University of Southern California Dissertations and Theses

Leafcutters: life simulation gameplay designed to evoke engagement with real-world subject matter

(USC Thesis Other)

Leafcutters: life simulation gameplay designed to evoke engagement with real-world subject matter

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content LEAFCUTTERS: LIFE SIMULATION GAMEPLAY DESIGNED TO EVOKE ENGAGEMENT WITH REAL-WORLD SUBJECT MATTER by William B. Graner A Thesis Presented to the FACULTY OF THE USC SCHOOL OF CINEMATIC ARTS UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree MASTER OF FINE ARTS (INTERACTIVE MEDIA) May 2011 Copyright 2011 William B. Graner Acknowledgements I would like to thank the faculty, staff, and students of the Interactive Media Division for their support of Leafcutters. Thank you to Fox Interactive for granting me the 2010 Fox Interactive Thesis Fellowship to fund this project. Thank you to Jeremy Gibson, Tracy Fullerton, Steve Anderson, and Jamie Antonisse, my advisors. Jeremy, thank you for your constant dedication, energy, and challenges toward improvement. Tracy, thank you for teaching me the art of game design, and for your mentorship and support. Steve, thank you for supporting this idea from the start. Jamie, thank you for demonstrating how to form and lead a game team and craft an amazing thesis. Thank you to my fellow IMD students; you have been my creative home for the last three years. May we never lose the spirit of discovery. Thank you to the Leafcutters team: Torin Borrowdale, Sean Bouchard, Kelly Curtis, Jim Lammlein, Vairavan Laxman, Sanghee Oh, Daniel Ponce, Evan Sforza, and Dai Yun. And thank you to my family for their support and encouragement throughout my career: my mother Darlene Graner, my father Dr. John Graner, and my brother Dr. John Graner. ii Table of Contents Acknowledgements ii ........................................................................................................... List of Figures iv ..................................................................................................................... Abstract v .................................................................................................................. Keywords v .............................................................................................................. Introduction 1 ......................................................... Short Description of the Interactive Work 1 ................................................................................................................. Prior Work 2 ................................................................................................................. A-Life 2 ....................................................................................... Life Simulation Games 3 ..................................................................................................... Swarm Games 6 ......................................................................................................... Virtual Pets 7 ................................................................................... Ants in Interactive Media 8 .............................................................................................................. The Concept 10 ................................... Gameplay Designed to Evoke Real-World Engagement 10 .............................................................................................. Intended Audience 11 ................................................................................................ The Interactive Work 11 ................................................................................................... The Experience 11 ........................................................................................... Aesthetic Treatment 15 .............................................................................................. The Game System 16 ............................................................................ Analysis of the Game System 20 ..................................................................................... The AI System in Detail 21 ................................................................................................... An AI Example 24 ........................................................................................ The Game Design Process 26 .................................................................................. Adaptation From Biology 26 ............................................................................................................. Iterations 27 ................................................................................................................. Evaluation 33 ................................................................................................................. Discussion 33 .................................................................................... How to Read Leafcutters 33 ..................................................................................................... Expressive AI 34 ............................................................................. Evocative Knowledge Object 35 ................................................................................................ Adaptation Game 36 .............................................................................................. Educational Game 36 ................................................................. Fiction and Nonfiction in Leafcutters 37 .................................... Play In Leafcutters: By the Book, Off The Beaten Path 38 ........................................................................ Expansion and Further Research 39 ................................................................................................................ Conclusion 39 ............................................................................................................. Bibliography 41 iii List of Figures Figure 1: Beginning of the game 12 Figure 2: Detail of ants carrying food 14 Figure 3: Swarmites at attention. 23 Figure 4: A simple ant AI 24 Figure 5: An ant who is outside and finds food will want to pick up food. 24 Figure 6: An ant who is inside and finds food will also want to pick up food. 25 Figure 7: A negative urge results in desired behavior. 26 Figure 8: Paper prototypes 28 Figure 9: Programming-style interface. 29 Figure 10: Flowchart-style interface. 30 Figure 11: Individual ant urge interface 31 Figure 12: Final interface 32 Figure 13: Final interface detail 32 iv Abstract Leafcutters is a life simulation game about leafcutting ants which is designed to evoke engagement with real world subject matter. In this game, players shape the behaviors of a colony of ants in order to establish complex behaviors such as foraging and fungus farming. The game system in Leafcutters is adapted from existing biological research on ants, with an emphasis on the accurate adaptation of a natural system into a game system. This project draws on previous works in artificial life, life simulation games, swarm games, virtual pets, and virtual ants. Leafcutters is a work of expressive AI, an evocative knowledge object, and an educational game. Keywords Simulation game, evocative knowledge object, educational game, video game v Introduction Leafcutters is a life simulation game about a colony of leafcutting ants. The player shapes the behaviors of these ants by changing their urges in response to various situations, giving rise to complex actions including foraging, nest construction, and fungus farming. Leafcutters allows a new kind of play with the natural system of an ant colony, with the intention of facilitating player engagement with real-world subject matter. Leafcutters draws inspiration from a range of previous work in Interactive Media, particularly: 1) Artificial Life, 2) Simulation Games, 3) Swarm Games, and 4) Virtual Pets. This project uses key contributions of these domains for the creation of a new game experience designed to evoke curiosity and inspire engagement with the natural world. In this regard, Leafcutters can be considered as a work of expressive AI, an evocative knowledge object, and an educational game. Short Description of the Interactive Work Leafcutters is a game in which the player nurtures and guides a colony of leafcutting ants by shaping and balancing the discrete, situational urges of the colony’s castes, such that the decisions and actions of individual actors combine to form an effective system, allowing the colony to thrive. In the process of establishing and shaping situational urges, the player is guided toward reinventing the signature behaviors of leafcutting ants, as described by current biological knowledge of Atta cephalotes, a leafcutting ant species. However, the game also invites exploration of the possibility space, allowing the player to create unrealistic, original behaviors and see their results play out. The look and feel of Leafcutters is realistic, with detailed creatures set in a specific rainforest backdrop. It seeks not only to represent the interesting systemic properties of its subject, but also to accurately represent leafcutting ants at an aesthetic, personal level. 1 Prior Work A-Life Our exploration of the works which influence Leafcutters begins with Artificial Life. Established by Christopher Langton as a subfield of Artificial Intelligence in the 1980s, Artificial Life, or A-Life, explores concepts of life by seeking to create artificial organisms which are not only lifelike, but truly alive. A-Life builds on von Neumann’s and Conway’s previous work with cellular automata. 1 As Langton writes, “A-life complements the analytic approach of traditional biology with a synthetic approach: rather than studying biological phenomena by taking living organisms apart to see how they work, we attempt to put together systems that behave like living organisms.” 2 In the introduction to his book Principals of Biochemistry, A.L. Lehniger writes, “The molecules of which living organisms are composed conform to all the familiar laws of chemistry, but they also interact with each other in accordance with another set of principles, which we shall refer to collectively as the molecular logic of the living state.” 