Designing Games for Learning

Editor's Notes

• #3: Early on I recognized that there are two important aspects of serious games. Rod Myers has great expertise on the gaming aspects, so I invited him to help out. Rod will now briefly summarize the introduction in the report.
• #4: When deciding whether to undertake the creation of a serious game, designers should understand the types of learning that games can and cannot promote in order best to leverage the distinctive features that games offer. First, a well-designed game can provide authentic practice in thinking and working in specific roles and contexts. In order to overcome obstacles, players must solve autnentic problems by formulating strategies, testing hypotheses, and reflecting on their failures and successes. Next, collaborative gameplay can provide experiences in which players learn to recognize and draw on the resources of their fellow players. These shared experiences can be collectively examined, discussed, and recalled when relevant to new situations. In the immersive game cycle of playing, failing, reflecting, and trying again, players experience feelings of pleasure and increasing motivation and gain a sense of control and autonomy. The resultant feeling of self-efficacy is an important influence on persistence and willingness to undertake new learning tasks.
• #5: Many professions that involve hazardous conditions and/or responsibility for the health and safety of others (e.g., military, police, fire fighting, surgery) have turned to games and simulations to provide practice in thinking and acting under pressure in critical situations. And finally, games can be designed so that they provide appropriate and variable levels of authenticity, which can be useful in reducing cognitive load for novices so that they can focus on the most critical aspects of a task. At the same time, the level of difficulty can be dynamically tailored to the learner’s current knowledge and skills to provide optimal challenge.
• #6: In addition, when deciding whether a serious game is an appropriate instructional strategy, we must consider factors regarding the potential effectiveness and efficiency of the desired learning experience. In terms of effectiveness, can the skills required to accomplish the learning goals be modeled as actions in a game with sufficient fidelity? Are the tasks sufficiently variable and complex such that elaboration is necessary to scaffold learners to the desired performance? Is there a risk of injury or death when performing in the real world? Each of these indicates that a serious game is a viable option. In terms of efficiency, a complex game that includes the necessary instructional overlay will probably require more resources for development than traditional instruction. However, if the potential audience is large or if instructor time is limited, a serious game may be worth the additional time and money. If inadequate instruction may have dire consequences, then effectiveness may be more important than efficiency. Now Charlie will summarize the “fuzzy vision” of serious games as described in the report.
• #7: Picture the scenario, roles, and contextual factors for the game as being highly authentic – that is, consistent with whole, real-world tasks, including portrayal of values, attitudes, beliefs, and cultures and provision of situational understandings. This means it is usually a multi-player game, though non-player characters (NPCs) are often created to play some or all of the other roles. This makes it a serous game. Picture the game as having levels of difficulty/complexity, each of which must be mastered by each player (role) before the players are allowed to progress to the next level, to avoid cognitive overload. Within a level of difficulty, cognitive load may be further reduced, if necessary, by reducing representational fidelity. Imagine 3 forms of scaffolding for a serious game … A virtual mentor can be used for both the coaching and instructing, which can be provided automatically at a set point in the game, or can be triggered by a player action, or can be requested by a player. The game is paused JIT for instruction, could be for coaching. This scaffolding makes the learning process quicker, easier, and more enjoyable.
• #8: During instruction, whatever needs to be learned at that point in the game must be mastered before the game can continue. So individual mastery is assessed in the instructional overlay, not in the game – team performance is assessed in the game. This ensures individual mastery, it ensures transfer, and it ensures automatization when important. Feedback is one of the most important features of a powerful serious game. We’ve identified four ways to provide feedback. … Motivation is key to learning. Several things can be done to enhance player motivation. Each role in the game can have a score based on performance. Motivation is also enhanced by collaboration (when that is a natural part of the task); it is enhanced by the authenticity of the task, and it is enhanced by building confidence or expectancy for success, as John Keller talks about in the ARCS model. Confidence is built by having progressive levels of difficulty, as in most video games, and by generating mastery in the instructional overlay. Now Rod will summarize the elements of the game space as described in the report.
• #9: The elements of the game space are the aspects that must be designed to create the necessary conditions for the game experience. We are not attempting here to prescribe a process for designing a game for learning. Instead, we are trying to provide guidance in making design decisions regarding these elements to promote learning.
• #10: Here is a depiction to aid in understanding the relationships among the elements of the game space. In essence, players enter the game space and interact with elements in the game environment, using objects and information to gain knowledge and skills, overcome obstacles, and achieve goals that are closely related to real-world learning goals.
