly used tools of data visualization. Their evolution over many centuries has allowed them to achieve a degree of sophistication that time-based and interactive representations have yet to achieve. Is the supremacy of printed (or print-like) data visualization to remain unchallenged? Would it even need to be? The authors contend that new approaches may be possible, and even necessary, but would require tapping into a different way of learning that was not strictly about managing the short term visual and auditory memory of the readers [3]. This learning would involve less the experience of reading and more that of direct experience through play and games. Jesper Juul contends that all games are learning systems [2]. That is, to play a game and become good at it, the player must learn the necessary skills and strategies to overcome their opposition. If the goal of data visualization is educational, it may be possible to use specific types of games as ways of representing specific types of data. It may be possible for a player to learn the system of the game and the system of the information together. The authors have built three game prototypes that illustrate the ways in which different forms of data can be represented in the form of digital games. The first prototype, Kimono Colors, is based on data from a cross-referenced table that describes the types of ingredients used to create traditional Japanese dyes in the production of kimonos. The core mechanic [4] of the game has the player “fishing” for colors using one of several dyes the player has collected. By fishing for these colors, players learn the relationships between materials and the manufacture of dye. The second prototype, Mannahatta: The Game, asks players literally to walk around Manhattan and connect the living and non-living elements of a directed graph representing the ecology of the island 400 years ago. Played over an iPhone, users place themselves in the middle of the dataset they are piecing back together. The third prototype, Trees of Trade, uses data from two directed graphs of relationships, ecological and commercial, in a Brazilian Atlantic rain-forest ecosystem. The game involves the players re-establishing the trophic levels of the forest by navigating through the relationships and inserting the missing species on an idealized map. Through play, the user will better understand the elements of a system that is typically illustrated in a static, two dimensional directed graph illustration. Two questions stem from these prototypes: can data create play and can play enlighten data? To answer the first question affirmatively, we need to find evidence that a system created by data has the ability to produce “choice molecules” [4]. That is, in the form of a game, does the structure of data allow the player to make interesting choices about how to proceed as he or she navigates and elaborates the data? If so, then the data in question can create play, which in turn can drive the development of a full-fledged game. As for the second question, if the first answer holds, then the player is, in the course of a game, playing with the data. If the choices made available to a player are established in such a way that the player “levels up” through the information, then the achievement of the games goals will be coincident with the understanding of the data itself. By actively manipulating and using the data to win the game, the player will need to understand the facts and relationships inherent in the data itself, thus producing the desired educational outcome and a greater sensitivity to the systems that data represents.