3 Langton’s seminal work of Artificial Life sets out to “explore the possibility of implementing the 'molecular logic of the living state' in an artificial biochemistry, based on interactions between artificial molecules.” 4 Langton traces the historical roots of A-Life from sculpture and automata, the ongoing desire and attempt to create life. Langton’s theoretical basis of A-Life uses the concepts of genotype: the rules which govern the actions of an organism, and phenotype: the collective effect of the genotype, seen in either the singular organism, or a collective of such organisms. Langdon states 2 1 Langton, Christopher G. "Studying artificial life with cellular automata.” Physica D: Nonlinear Phenomena. 22.1-3 (1986): 120-149. Print. 2 Langton, Christopher G. "Artificial Life." SFI Studies in the Sciences of Complexity. (1988): Print. 3 Lehninger, Albert L. Principles of Biochemistry. 1st ed. New York: Worth, 1982. Print. 4 Langton, “Studying artificial life with cellular automata,” 120. that “The most surprising lesson we have learned from simulating complex physical systems on computers is that complex behavior need not have complex roots.” 5 One of Langton’s first A-Life creatures is a “virtual ant” or “vant” which, following a simple set of rules, and influencing other vants via pheromone trail-styled markings, creates a variety of fascinating emergent patterns. One such pattern is a “highway” or “tunnel” of pheromone marking which extends indefinitely after an initial lengthy period of chaos. 6 Langton’s virtual ant is perhaps the earliest example of ants represented in Interactive Media. Importantly, Langton’s ant is decidedly abstract, and its value is measured in the shapes it creates via its emergent properties. This brings us to the limitations of the Artificial Life as a mode of inspiration for Leafcutters and as a lens through which to understand it. A-Life concerns itself with creating truly living artificial organisms, not with delivering the experience of being in the presence of something which is alive. This has wedded the field to a mathematical, abstracted aesthetic and area of exploration. As evidenced by experimental works of Artificial Intelligence such as Tale- Spin, sometimes rich and deep AI systems in fact deliver a less lifelike experience than do simple, cleverly constructed works. 7 A-Life focuses on abstract systems, to the exclusion of aesthetic design. Leafcutters differs from A-Life in that it uses a realist aesthetic design in order to make its subject matter more lifelike. Life Simulation Games Life simulation games area a genre of digital game which draws heavily from Artificial Life. While the first such work, Atari’s Little Computer People, predates the first A-Life 3 5 Langton, Christopher G. Artificial Life. Redwood City, CA: Addison-Wesley, 1989. Print. 6 Langton, Christopher G. "Studying artificial life with cellular automata." Physica D: Nonlinear Phenomena. 22.1-3 (1986): 120-149. Print. 7 Meehan, James R. "Tale-Spin, an interactive program that writes stories." Proceedings of the 5th International Joint Conference on Artificial Intelligence. San Francisco, CA, Morgan Kaufmann Publishers Inc.. 1977. Print. Conference, later entries into the genre drew directly from A-Life. SimLife describes its virtual “orgots” in Langton’s terminology of genotypes and phenotypes, and it is, like many A-life projects, a game about the interesting emergent behaviors of large populations of simple, similar organisms. 8 It should be noted that SimLife clearly grapples with the introduction of A-life and simulation into the realm of games: the game explicitly states in an introductory demo sequence that SimLife “lets you break the barriers between games and simulations, playing and learning— even between machines and living beings.” 9 Maxis marketed its games at the time as “software toys,” highlighting their focus on self-directed play. 10 These games feature interim achievement- style goals, but rely on the player to set his own long-term objectives. The body of simulation games can be categorized in terms of the number of actors being simulated and, correspondingly, whether these actors are represented through individual game objects or through abstracted collections. Simulation games with similar numbers of actors and modes of representation of these actors also tend to share similar control mechanics. SimCity is an example of a game with many non-discrete actors—“sims”—who inhabit a city. Rather than interacting directly with a particular sim, the player interacts with the city by placing buildings and establishing urban zones, as well as adjusting values such as budgetary allocation, city ordinances, etc. 11 Also, as in SimLife, the player can instigate disasters in his city —revealing the game’s focus on free play and experimentation rather than pursuit of a goal, since these fun activities actually set the player back in terms of success within the game system. One step closer to individual simulation are simulation games in which the player manages many discrete autonomous actors. These include Populous, SimTower, Dungeon 4 8 Cliff, Dave, and Stephen Grand. "The Creatures Global Digital Ecosystem." Artificial Life 5. (1999): 77-94. Print. 9 SimLife. Maxis, 1992. 10 Cliff, Dave, and Stephen Grand. "The Creatures Global Digital Ecosystem." Artificial Life 5. (1999): 77-94. Print. 11 SimCity, Maxis, 1989. Keeper, Dwarf Fortress, and SimAnt. 12, 13, 14, 15, 16 In these games, autonomous actors make their way through the simulation, navigating an architectural space. Rather than directly modifying this space, the player directs the actors to modify the space around them. In this way, the player retains exacting architectural control but must leave the realization of this design to the agents in the game. The player also manages traffic, often through the use of waypoints which attract agents to them. In the game SimAnt, this spatial attraction system ties into the subject matter of ants, using their pheromone trails as the attractive game object. In these games, the player generally seeks to move the game actors into place, at which point the autonomous actors carry out beneficial tasks, such as resource gathering and organization. The player acts as a manager, directing, but not controlling, the simulated agents toward progression in the game. However, some of these games allow the player to temporarily take control one of the actors, using it as an avatar for direct interaction and exploration. SimAnt provides this in the player’s control of a yellow ant, which serves as an avatar, allowing the player to directly shape the nest (in fact, digging much more quickly than a non-yellow ant), forage for food, and organize attacks against the competing red ant colony or predators. While this avatar provides the player with an avenue for engagement with the world of the simulation and immersion within the space of the playfield, it comes at the cost of disrupting the simulation by injecting human problem-solving into the system. 5 12 Populous. Bullfrog, 1989. 13 SimTower. Maxis, 1994. 14 Dungeon Keeper. Electronic Arts, 1997. 15 Slaves to Armok: God of Blood Chapter II: Dwarf Fortress. Bay 12 Games, 2006. 16 SimAnt. Maxis, 1991. Finally, there are sim games in which few, complex actors rely on the player’s guidance for their survival, such as Creatures, Black and White, and The Sims. 17, 18, 19 Creatures and Black and White exhibit commonalities in terms of their creature learning model, wherein the player uses rewards and punishments—in this case tickles and slaps, to reinforce or discourage the actors’ behaviors. The Sims, rather than using a learning model, makes use of an urge-based system mapped to the space of the sims’ homes, which was inspired by Will Wright’s work on the pheromone systems of SimAnt. 20 Swarm Games Swarm games are a relatively new genre of video game and feature large groups of similar or identical actors which the player controls either directly or indirectly. It is this amount of direct control, the limited complexity of individual actors, and the presence of these creatures in crowds which differentiate swarm games from simulation games (although there are overlaps between the two). Examples of swarm games include Lemmings, Pikmin, and Swarm, as well as experimental games based on swarm AI algorithms such as boid flocking. 21, 22, 23 Sebastian von Mammen and Christian Jacob of University of Calgary present two game prototypes using boids. 6 17 Creatures. Mindscape, 1996. 18 Black and White. Electronic Arts, 2001. 19 The Sims. Electronic Arts, 2000. 20 "Will Wright explains what The Sims and an ant colony have in common." Joystiq.com. 2010. 21 Mar. 2011 <http:// www.joystiq.com/2010/11/08/will-wright-explains-what-the-sims-and-an-ant-colony-have-in-com/>. 21 Lemmings. Psygnosis, 1991. 22 Pikmin. Nintendo, 2001. 23 Swarm. Hothead Games, 2011. 24 In the first game “Herd the Cows,” the player controls one boid, which other boids in viewing range will follow and flock around; the player must guide as many boids as possible into a goal area in as short an amount of time as possible. Their second game, “Feed the Crows,” replaces this direct control scheme with the less direct controls, including the placement of waypoint tiles and parametric controls which affect the boids’ behaviors. Interestingly, this game also allows the player to make use of automatic, artificial evolution of boid behaviors to aid the player in what would otherwise be a prohibitively difficult game. Virtual Pets Virtual Pets are a genre of interactive object which emerges from A-life, simulation games, and automata, but with the distinctive goal of fostering empathy between the user and the lifelike virtual object. They place the user into a caretaker role as pet owner or parent. Virtual pets draw on automata such as the early Mechanical Turk, to the more recent Senster, in that they do not seek to function as actual life forms, but rather to present lifelike qualities. 25, 26, 27 Since the first use of the term Virtual Pet with PFMagic’s Dogz and Catz, the genre has had many 7 24 von Mammen, Sebastian, and Christian, Jacob. "Swarming for Games: Immersion in Complex Systems." Applications of Evolutionary Computing. Ed. Mario Giacobini, Anthony Brabazon, Stefano Cagnoni, Gianni Di Caro, Anikó Ekárt, Anna Esparcia-Alcázar, Muddassar Farooq, Andreas Fink, and Penousal Machado. Heidelberg: Springer Berlin, 2009. 293-302. Print. 25 Benthall, Jonathan. Science and Technology in Art Today. London: Thames and Hudson, 1972. Print. 