• #11: We think it makes sense to begin with the desired learning outcomes and the kinds of tasks that the learner will undertake in the post-instructional environment. These should guide all subsequent design decisions for the rest of the game elements. The fundamental principle here is the authenticity of the goals and tasks. Normally the goals will be broken down into subgoals that require the completion of a variety of whole, authentic tasks of optimal difficulty for the learner. These subgoals will usually be structured as levels of increasing difficulty and complexity that build on the learning achievements of previous levels. This game cycle scaffolds the learner toward self-efficacy and mastery.
• #12: Game mechanics are the fundamental actions a player may take to complete tasks. In performing these actions, players come to understand the underlying rules of the game and to formulate strategies for leveraging those rules. Therefore, instructional designers should conceive of the desired learning outcomes as actions—including cognitive actions—that form the basis for playing the game. We have identified three types of mechanics and their relationship to learning. Core mechanics are most fundamental in accomplishing the goals of the game. As such, they should recur frequently to provide sufficient practice and become skill-based. Compound mechanics are combinations of mechanics, usually including one or more core mechanics that have already become skill-based. For example, in popular videogames, once the player has learned to drive vehicles and shoot guns, he may be required to do both at once. Peripheral mechanics are optional and thus require more cognitive processing. While they are not directly related to the learning goals, their novelty may increase engagement.
• #13: Game mechanics and rules are closely related as rules govern the possibilities of and constraints on actions. When determining the rules for game mechanics, designers must consider both the ways in which they have been instantiated in other games and how they work in the real world to promote transfer of expertise. Games are complex systems, and it’s not always possible to predict the effects of even small changes to rules, mechanics, and other game elements. Therefore, designers usually conduct frequent playtests, a kind of formative evaluation during prototyping, to observe players’ decisions, questions, difficulties, and overall experience of the game.
• #14: Just as we endeavor to design learner-centered instruction, game designers strive to create player-centered gaming experiences. Given the social nature of learning, it makes sense whenever possible to create multi-player games in which players take on roles and either compete or cooperate (or both) to achieve the goals of the game. Avedon identified eight possible configurations of players seen here. Game dynamics are not a designed element but instead emerge as a result of players interacting with the mechanics and rules of the game. In poker, bluffing is an example of an emergent behavior; it is not part of the rules but was invented by players trying to gain an advantage within the constraints of the game.
• #15: Game environments range from single-setting board games like Clue to videogames with numerous, diverse settings such as the spaceship and locations on several planets in Mass Effect. How the player moves within the environment depends on the structure and dimensionality of the environment and is related to how the player perceives the environment. So-called “god games” are generally isometric or top-down with the players acting on the environment, while in first- and third-person games the players are acting in the environment. Physics is important in some analog games like billiards and tiddlywinks, while in videogames a physics engine may handle the rules governing how objects move and respond to force. The important thing here is determining the degree of fidelity required to achieve the desired learning outcomes. And finally, designers must recognize the distinction between play time—the time it takes to play the game—and event time—the time taken in the game world. Event time may be real, compressed, extended, or variable, and it may be used by the designer to adjust difficulty and challenge for the player.
• #16: Objects are all of the components that players use or take action on through game mechanics. Borrowing from the vocabulary of cinema, diegetic objects exist in the game and, when the game includes an avatar for the player, are accessible by the avatar. Examples include vehicles, tools, buildings, furniture, and so forth. Non-diegetic objects are accessible to the player, mainly through an on-screen interface. Examples include menus, maps and heads-up displays, and other means of obtaining information about or controlling the game. Objects have properties with static or dynamic states. For example, an avatar’s backpack may have a static property for the amount of food it can hold and a dynamic property for the amount of food it currently contains.
• #17: Every action creates a change in the state of the game system. Players’ decisions regarding actions are guided by the available information about the game state. Many game objects are conduits for information, which may be presented as text (e.g., a popup window with instructions; a letter from a non-player character), icons (e.g., a weapon icon that indicates which weapon is currently active; a health meter), or visual/aural attributes of objects that serve as cues of state (e.g., a clicking sound to indicate that the chosen door is locked). We have identified five types of information that may be available to the player: information about avatars, about objects, about events—which includes feedback for actions as well as narrative information—about the environment, and about the system itself, including access to the instructional overlay when a player needs practice or assistance beyond that provided in the game setting. Access to information is on an accessible-inaccessible continuum, the exact position depending on a number of factors including authenticity (how much the player would know in the real-world situation being simulated), level of difficulty (withholding information can increase difficulty), and cognitive load (too much information overloads working memory).