26 Schaffer, Simon. "Enlightened Automata." The Sciences in Enlightened Europe. Ed. Clark et al. Chicago and London: University of Chicago Press, 1999. 126-165. Print. 27 Kac, Eduardo. "Origin and Development of Robotic Art." Art Journal, Digital Reflections: The Dialogue of Art and Technology, Special issue on Electronic Art. 56.3 (1997): 60-67. Print. entries such as Tomagotchi, Furby, Nintendogs, Pleo, Webkinz, and Pluff, as well as the previously mentioned simulation games The Sims and Creatures. 28, 29, 30, 31, 32, 33, 34, 35 Ants in Interactive Media Ants have had a long history as subject matter in works of Interactive Media. These tend to fall into three categories: children’s educational games, bug-squashing games, and ant colony simulations. Children’s educational games about ants feature a wide variety of mechanics and subject matter, but generally share a cartoonish art style and a tendency to anthropomorphize ants. Examples include Ant War, Pest Detective, and Archibald’ s Adventure. 36, 37 , 38 Bug-squashing games are interesting in that they are very simple and numerous, and anti- empathetical. They include Ant Splat, Ant Squash, Ant Smasher, and even Mario Paint, among 8 28 Cliff, Dave, and Stephen Grand. "The Creatures Global Digital Ecosystem." Artificial Life 5. (1999): 77-94. Print. 29 Pets (Dogz and Catz). PF Magic, 1995. 30 Tamagotchi. Bandai, 1996. 31 Furby. Tiger Electronics, 1998. 32 Nintendogs. Nintendo, 2005. 33 Pleo. Ugobe, 2006. 34 Webkinz. Ganz, 2005. 35 Hughes, Diana. "Pluff: Creating intersections between tactile interface devices and fabric based electronics". MFA thesis. University of Southern California, 2009. 36 Ant War. Anarchy Enterprises, 2003. 37 Pest Detective. 21 Mar. 2011 <http://www.pestworldforkids.org/pest_detective/index.html>. 38 Archibald’s Adventure. 21 Mar. 2011 <http://www.pestworldforkids.org/archibald/index.html>. many others too numerous to list. 39, 40 , 41 , 42 These are essentially performative reaction and aiming games with a negative, if cartoonish, fiction and are perhaps the simplest type of game in which killing is the primary activity. Ant colony simulations include Myrmedrome, 3D Ant Farm Simulation, and SimAnt. 43, 44, 45 In the three named simulations, the ant species being simulated is unspecified. Each simulation also features a colony of hostile red ants, which function identically to the black ants, but compete and fight against them. The ant species are distinctly abstracted, perhaps in order to signal that the simulations are meant to apply to ants in general. Likewise, the space of the simulations is abstracted, formed from generic brown dirt. Although SimAnt is more specific in this regard than the other examples, in that each region is placed within the larger area of the yard, the similarity of each local playfield suggests an abstracted space. Another interesting similarity among the simulations is their winking reference to human interaction with ants. Myrmedrome and 3D Ant Farm Simulation do this through the food which the ants seek out in their environment, one representing food as discarded brownies, and the other denoting the food placement tool with a spaghetti icon. These choices—the non-specificity of location, the abstraction of ant species, and the self-referential choice of game elements—highlight the abstracted style of these games. The abstraction of ants in Interactive Media goes as far back as Langton’s early work in A-life, with his virtual ant. In terms of genome, a cellular automaton only resembles an ant in that it places signals onto its environment which it, and other virtual ants, can later read. At the 9 39 Ant Splat. 12 Mar. 2011 <http://arcadevoid.com/games/ant-splat>. 40 Ant Squash. Best Cool & Fun Games, 2010. 41 Ant Smasher. 12 Mar. 2011 <http://www.schvarts.com/games/ant-smasher/>. 42 Mario Paint. Nintendo, 1992. 43 3D Ant Farm Simulation. 12 Mar. 2011 <http://www.forgefx.com/casestudies/prenticehall/ph/ants/ants.htm>. 44 Myrmedrome. 12 Mar. 2011 <http://www.not-equal.eu/myrmedrome/>. 45 SimAnt. Maxis, 1991. phenome level, however, Langton’s ant shares a core characteristic with biological ants: its simple decisions and actions lead to complex emergent behaviors. Langton’s ant set a precedent for abstract representation of ants in Interactive Media. Leafcutters attempts to bring representational authenticity to its subject matter, both at the level of the individual insect and the colony level of complex, emergent behavior. The Concept Gameplay Designed to Evoke Real-World Engagement Leafcutters was conceived as an interactive game experience that evokes engagement with subject matter drawn from the natural world. We use the term engagement to refer to curiosity, interest, and knowledge in the game’s area of content. The subject of ants was chosen because cursory research revealed that the subject matter was full of interesting features at a variety of depths, so that an ongoing play experience could continue to reveal facets of interest to an engaged player. The world of ants is a mysterious, alien universe that is also quite close to home for anyone on the planet. The choice to create a digital game was not made lightly. For example, an interaction with real ants could more directly evoke engagement with them as a subject. However, the video game medium allowed for interactions and forms of play that would otherwise be impossible. The traditional monitor, mouse, and keyboard were chosen to allow for the greatest accessibility of the game via online distribution. Making the game available online also allows for discussion between users, which could serve as another tool for player engagement. The core mechanic of Leafcutters is the player’s encouragement and discouragement of urges in the minds of the ants in response to specific situational cues by which the ants understand their surroundings. The ants ultimately make their own decisions in response to these urges, and the player never has direct control over individual ants. 10 The ants’ free will helps to make them feel alive to the player. This perception is designed to encourage the player to invest emotionally in the ants, therefore increasing engagement with the subject matter. It is a core design assumption of Leafcutters that the player’s experience of the game as a window into a living world (rather than a technical simulation) is key to the game’s efficacy as an evocative experience for engagement with real- world subject matter. The desire to maintain the evocative properties of the game placed certain high-level restrictions on the design of Leafcutters. For one, the game had to adhere to current biological knowledge of leafcutting ants, so that it would remain consistent with a player’s further explorations in the subject matter. Therefore, fun of gameplay and the direct satisfaction of players became secondary priorities during the initial period of design. However, since the game would also fail as an evocative experience if users were not engaged with the game itself, user- centric design techniques were emphasized after the initial basic design was established. Intended Audience As an installation piece, Leafcutters is designed to appeal to a wide audience. Previous experience with video games will be helpful to new players, but the game accommodates anyone with basic computer familiarity. As a downloadable game, Leafcutters invites further exploration by hobbyist players. Leafcutters is designed for players of middle school age and older. The Interactive Work The Experience In its gallery installation incarnation, Leafcutters is played on an HD television screen, set in front of a couch that is large enough for three people. On a small, movable stand in front of the couch are a wireless mouse and keyboard. The player approaches the game, perhaps with a friend or two to observe and share the experience. Already visible on the screen is the cross 11 section view of a small, simple colony of ants. The nest has one room with one tunnel leading upward to the surface. In the room are a large queen ant, a handful of large worker ants, and a blob of fungus which resembles fuzzy gray coral. The worker ants walk up and down the tunnel, pacing in and out of the nest. Music plays which is calm and pastoral, yet alien. After a few seconds, a small message appears on the screen: “Try clicking on an ant.” Figure 1: Beginning of the game This message is the beginning of a tutorial which leads the player through the basics of the game. This includes the navigation arrows in the upper corner of the screen, which allow the player to move the camera up, down, left, and right, as well as zoom in and out –– alternately revealing the details of an individual ant, and bringing into view the growing swarm of the colony. Moving the camera upward, the player leaves the underground and emerges in a lush rainforest surrounding the small ant mound. Through the tutorial’s guidance, the player comes to understand the series of icons and symbols through which he can interact with the ants. By clicking on symbols which represent 12 certain conditions, such as “Outside,” “Near food,” or “Inside nest,” the player specifies a set of conditions, and all of the ants matching that condition become visibly highlighted in yellow. By clicking on icons related to urges, such as “Pick up food,” “Go inside,” or “Drop what you’re holding,” the player encourages or discourages those urges, among the given set of conditions. Another message appears to the player: “The larvae are hungry for food.” This time, the message contains small icons which show the larvae to be the small white squiggly creatures inside the nest, and the food to be the green globules outside. Guided by the tutorial, the player selects all of the ants inside the nest, and clicks on the “Explore Outside” icon. A green bar lights up above the icon, and the ants start to walk out of the nest and wander around outside. As they leave the nest, they glow more faintly, since the selected condition icons no longer exactly match their situation. The player spots one of the ants passing by a piece of food, and he clicks on the ant. The ant glows for a moment, then fades. The player sees that as his ants pass by pieces of food, they each glow for a moment. He clicks the icon to encourage the urge to “Pick up food.” Over the next few seconds, he sees each of his ants pick up a piece of food, and continue wandering outside. 13 Figure 2: Detail of ants carrying food He uses the same process of making suggestions to get the ants to go back inside, then drop off the food they’re holding. The larvae immediately seize the food, consume it, and hatch into ants. Because these new ants are now inside the nest, and therefore match the conditions established earlier, they get the urge to explore outside, crawling outside in search of more food to gather, as the queen births more larvae. The player follows a series of prompts toward intermediate goals which guide him toward the leaves, and the growth of a fungus garden. As he learns how the various elements of the ecosystem interact, he also witnesses the way his ants work: they accomplishing tasks en masse, with individual inefficiency but strength in numbers. As his nest fills with fungus, ants, and larvae, he instructs the smaller ants to dig new tunnels whenever they are in a room that is full, and the nest begins to expand deeper into the ground. 14 Suddenly, the music changes, taking on an ominous tone, and the player’s eye is drawn to new creatures in the area: a group of large hostile ants have wandered close to the anthill. When one of the player’s ants encounters the hostile ants, they battle, and the player’s worker ant is killed. In reaction, the player selects all of his outside ants and gives them a strong urge to go inside. In fact, the ants nearest to the hostile ants associate this urge with the strangers’ smell, and will remember this lesson (a key point, since such accidental learning by reinforcement will continue to influence the player’s interaction with the ants). After the hostile ants have left, the player tells his underground ants to go outside again. Eventually, the colony is booming, like a swarming factory. The music has become busy and celebratory, in keeping with the colony’s growth. The player has achieve all of the goals with which he has been presented, and a message appears to congratulate him. The title, “Leafcutters,” fades into the foreground, with his bustling colony as its backdrop. The player may continue to play, and when he chooses to leave, the screen fades slowly to black, eventually resetting for the next player. Aesthetic Treatment The look and feel of Leafcutters is rooted in realism and specificity. The ants are modeled and animated as accurately as possible after Atta cephalotes, a species of leafcutting ant, with bright orange coloration, lanky legs, and enormous jaws for cutting leaves. They are not the ants an American would find in her back yard. Rather, they are alien, but also related to ubiquitous ants. The environment art is similarly alien but familiar, a rainforest scene with deep green tree roots and thick grassy underbrush. The game is built in the Unity 3D engine, using 3D models for the creatures and environment. The 3D perspective view creates a sense of physical depth, particularly highlighting the contrast of the flat space of the underground nest with the deep space of the rainforest exterior. 15 The sound design of the game is gritty and realistic, like the audio from a nature documentary film’s closeup scenes of insects. In contrast, the music is cheerful and eclectic, pairing exotic acoustic and percussive instruments with electronic embellishments. The audio design is intended to give Leafcutters the feeling of a window into the real world, but with some reference to its video game heritage. The overall aesthetic is designed as a vignette into a world which is both familiar and alien, to present the real world as a space for exploration. The Game System We will now examine the game system of Leafcutters, based on Tracy Fullerton’s list of formal elements of games. 46 Leafcutters is a single-player, player-vs-system game, the objective of which is to grow a colony population of 50 ants. In this description, the syntax “Trigger → Urge” represents a trigger and urge pairing, such that when the trigger is true, the ant will experience the listed urge. The term “cooldown,” applied to urges and behaviors, refers to the amount of time which must pass between executions of the same behavior by a given ant. In Leafcutters, ants are delineated into four castes: forager, midden worker, nurse, and queen. While the names and physical appearances of these castes are based on scientific knowledge of leafcutting ants, the three worker castes (forager, midden worker, and nurse) are functionally equivalent in the game. Because each caste can be given its own set of triggers and urges, the separation of castes allows the player to create specialized groups of workers. The appearances and names of the castes, while not systemically significant, hint to the player how such groups of workers might be specialized, and they are meant to prompt the user’s curiosity regarding the roles of these castes in real ant colonies. 16 46 Fullerton, Tracy. Game Design Workshop. 2nd. Elsevier, Inc., 2008. Print. Resources: food, larvae, worker ants (foragers, midden workers, and nurses), fungus, leaf pulp, the queen ant Anti-resources: leaf waste, dead ants Dangers: old age (for worker ants), hostile ants Procedures: Encourage an urge for a caste/condition combination; discourage an urge for a caste/condition combination. See the list of ant urges below. Rules: Ants’ Actions: • Each frame, each ant which is not undertaking a task, or whose set of conditional triggers has changed since last frame, determines which of its triggers are true (see list of triggers below). • It then determines its current urges as follows: for each of the ant’s urges, if one or more of the triggers related to that urge are true, that urge is added to the active urges for this frame, with a calculated weight: the base weight of the urge is multiplied by the squared proportion of that urge’s triggers which are currently true. • Finally, the ant uses a weighted random to determine which urge to pursue as its next action. Ant death: • Worker ants die of old age after 3 minutes. • Dead ants disappear after 1 minute. Ant Triggers (Conditions): • Inside Nest • Outside • This room is full • Near food/leaf/pulp/leaf waste/larva/fungus/hostile ant/dead ant 17 • Holding food/leaf/leaf waste/pulp/larva/fungus/dead ant/nothing Ant Urges: • Pick up food/leaf/leaf waste/pulp/larva/fungus/dead ant: If you’re already holding something, drop it, then pick up the specified object if it is nearby. • Drop what you’re holding. • Explore outside: The ant explores outside near the nest. The longer this action goes uninterrupted, the farther from the nest the ant will explore. • Explore inside: The ant walks around inside the nest, randomly choosing destination rooms and walking to them. • Dig a new room: If inside the nest, the ant digs a tunnel from its current room to a new room (10 second cooldown). • Birth larva: This ant creates a larva (Queen only, 5 second cooldown). • Chew leaf into pulp: If not holding anything, this ant picks up a nearby leaf. If holding a leaf, this ant chews the leaf. If the leaf has been chewed for 2 seconds total, it becomes leaf pulp. • Tend fungus: Turns nearby fungus healthy (see fungus rules below; 5 second cooldown). • Attack hostile ant (see hostile ant rules below). Castes: • There are three castes of worker ant: foragers, midden workers, and nurses. These castes do not have inherent differences, but can each have different configurations of triggers and urges, allowing the player to specialize them as desired. • At the start of the game, foragers are the only available cast of worker ant. 18 • When the colony’s population has grown to 3 ants greater than the starting number, the nurse caste is unlocked. • When the colony’s population has grown to 10 ants greater than the starting number, the midden worker caste is unlocked. Larvae: • If a larva is near food, it will eat the food and hatch into a worker ant of a randomly selected caste, from the unlocked worker castes: forager, midden worker, or nurse. Fungus: • Fungus starts healthy (gray) but turns unhealthy (yellow) over the course of 2 minutes. • If a healthy fungus is near a piece of leaf pulp in a room which isn’t full, the fungus will eat the leaf pulp and produce another fungus and a piece of leaf waste. • A fungus which is not completely unhealthy will produce a piece of food every 20 seconds. Hostile Ants: • Hostile ants enter the scene in group of 1–3, approximately once every 5 minutes. • Hostile ants have the following urges: Outside → Explore Outside (weight 1) Find Food → Pick up Food (weight 1) Carrying Food → Leave Scene (weight 3) Find Player’s Ant → Attack Player’s Ant (weight 5) • When one ant attacks another, both ants will begin to fight. If another ant attacks an ant which is already fighting, it will join the fight. After 19 10 seconds of fighting, whichever side has the most ants involved in the fight will win, and the other ants will die. Room Capacity: • Each room in the nest has a maximum capacity of 10. • Each game object has mass: Ants, larvae, food, and leaf waste have mass 1; fungus has mass 3. • If the game objects in a room have a total mass greater than the room’s capacity of 10, the room is full, and it glows blue. • In a full room, new larvae cannot be born, fungus cannot produce food, and fungus cannot produce more fungus. Starting Conditions: At the start of play, the nest consists of one room with a tunnel connecting to the nest entrance. In the nest room is the queen ant, one fungus, and three foragers. Outside, there are 14 pieces of food throughout the map. There are also 10 leaves in the grass near the nest. At the top of the tree to the left of the nest, there is an infinite number of leaves (these replenish as collected). The ants have the following urges at game start: Queen: Inside → Birth Larva (weight 1) Forager: Inside, Outside → Explore Outside (weight 1) Analysis of the Game System To succeed in Leafcutters, players must finesse their ants’ urges into an economy and then balance that economy. The greatest opposing factor to this success is worker ant death, from old age or hostile ants. A limited life span encourages ants to explore outside as much as possible 20 for maximum returns; however, hostile ants outside mean that exploration is dangerous and wasteful. Since the loss of a young ant is a greater cost than the loss of an older ant, it is more efficient for older ants to forage. The player must also manage space in the nest. Because each room