• #18: This is a vast and ever-changing element. Technologies used in games range from stones to paper and pencil to molded plastic to joysticks and motion detectors. As new technologies are invented, old games will be ported to them and new games will emerge as designers explore the new affordances and possibilities. Digital game-based learning offers the ability to gather detailed data about a player’s choices and actions. These data can be used to assess performance and also to personalize gaming experiences, especially by dynamically adjusting difficulty for a player to promote immersion and flow.
• #19: Perhaps the most important aspect of narrative in games for learning is that it taps the power of episodic memory for structuring and storing our experiences. A clear narrative structure such as the hero’s journey identified by Joseph Campbell can provide both a familiar frame of reference (an idea of what to expect) and a cognitive frame of reference (or schema) to promote recall. Genre can influence the narrative structure of a game. A first-person shooter like Halo has some narrative trappings like a linear plot held together by cut scenes, while a role-playing game like Skyrim has customizable avatars and multiple storylines that vary according to the player’s actions and choices. But even within a particular genre, narrative may be employed in very different ways. Games often require the player to assume a role within the narrative and to take action in a manner consistent with that role. Shaffer has argued that a player’s role in a game for learning should be based on an epistemic frame, which he defines as a set of “skills, knowledge, identities, values, and epistemology that professionals use to think in innovative ways.” Clearly this approach is related to the authenticity of a game.
• #20: When we talk about the aesthetics of games, we are referring to the emotional responses and felt experiences that arise in the player(s) through interaction with/in the game system. Design decisions for all of the other game components create the overall aesthetic experience of the game, so deciding how a player will feel while playing the game is a crucial part of the fuzzy vision early in the design of the game. A short list of terms for describing the aesthetics of games might include feelings of challenge, fellowship, discovery, expression, and fantasy. Naturally, more than one of these feelings may be present at any given time, and all may occur throughout gameplay. If the game follows a narrative, the type and timing of events and their emotional flow may be dictated by the story arc. Decisions regarding aesthetics are directly influenced by the degree of authenticity required to achieve learning and transfer. The dimensions of authenticity should be consistent in their levels of realism (or fidelity), where realism ranges from completely abstract to highly realistic. Types of realism include physical, perceptual, functional, cognitive, and emotional.
• #21: A game is a designed experience comprised of the ten elements described above. Game designers can create the conditions for the desired experience by applying the fundamental principles described previously when designing these elements. Through this process, designers iteratively refine the fuzzy vision of the game. However, as with any medium used for instruction, in serious games special attention must be given to the instructional methods to ensure that players attain the learning objectives. Now Charlie will summarize elements of the scaffolding as described in the report.
• #25: Formats: There are two formats, each of which exists on a continuum. One continuum is part-task selection, which concerns whether the system teaches all the part-tasks to all players in a given role on one extreme, or diagnoses each player’s needs regarding part-task instruction on the other extreme. A midpoint on this continuum is for the instruction on each part-task to start with medium-difficulty practice (with feedback) using a computer-adaptive testing algorithm, and teach more or fewer part tasks as needed. A second continuum for format is use of a virtual mentor, which ranges from extensive use of virtual mentor at all stages of the instruction to no use of a virtual mentor for any stages of it
• #32: To summarize, we began by describing six fundamental design principles that can be used to create a holistic view of a game for learning. Decisions regarding these principles then guide subsequent design decisions for the elements of the game space and the elements of scaffolding. The elements of the game space are the aspects of the game that must be designed in order to create the necessary conditions for the game experience. These elements are so intricately inter-related that decisions regarding one invariably influence decisions about the others. We think that it is especially important that the goals of the game are closely aligned with the learning goals so that winning the game requires achieving the learning goals. The elements of scaffolding are focused on ensuring that those learning goals are achieved by the players. Adjusting aspects of the game is the subtlest approach as it occurs within the flow of the game. It is closely related to the levels of difficulty, game mechanics, and rules. Coaching is more overt, but it may occur within the game space or preceding or following gameplay. Instructing generally requires that the game be paused. Strategies for instruction and assessment vary depending on the kind of learning involved.