in the nest has a maximum capacity beyond which new food, fungus, and larvae will not be produced in the room, the player must dig a nest of appropriate size and organization to house his colony. This process is important to the player’s further understanding of the decentralized nature of ant architecture; since the player cannot designate where new rooms will be built, he must express his architectural instructions as situational urges, in keeping with the process by which real ant nests are shaped. Because the food in the game environment is limited, in order to achieve a population of 50 ants, the player must learn to grow and tend a fungus garden. This is a complex activity which includes several smaller tasks: gathering leaves, digging new rooms in the nest, removing leaf waste from the nest, distributing the fungus throughout the nest, and gathering the food from the fungus throughout the nest. Completing these subtasks requires that the player demonstrate an understanding of the decentralized logic that drives an ant colony. The AI System in Detail Each ant has a caste (forager, midden worker, nurse, or queen), and each caste has a behavioral model that is represented by triggers and urges. A trigger is a boolean statement about the ant’s situation which is updated each frame. Triggers include “inside nest,” “outside,” “carrying food,” “smells leaf” and so on. The configuration of these triggers each frame represents the ant’s entire “understanding” of its environment. Urges represent the ant’s desire to perform a certain action. Each urge is tied to one or more triggers, and each urge has a strength which is determined by the player. When all of the triggers tied to an urge are true, that urge occurs at full strength. When only some of the triggers 21 are true, the urge will occur at partial strength. The strength of an urge drops exponentially with the number of false triggers to which it is connected. Because of this, an ant will have a stronger urge to do something when the situation is exactly correct than the strength of its urge to do something only partially fitting the situation. Just as ants can have positive urges, they can also have negative urges, which dissuade them from taking certain actions in response to certain situations. Since urges occur at partial strength when only some of their triggers are true, negative urges are an important feature because they allow the player to teach ants not to do things in certain situations. For example, a foraging ant ought to pick up food when it isn’t carrying anything and it smells food outside, but it should be dissuaded from doing this when inside. As described, the ants’ triggers affect the strength of their urges at each moment. However, it is ultimately up to each ant to decide what to do. When an ant finishes its current action or is confronted with a change in situation (represented by a change in triggers), it chooses a new action based on a weighted random choice from among its current urges. The ant will then pursue this new action until the action is completed or the ant encounters a different set of triggers. The player interacts with the ants by making positive or negative behavioral suggestions in response to an ant’s current situation. When the player makes such a suggestion, it establishes or alters the weight of the urge, tied to the triggers which are currently true for the selected ant. This change alters the urges for the entire caste of ants to which the selected ant belongs, and therefore the results of the player’s suggestions are immediately reflected throughout all the ants of that caste. This model is similar to that demonstrated in a video of an early version of Swarm, by Hothead Games. 47 In order to clarify the mass teaching process of the swarm, the designer has 22 47 "Swarm – Hothead Games." YouTube. Web. 21 Mar 2011. <http://www.youtube.com/watch?v=F_xRPB5Kijk>. included a distinctive visual cue: when one member of the swarm is selected for manipulation, the other Swarmites turn toward it attentively, watching to see what it will do (at the player’s behest). Behavioral changes spread through the swarm over the course of several seconds, simulating an organic process of group learning. This is an elegant design solution, but Leafcutters does not allow the same visual cue; Swarm benefits from the humanoid form of its Swarmites, as well as the relatively flat, open space which they inhabit, both of which make this group watching situation easy to relate. In Leafcutters, it is much more difficult to signal “stopping and watching,” since the ants are non-humanoid and their lines of sight are usually blocked. Instead, the ants who will be affected by changes are highlighted yellow, with a brightness depending on how well their current active triggers match those selected by the player, and therefore how strongly the player’s suggestions will affect their behavior. Figure 3: Swarmites at attention 23 An AI Example This figure shows a simple example of the triggers and urges that make up the AI for a caste of ants. When a trigger on the left is true, the ant will experience all of the urges connected to that trigger. Outside Inside Nest Find Food Pick up Food: 3 Find Spider Run Away: 5 Triggers Urges (and Weights) Figure 4: A simple ant AI The figure below shows what will happen when the ant smells food while outside. Both triggers are activated, and the urge to “Pick up Food” occurs at its full weight of 3. Outside Inside Nest Find Food Pick up Food: 3 Find Spider Run Away: 5 Triggers Urges (and Weights) Resulting Urge Pick up Food: 3 Figure 5: An ant who is outside and finds food will want to pick up food. In the following figure, only one of the triggers associated with the “Pick up Food” urge is true, and therefore the urge occurs, but at a lesser weight. Since this partial weighting reduces 24 exponentially rather than linearly, the resulting weight is approximately 1, less than half the full weight of the urge. Note that this might cause the ants to pick up food inside the nest, which could be counterproductive if, for example, the player intended the ants to gather into the nest from outside. Outside Inside Nest Find Food Pick up Food: 3 Find Spider Run Away: 5 Triggers Urges (and Weights) Resulting Urge Pick up Food: 1 Figure 6: An ant who is inside and finds food will also want to pick up food. A negative urge can alleviate this problem, as shown in the figure below. If the player discourages the ant from picking up food when that ant is inside the nest and smells food, the ant will gain a negative urge in that situation. As shown below, the negative urge will prevent the ant from picking up food inside the nest by reducing the urge’s weight to a negative number. However, it will not prevent the ant from experiencing the urge to pick up food outside (although it will slightly dampen it). 25 Outside Inside Nest Find Food Pick up Food: 3 Find Spider Run Away: 5 Triggers Urges (and Weights) Resulting Urge Pick up Food: 1 Pick up Food: -3 Pick up Food: -3 Pick up Food: -2 Resulting Urge Figure 7: A negative urge results in desired behavior. This system of triggers and overlapping positive and negative urges allows the player to quickly establish a robust system of urges based on many triggers at once. Since clicking on an ant will select all of that ant’s active triggers, the player is able to create basic patterns of AI in relatively few clicks, and refine them later using the more precise icon interface. The Game Design Process Adaptation From Biology Leafcutters was designed as an adaptation of a natural system into a game system. The game system is as true to biology as possible, with necessary simplifications and omissions. By maintaining a depth of accurate content, the game is designed to inspire the player to engage with and explore the subject matter. Game design began with research into ant biology. The system of castes, urges, and triggers is inspired by Wilson and Hölldobler’s descriptions of ant castes, tasks, and roles, in which they go so far as to divide worker behavior into 29 discrete tasks, from which emerge the entire behavior of the colony. 48 Wilson and Hölldobler write that the emergent behavior of an ant 26 48 Hölldobler, Bert, and Wilson, Edward O. The Ants. Belknap Press, 1990. 299. Print. colony is comprised of individuals’ “elementary decisions based on local stimuli that contain relatively small amounts of information.” 49 They describe examples of such ant logic: ...continue hunting for a certain foodstuff if the present foraging load is accepted by nestmates; follow a trail if sufficient pheromone is present; feed the queen more if final-instar larvae are present; and attend the larvae and other immature stages if regular nurse workers are absent. Each of these rules is easily handled by the individual worker, even when we allow for brains as small as a tenth of a cubic millimeter. Each action is also performed in a probabilistic manner with limited precision. Yet when the actions are put together in the form of dense hierarchies involving large numbers of workers, the whole pattern that emerges is strikingly different and more complicated in form, as well as more precise in execution. 50 The system of leaf cutting and fungus gardening, including leaf pulp, leaf waste, and food production, as well as the “sickening” of the fungus, is based on the description of fungus gardening and escovopsis microfungus infestation as described by Wirth, Herz, Ryel, Reyschlag, and Hölldobler. 51 Leafcutting ant colonies must balance a variety of factors to maintain efficiency, including foraging outside versus the danger of losing ants to predators; maintaining the fungus garden versus spending time foraging; and opening nest entrances to vent the CO2 released by the fungus, versus the danger of flooding during rainstorms. 52 Some of these balancing tasks were included in the game, and others, such as considerations of CO2, would make exciting material for future development. Iterations Leafcutters underwent many changes throughout its design, with the primary challenge being establishment of the player’s mode of interaction with the ants. While the game has always 27 49 Hölldobler and Wilson 358 50 Hölldobler and Wilson 358 51 Wirth, R., H. Herz, R.J. Ryel, W. Beyschlag, and B. Hölldobler. Herbivory of Leaf-Cutting Ants: A Case Study on Atta colombica in the Tropical Rainforest of Panama. Berlin, Heidelberg: Springer-Verlag, 2003. 18. Print. 52 Tsutsui, Neil. Personal Interview by William Graner. 1 Nov. 2010. focused on modifying the conditions and resulting actions of the ants, the expressive AI system and user interface underwent many strikingly different iterations. Very early in the design process, paper prototyping suggested that simple rules made from pairings of conditions and actions could lead to emergent behaviors which would allow the player to solve problems by indirectly controlling ants in the colony. However, it soon became apparent that such emergence at a sizable scale would rely on digital prototyping. Figure 8: Paper prototypes In the initial digital prototype of the game, the player constructed behavior sentences out of conditions and actions, in a process similar to computer programming. One problem that emerged from this prototype was the amount of “debugging” which the player had to do in order to get his intentions across. At a deeper level, however, the prototype was unsatisfying because users reported that the ants felt more like robots than like living creatures, which broke the design goal of creating an experience which could evoke engagement with the real world. 28 Figure 9: Programming-style interface. The next prototype tested a flowchart style user interface and AI system, in which the ants’ current behaviors were mapped onto nodes on a graph, with arrows between the nodes representing conditions which would cause the ant to switch actions. While this iteration solved the debugging problem by providing a clearer view of the ants’ internal logic, it did not make the ants seem any more lifelike. To address this in the next iteration, the AI system was partially obscured, and weighted randomness was introduced to simulate the ants’ autonomous decision making. 29 Figure 10: Flowchart-style interface. This led to the first iteration of the AI system based on urges and triggers. The user interface was revised to require the player to select and interact with one ant at a time, with the results of these interactions spreading to all of the ants of the same caste. While informal user testing suggested that this design gave players a stronger sense that the ants were alive, it also introduced some confusion since players were, in fact, affecting all similar types of ants when apparently interacting with one. Also, while this user interface worked with small numbers of ants, it became prohibitively difficult to use with large swarms. 30 Figure 11: Individual ant urge interface Finally, the current user interface was developed, which maintains some of the intimacy of the urge interface, as well as the organic feel of the urge AI, while allowing the player to manage large groups of ants via icon-based selection. The player interacts with the ants using the icon panel at the bottom of the screen. The icons on the far left represent the castes of ants. To the right of these are icons representing all of the triggers that the player’s ants have experienced so far. When the player selects a caste icon and one or more trigger icons, a set of urge icons appears to the right of these, showing the corresponding urge strengths for the given situation. The player clicks the arrows above or below an urge to set change that urge’s strength for the given situation. Ants whose caste and active triggers match those selected on the icon panel will glow yellow. By clicking an ant, the player can quickly select that ant’s caste and active triggers. 31 Figure 12: Final interface Figure 13: Final interface detail 32 Evaluation Evaluation for Leafcutters occurred in the form of frequent informal user testing during the design process. In addition to testing for usability and user satisfaction, it was key to evaluate the game as an experience which would lead to engagement with the real world. After playing the game, subjects were asked the following questions: 1. Describe the ants in the game. Did they seem like robots? Did they seem lifelike? 2. What, if anything, about the game’s subject matter stood out to you? 3. To the best of your ability, describe how a leafcutting ant colony functions. 4. Would you recommend this game to a friend? Question 1 was designed to gauge the player’s emotional experience with the virtual ants. Question 2 was designed to determine the player’s general interest and curiosity in the subject matter and to determine what portions of the game’s science-based content were most memorable. Question 3 revealed the player’s mental model of the game system; also, by framing the question in terms of real ant colonies rather than discussion of the game itself, it provided some measure of the player’s willingness to apply knowledge of the game system to the real-world subject matter. Finally, Question 4 was used to gauge the player’s enjoyment and perceived worth of the game. This interview methodology was used during game design to guide the design process. The pace of development required that testing be quick and informal; while useful for the design process, these results could not be used to reliably evaluate the effectiveness of the game. This presents the future opportunity for a more rigorous study to evaluate the effectiveness of Leafcutters as a game which increases engagement with real-world subject matter. Discussion How to Read Leafcutters As a work which draws from a variety of lineages of Interactive Media, Leafcutters can be interpreted through several lenses. As discussed earlier, Leafcutters is a simulation game. It is 33 also a work of expressive AI, an evocative knowledge object, an adaptation game, and an educational game. Expressive AI Michael Mateas presents the concept of expressive AI: Expressive AI is a new interdiscipline of AI-based cultural production combining art practice and AI research practice. Expressive AI changes the focus from an AI system as a thing in itself (presumably demonstrating some essential feature of intelligence) to the communication between author and audience. 53 This point of view describes the design priorities and process of Leafcutters. The game is not a work of AI research, yet the artificial intelligence system which drives the individual ants and allows for the emergence of large-scale collective behavior is key to the user experience. This influenced the methodology of the game’s creation. While Fullerton’s design methodology would suggest user playtesting via paper prototyping before technical implementation, since the artificial intelligence in Leafcutters is indeed “part of the concept of the piece, part of the larger interpretive context of people theorizing about the piece,” early implementation of this artificial intelligence was key to the design process. This allowed for iteration in order to tune the artificial intelligence and achieve the game’s goal of lifelike creatures. 54, 55 It is vital to Leafcutters that the ants seem alive and autonomous, and this effect could only be elicited via a relatively intensive, early technical implementation. 34 53 Mateas, M. "Expressive AI: A hybrid art and science practice." Leonardo, Journal of the International Society for Arts, Sciences, and Technology. 34.2 (2001): 147-153. Print. 54 Fullerton, 179-212. 55 Mateas, 151. Evocative Knowledge Object Evocative knowledge objects, introduced by Rich Gold in The Plenitude, have been described by John Seely Brown as “metaphors, sayings, or experiences which then help the other person rapidly construct their own understanding,” and by Steve Anderson as “tools, systems and architectures that allow us to think differently about the world,” or more simply, “objects we think with.” 56, 57 , 58 Leafcutters was designed as an evocative knowledge object, in that it seeks to evoke engagement with a natural system and with the outside world in general. The game is an object in that it is a digital artifact, a “toy” for the user to play with, echoing Maxis’ description of their Sim games. 59 This philosophy impacted many design decisions in the creation of Leafcutters. First, the game guides the player only very loosely, with achievement-style intermediary objectives which may be pursued or ignored. Second, the game contains very little didactic information, instead relying on the player’s interaction with the system for the evocation of knowledge and understanding. Third, the game prompts more questions than it answers, inviting player speculation about its subject matter. Such “questions” include the use of a specific ant species with a unique appearance, but a lack of explicit information about this species; a selection of foliage and environment art which suggests a specific location, but again, a lack of explicit location information; and the inclusion of expert subject matter in form of game objects, such as the ants’ fungal gardens and the effects of the microfungus escovopsis, presented in a manner which allows for successful play but which does not explicitly identify these objects. 35 56 Gold, Rich. The Plenitude: Creativity, Innovation, and Making Stuff. The MIT Press, 2007. Print. 57 Brown, John Seely. Upside Magazine. Intervew by Marcia Conner. 1993. Web. 21 Mar 2011. <http:// www.johnseelybrown.com/linklearn_int.html>. 58 "Interactive Media Division Forum for 9/24/08." YouTube. Web. 21 Mar 2011. <http://www.youtube.com/watch? v=3bn327W5y_g>. 59 Cliff and Grand 82 As an evocative knowledge object, Leafcutters seeks to raise questions while engaging the player in play, thereby cultivating curiosity about the subject matter. This pattern of engagement is also utilized in The Cat and the Coup, which takes a similar approach to its rich subject matter: the game provides enough prompts and bits of knowledge to engage the player’s curiosity—without filling in all of the details. 60 Adaptation Game Leafcutters is a game designed by adaptation from a natural system. At the most obvious level, the narrative and artistic elements of the game are modeled on the subject matter. The game system is also an adaptation, designed to accurately represent the natural systems of leafcutting ants. Furthermore, the play strategies which arise in this game system mirror behaviors found among real ants. Leafcutters combines these multiple layers of adaptation in order to faithfully represent leafcutting ants in video game form. Educational Game Leafcutters also functions as an educational game. James Paul Gee presents a list of principles for games as learning tools. 61 From this list, Leafcutters particularly features the concepts of co-design, system thinking, and leveraging the concept of meaning as action image. Gee states, “In good games, players feel that their actions and decisions—and not just or primarily the designers’ actions and decisions—are co-creating the world they are in and the experiences they are having.” 62 Leafcutters provides the player with such co-creation in three ways. First, the player’s shaping of the ants’ behaviors results in visual patterns of movement of 36 60 The Cat and the Coup. Peter Brinson and Kurosh ValaNejad, 2011. 61 Gee, James Paul. "Learning by design: Games as learning machines." Interactive Educational Multimedia. 8. (2004): 15-23. Print. 62 Gee 17. both the ants and the various objects in their ecosystem, allowing the player to author patterns of change in the environment. Second, the population of the colony grows over the course of normal play, amplifying this customized systemic pattern. Third, the ants dig at the player’s instruction, forming a unique nest architecture, resulting from this interaction. These three factors allow the player to feel ownership and investment in the colony, thinking of it as his own. System thinking is the heart and soul of Leafcutters. The game guides the player through a process of connecting the actions of an individual ant to the complex actions of the colony as a whole. By allowing (in fact, requiring) the player’s interactions to alter the decision making of multiple ants simultaneously, Leafcutters demonstrates the multiplicative affect of individual actions taken by many actors. Gee writes that meanings of concepts are best learned when the student can associate them with an action image—a lived experience. For games to do this, they must “make the meanings of words and concepts clear through experiences the player has and activities the player carries out, not through lectures, talking heads, or generalities.” 63 Leafcutters allows the player to live through the growth (or death) of an ant colony, thereby tying concepts of the workings of an ant colony to very particular action images. The game provides the experience of guiding and living with an ant colony that is not possible in the real world, but which remains authentic to reality, so that these action images provide valid understanding of natural phenomena. Fiction and Nonfiction in Leafcutters Leafcutters blends fiction and nonfiction in an attempt to accurately represent ants in a way that is prohibited by real-life constraints. 37 63 Gee 22. Since the game is necessarily abstracted and oversimplified, its nonfiction elements are adaptations of reality, and therefore no aspect of the game is entirely untouched by design considerations. One touchstone of realism in the game is the ants’ behavior-making autonomy. Each ant makes decisions, and this group of independent actors, acting in dense proximity (but not formation) gives rise to the sense of the colony’s intentions and actions. Never in Leafcutters will one actor give orders to another. The queen of the colony is not a monarch; as in reality, she never commands anyone. Leafcutters seeks to faithfully recreate the ecosystem in which the ants live—presenting a cross-sample of the types and magnitudes of danger which the ants might experience in the form of a simplified representation—for example, one type of “food” object represents all foods which the ants eat. And of course, this adapted ecosystem is balanced for the purposes of entertaining play. For example, since the scope of the game has been artificially limited to one colony, this colony must not be as prone to sudden destruction as might truly be the case. Likewise, the number of ants in the player’s colony is much lower than the population of a real colony. Such deviations from the biology were accepted since they allowed the game to function well, while maintaining the spirit of the natural system. Play In Leafcutters: By the Book, Off The Beaten Path Leafcutters is designed to support two extreme modes of play. On the one hand, the game supports a “by the book” style of play in which the player follows biology faithfully, and recreates a true-to-life system of an ant colony. In fact, this mode of play is encouraged by the interim goal prompts which the game provides to the player. On the other hand, the game supports a form of play in which the player attempts to explore the outer limits of the system, by attempting to break the game or create outlandish behavior. For example, an early playtester deliberately attempted the unique strategy of leading 38 the queen outside of the nest to birth larva directly at the sources of food. Rather than outlaw such exploratory play, Leafcutters serves as a virtual laboratory for player experimentation within an abstracted ecosystem. In this way, Leafcutters harnesses a strength of games as learning tools: games allow for the exploration of the limits of systems, allowing the player to challenge the provided model and determine its breaking points and outer limits. Expansion and Further Research As described earlier, Leafcutters has been designed with the specific intention of evoking real-world engagement. Further research could formally assess whether the project achieves this goal and, if so, how such success could translate into design considerations for future projects. Such knowledge could benefit the design of educational games which seek to evoke learner engagement in educational content. More generally, these principles could point the way to a genre of game which achieves unique richness and resonance by adapting evocative systemic elements of the real world. Conclusion Leafcutters emerges from the fields of A-life, life simulation games, virtual pets, and swarm games, leveraging the specialties of each in order to present an experience of play designed to evoke engagement with a natural system. The game was designed through a process of adapting the natural systems of leafcutting ants in order to allow a previously impossible form of play with this real-world subject. By representing the virtual ants in a lifelike manner, the game seeks to establish player empathy with the creatures. The game requires the player to solve problems using the decentralized logic of an ant colony. While the game is built on scientific knowledge, it presents this knowledge in an implicit, rather than explicit, way, seeking to raise the player’s curiosity about the subject matter. By facilitating this mental and emotional engagement with a natural system, Leafcutters is designed to evoke engagement with the real world. Further 39 research could determine the project’s efficacy toward this goal and establish design principles which could benefit educational game design as well as point the way for deeper exploration of games based on adaptations of natural systems. 40 Bibliography Benthall, Jonathan. Science and Technology in Art Today. London: Thames and Hudson, 1972. Print. Black and White. Electronic Arts, 2001. Brinson, Peter and ValaNejad, Kurosh. The Cat and the Coup. 2011. Brown, John Seely. Upside Magazine. Interview by Marcia Conner. 1993. Web. 21 Mar 2011. <http:// www.johnseelybrown.com/linklearn_int.html>. Cliff, Dave, and Grand, Stephen. "The Creatures Global Digital Ecosystem." Artificial Life 5. (1999): 77-94. Print. Creatures. Mindscape, 1996 Dungeon Keeper. Electronic Arts, 1997. Furby. Tiger Electronics, 1998 Gee, James Paul. "Learning by design: Games as learning machines." Interactive Educational Multimedia. 8. (2004): 15-23. Print. Gold, Rich. The Plenitude: Creativity, Innovation, and Making Stuff. The MIT Press, 2007. Print. Hölldobler, Bert, and Wilson, Edward O. The Ants. Belknap Press, 1990. Print. Hughes, Diana. "Pluff: Creating intersections between tactile interface devices and fabric based electronics". MFA thesis. University of Southern California, 2009. "Interactive Media Division Forum for 9/24/08." YouTube. Web. 21 Mar 2011. <http://www.youtube.com/ watch?v=3bn327W5y_g>. Langton, Christopher G. "Studying artificial life with cellular automata." Physica D: Nonlinear Phenomena. 22.1-3 (1986): 120-149. Print. Langton, Christopher G. "Artificial Life." SFI Studies in the Sciences of Complexity. (1988): Print. Langton, Christopher G. Artificial Life. Redwood City, CA: Addison-Wesley, 1989. Print. Lehninger, Albert L. Principles of Biochemistry. 1st ed. New York: Worth, 1982. Print. Lemmings. Psygnosis, 1991 Mateas, M. "Expressive AI: A hybrid art and science practice." Leonardo, Journal of the International Society for Arts, Sciences, and Technology. 34.2 (2001): 147-153. Print. Meehan, James R. "Tale-Spin, an interactive program that writes stories." Proceedings of the 5th International Joint Conference on Artificial Intelligence. San Francisco, CA, Morgan Kaufmann Publishers Inc.. 1977. Print. Nelson, Randy. "Will Wright explains what The Sims and an ant colony have in common." Joystiq.com. 2010. 21 Mar. 2011 <http://www.joystiq.com/2010/11/08/will-wright-explains-what-the-sims- and-an-ant-colony-have-in-com/>. 41 Nintendogs. Nintendo, 2005 SimAnt. Maxis, 1991 SimCity. Maxis, 1989. SimLife. Maxis, 1992. SimTower. Maxis, 1994. Sims, The. Electronic Arts, 2000. Slaves to Armok: God of Blood Chapter II: Dwarf Fortress. Bay 12 Games, 2006. Swarm. Hothead Games, 2011 Tamagotchi. Bandai, 1996. Pest Detective. 21 Mar. 2011 <http://www.pestworldforkids.org/pest_detective/index.html>. Pets (Dogz and Catz). PF Magic, 1995 Pikmin. Nintendo, 2001 Pleo. Ugobe, 2006 Populous. Bullfrog, 1989. Schaffer, Simon. "Enlightened Automata." The Sciences in Enlightened Europe. Ed. Clark et al. Chicago and London: University of Chicago Press, 1999. 126-165. Print. “Swarm - Hothead Games” 21 Mar. 2011. <http://www.youtube.com/watch?v=F_xRPB5Kijk> Tsutsui, Neil. Personal Interview by William Graner. 1 Nov. 2010. von Mammen, Sebastian, and Christian, Jacob. "Swarming for Games: Immersion in Complex Systems." Applications of Evolutionary Computing. Ed. Mario Giacobini, Anthony Brabazon, Stefano Cagnoni, Gianni Di Caro, Anikó Ekárt, Anna Esparcia-Alcázar, Muddassar Farooq, Andreas Fink, and Penousal Machado. Heidelberg: Springer Berlin, 2009. 293-302. Print. Webkinz. Ganz, 2005. Wirth, R., H. Herz, R.J. Ryel, W. Beyschlag, and B. Hölldobler. Herbivory of Leaf-Cutting Ants: A Case Study on Atta colombica in the Tropical Rainforest of Panama. Berlin, Heidelberg: Springer- Verlag, 2003. 18. Print. 42

Abstract (if available)

Abstract Leafcutters is a life simulation game about leafcutting ants which is designed to evoke engagement with real world subject matter. In this game, players shape the behaviors of a colony of ants in order to establish complex behaviors such as foraging and fungus farming. The game system in Leafcutters is adapted from existing biological research on ants, with an emphasis on the accurate adaptation of a natural system into a game system. This project draws on previous works in artificial life, life simulation games, swarm games, virtual pets, and virtual ants. Leafcutters is a work of expressive AI, an evocative knowledge object, and an educational game.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

By nature: an exploration of effects of time on localized gameplay systems

PDF

Day[9]TV: How interactive Web television parallels game design

PDF

Exploring empathy through negotiation

PDF

Aquariyum! Designing an organic asynchronous gameplay experience

PDF

Reality ends here: environmental game design and participatory spectacle

PDF

The moonlighters: a narrative listening approach to videogame storytelling

PDF

Toward a theory of gesture design

PDF

Ascension: an analysis of game design for speech recognition system usage and spatialized audio for virtual reality

PDF

SomeDay: designing a game about different thought processes

PDF

The witch: Identity construction and the fairy tale in interactive narrative

PDF

Fall from Grace: an experiment in understanding and challenging player beliefs through games

PDF

Garden designing a creative experience with art and music orchestra

PDF

Aesthetic driven design of Space Maestro

PDF

BeachcomberVR: inspiring exploration: concepts in the production of a virtual reality game

PDF

There You Are: an exploration of storytelling methods using in video games

PDF

CODA: the making of a team reliant first person shooter game

PDF

Wetware: designing for a contemporary dilemma

PDF

Paralect: an example of transition focused design

PDF

Penrose Station: an exploration of presence, immersion, player identity, and the unsilent protagonist

PDF

The tree as storied experience: an experiment in new narrative forms

Asset Metadata

Creator Graner, William B. (author)

Core Title Leafcutters: life simulation gameplay designed to evoke engagement with real-world subject matter

School School of Cinematic Arts

Degree Master of Fine Arts

Degree Program Interactive Media

Publication Date 05/05/2011

Defense Date 05/05/2011

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag educational game,evocative knowledge object,OAI-PMH Harvest,simulation game,video game

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Gibson, Jeremy (committee chair), Anderson, Steven F. (committee member), Fullerton, Tracy (committee member)

Creator Email bill@bgraner.com,granerw@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3919

Unique identifier UC1293017

Identifier etd-Graner-4578 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-469526 (legacy record id),usctheses-m3919 (legacy record id)

Legacy Identifier etd-Graner-4578.pdf

Dmrecord 469526

Document Type Thesis

Rights Graner, William B.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu