Labyrinth is the working title for a game for learning still in development at The Education Arcade at MIT, for which I have been a designer for more than a year. I've chosen this game as my case study because of my familiarity with the game and the reasons behind the design decisions we made. Also, the game has many of the features this thesis is focusing on, including clearly increasing representations of mastery, Web and mobile components, a system of reputation, and multiplayer collaboration.
First, let me describe the game. Labyrinth aims to help middle school students learn math and literacy. The game is meant to be played with a small group of classmates, though not necessarily in the classroom or during school. There is little or no time pressure during any of the game's challenges, so two or three players could sit together at the same computer and discuss strategies without fear of needing fast reflexes or running out of time. Players engage with the math component of the game through solving puzzles. They improve their literacy skills through the in-game message board, where they read and write to share strategies with teammates. They also improve their literacy skills by reading the story within the game, which is presented in comic book style. The game includes a dynamic difficulty system, matching the challenges it presents to players' skill levels. How do the design decisions made for Labyrinth affect mastery motivators?
Personal Relevancy in Labyrinth
Labyrinth aims to create personal relevancy in a number of ways. There are 12 puzzles in the game, each with four levels of difficulty. Labyrinth gives players a range of choices about which puzzles to play at any given time. If a player finds a particular puzzle boring or frustrating, she can simply choose to spend time with other puzzles. If she finds the puzzle especially interesting, she can spend more time with it, working through each of its four levels of difficulty. This freedom of choice allows players to maximize personal relevancy. True, the game requires players to achieve some level of mastery with all of the puzzles, but since teams play together, someone else on the player's team will probably solve each puzzle eventually and contribute strategies to the rest of the team. So, any individual player who dislikes a particular puzzle can simply wait for other teammates to tackle it and benefit from their strategies. Players hoping to achieve a competitive personal score, though, will have to contribute strategies of their own for at least some of the puzzles, so the game doesn't encourage them to simply sit back and let teammates do all the work.
Labyrinth also creates personal relevancy through avatar choice. Players can't customize every aspect of their avatar, as some games allow, but they are able to choose a mask that uniquely identifies them within their team. Even this small level of customization makes players more attached to their avatars. This representation of them shows up in the gaming world, in the story comics, and in the message board. They become more attached to their avatars for two reasons. It represents a choice they made. If the game requires them to use a particular avatar, even if it was the same one they would have chosen from a list of options, players don't feel as much ownership unless they can make the choice themselves. They also likely choose an avatar that appeals to them more than a default avatar would. We have chosen not to represent gender or race through the avatar. The masks available to players all look like different creatures or monsters, so race and gender are largely irrelevant. This simplifies the design, reduces production requirements, blends well with the story, allows us to target a wider age range, and helps the game keep from going out of date as fashions change. Playing an unfashionable or too-young character reduces personal relevancy tremendously, so we aimed to avoid that.
The story also helps create personal relevancy. It helps entertain players, convincing them that the game is worth their time and attention (relevant to them). However, if the story fails to entertain some players, it may reduce personal relevancy. It also situates them in the game world in a heroic role, adding significance to all of their actions. Finally, it creates mystery, leveraging players' curiosity to drive them forward in the game.
The physical space of the game world creates personal relevancy for players in much the same way the story does. It entertains them and convinces them that the game is worth their attention. It makes the context in which their avatars exist seem more believable, adding weight to any actions and accomplishments in that context. It supports the story, letting players interact with the same characters and space they see in the comics. It supports the avatar choice, letting players see their characters in action. And, it supports puzzle choice, anchoring the existence of puzzles in a physical space and representing completed puzzles as obliterated (crumbled, overrun spaces). Seeing obliterated puzzle rooms in a physical space rewards prior successes with puzzles and motivates future successes.
Mobile and online access to Labyrinth creates personal relevancy through accessibility. The more accessible Labyrinth is to its players, the more they will keep the game in mind and the more it will become relevant to them. That's why the game will be playable from any computer with Internet access. Players can log in sometimes from school and other times from home or a friend's house, whatever his most convenient. To make the game even more accessible, we are designing a way to play the game on mobile devices as well. We haven't finalized the mobile platform for Labyrinth, but for the sake of this thesis, let's imagine it being implemented on the Nintendo DS. The DS is a portable gaming device that folds in half. Each half has a screen, and the bottom screen is touch-sensitive (it comes with a stylus). The console also has a microphone, speakers, a direction pad and several other digital buttons, and a wireless network connection (802.11b WiFi). The device is small enough that middle school students could easily keep it in their backpacks at all times. They could then play the game whenever they had a few free minutes, such as bus rides, lunch breaks, after school, and at home. All of the preset game data would be kept on a cartridge in the DS, including artwork, puzzles, etc. Communication with the online game server would happen whenever the player was within range of a wireless hotspot, at which point the game would synchronize data like the player's progress and messages to and from teammates (which would also be saved locally on the game cartridge). The asynchronous design of the connectivity elements allows this kind of play style, being online sometimes and off-line other times. Having access to the game from any connected computer as well as from mobile allows players to time their play to best fit their mood and schedule. Knowing that Labyrinth will always be there encourages players to think more about the game and develop more of an attachment to it
Finally, players' individual scores increased personal relevancy. There are several ways that players can increase their individual scores in Labyrinth, including freeing pets by solving puzzles, solving puzzles at higher levels of difficulty, writing strategies that teammates find helpful, and completing the game. In addition to the more precise representation of individual mastery within the game, this score gives players a general sense of increasing mastery. Players can use this score to compare their performances with others in an online leaderboard. They can also ignore the leaderboard, if they're only interested in their own progress relative to before.
There are a few themes from all of these examples of heuristics for creating personal relevancy. Giving players meaningful choices increases their attachment to the results of those choices. Simply entertaining players motivates them to take the game seriously and want to succeed within its rules. And, making the design coherent allows features to support each other and enhance personal relevancy (as adding the physical space did for the other features).
Social Relevancy in Labyrinth
Labyrinth creates social relevancy for players in many of the same ways it creates personal relevancy. It's a multiplayer game. Simply having others there to notice what the player does adds consequence to his actions. The team score represents this directly. Each team member contributes to the team score, so if one player doesn't pull his weight he will drag the whole team down a notch. Conversely, a strong contribution to the team score could make him a hero. The weight of the consequence for any particular contribution to the team score depends upon whether the team takes its score seriously. It's possible that no one on a particular team cares about score. Ignoring score is a form of choice that can increase personal relevancy, even if it decreases social relevancy. Team scores will be entered in the online leaderboard, so the team can see how it compares to other teams. We envision there will be some teams that want to compete directly with each other, such as different teams in the same class or school or teams at rival schools. The team score will support this kind of play, which enhances social relevancy.
Mobile and on-line access to Labyrinth create social relevancy through accessibility and by bringing play to different social situations. Since players can connect to the game from mobile devices or any connected computer, they can frequently play together. They can play on the same computer or the same mobile, or they can play side-by-side on different devices. Either way, they become more interested in each other's status and progress by virtue of being within sight and therefore within mind of each other. The primary side effect of social relevancy that we are trying to encourage in addition to increased motivation for mastery is encouraging players to discuss their strategies with each other. This process of discussion helps players solidify their own thinking, figure out how best to communicate their thinking, and become stimulated by others' thinking. In our weekly play testing sessions with middle school students, we noticed that some students who played together did discuss and share strategies, and others took turns to see who would solve the puzzle first on his own. Players who solved the puzzle on their own or with the group seemed to experience the satisfaction of increasing mastery, but players who sat by and watched as a teammate solved the puzzle seemed to have reduced personal relevancy. For the successful players, at least, the social relevancy and visibility strengthened personal relevancy.
Allowing players to choose which puzzles to focus on and in which order creates social relevancy. Since one of the main ways that players can contribute to their team's success is by solving a puzzle first and contributing a written strategy for that puzzle to the in-game message board, being able to choose which puzzle to play is important. As teammates branch off and solve different puzzles first, each can feel valuable by contributing strategies. If there were a fixed order to all of the puzzles, teammates would feel that they were more in competition with each other to solve the next puzzle first. Competition within the team can create social relevancy as well, but we are actively trying to diminish anxiety about math among students. Therefore, having at least the team unit be as cooperative as possible is an opportunity to create a supportive atmosphere where players feel comfortable asking questions. We essentially want to create an affinity space (Gee 2004) for the game that's also in the game. We want players of different skill levels to share the space, engaged in a common endeavor, all bringing their unique knowledge and skills and contributing in the ways they prefer. That cooperative environment is important for supporting both the math and literacy goals of the project.
By contributing strategies to the team message board, players create an opportunity to build a positive reputation within the team. Each strategy contribution can be rated by other team members as helpful or not helpful, à la product reviews on Amazon. This discourages players from contributing unhelpful or incomplete strategies. If a strategy makes sense but is poorly explained, it may be rated as helpful by some team members and not helpful by others. All of these ratings affect a player's individual score and reputation within the team on the message boards. If a player contributes useful strategies and gathers a handful of positive ratings, team members will probably pay closer attention to that player's strategies in the future. Since we only planned for six players per team, though, every message board post may get plenty of attention anyway. Regardless, a positive reputation given to a player by the team will be respected and valued by that team as well. That's social relevancy.
There are a couple of features that have the potential to both increase and decrease social relevancy. First, the counterpart to contributing strategies is asking for help. Different teams will respond differently to players who ask for help. Some teams will be supportive, supplying answers and encouraging further questions. Other teams will likely be critical, making fun of the perceived ignorance or incompetence of players who ask questions. In the supportive example, asking and answering questions becomes part of the cycle that builds a supportive atmosphere within the team and increases social relevancy. Additionally, contributing a strategy feels more socially valuable when a teammate has requested that strategy than when it comes unsolicited. In the negative examples, players may perceive the team as not worth impressing and disengage, decreasing social relevancy. We think that most teams will be supportive, but certainly not all.
Another feature that could either increase or decrease social relevancy is the game's support for paired play. We designed the game without strict time pressure or need for reflexes, in part because we wanted more than one player to be able to play the game together at the same computer. We expect that most of the time, especially outside the classroom, players will play one to a computer. However, in the cases where the game is used in the classroom, there is a high probability that there are more students than computers and the teacher may assign two or three students to a computer. This kind of play generally increases social relevancy within the small group of players sitting at the computer. If there are two players at the same computer and one of them figures out how to solve a puzzle, she has another person sitting next to her who she can immediately impress. On the other hand, neither player now has full control over the choices of which puzzles to play, which avatar to use, when and how to contribute strategies, etc. Neither player feels completely responsible for progress through the game, and therefore may not feel that the individual score, the number of pets freed, or the message board reputation accurately reflect her contribution, decreasing personal relevancy. This is a fair compromise, because of the opportunity for players to discuss strategies back and forth vocally and make better progress in the game and potentially in their thinking.
What themes or design heuristics can we draw from these examples? Simply adding other players into the mix increases social relevancy. Cooperation increases relevance anymore for some players and competition increases it more for others. In Labyrinth, we chose a hybrid model, players collaborate with each other to form teams and then compete as a group with other teams. All competition is automatic but can be ignored.
Personal Visibility in Labyrinth
We've put a number of features into the game specifically to help increase personal visibility of mastery. Most obviously, the player's goal is to free as many of the trapped pets in the factory as possible, and as the player achieves more success in the game, she frees more pets. She can see the number of pets she has freed so far in the upper left corner of the screen. She can see this number of pets increasing as she makes progress within a puzzle. However, due to constraints on the number of art assets we could produce for the game, we have chosen not to animate pets being freed. That's a missed opportunity for personal visibility of mastery and an area we could have improved with a larger budget. The player can also visit a room in the factory where all of the free pets have gathered, visually representing success within the physical space of the game world.
Once players successfully solve a puzzle three times, they graduate to the next highest level of difficulty for that puzzle. When they solve all difficulty levels of the puzzle, they obliterate it. Both are made visible in the game to indicate progress to the player. When a player graduates to the next difficulty level of a puzzle, she receives a key and a clue about where the room containing that puzzle is within the factory. That key and the search for the next room both reinforce for the player that she is making forward progress in the game. When she finally obliterates a puzzle, she will feel the same progress to an even greater degree. The danger of making all this so visible is that players who are unable to obliterate one or more puzzles will be fully aware that they have not overcome all of the challenges presented. This visibility of the lack of complete success will motivate some players to invest more effort and will frustrate some players and encourage them to disengage. Frustration will probably occur more frequently if a person is playing alone, because those with a team to support them will likely progress faster. Some of the players going through the game alone who become frustrated will then seek out a team experience, and that's good.
Physical exploration helps players visualize progress in the game as well. Each puzzle exists in a room, and players will have to traverse a physical space to find the correct room before and during any puzzle. Exploration of the physical space makes players more familiar with possible room locations and helps them make educated guesses about where puzzle rooms might be. Exploration is its own kind of mastery, visually represented, but in this case we have also linked it to mastery of the puzzle. When players successfully complete key challenges (finding which door a key goes to), they actually leave one space (the hallways linking all of the puzzle rooms) and enter another space (a puzzle room). Each puzzle has its own interesting set of artwork, so the game rewards the player immediately with a new scene they haven't seen before (or at least a variation on an old scene).
Increased accessibility of the game through mobile and online play keeps the game and mastery within it visible to players. Because the game is so accessible, players will interact with it more frequently. By interacting with it more frequently, they will focus more on the game and what it emphasizes, namely their growing mastery.
Finally, individual and team scores help players instantly and quantifiably identify progress in the game. Because individual scores are entirely the result of a single player's actions, players will likely look to that score first for confirmation of growing mastery. The team score likely won't carry much significance except in comparison with the scores of other teams. Thus, if a player is tracking her team's progress compared with a rival team or a leaderboard, she may check the team score during play, but otherwise may not care until she finishes the game. At that point, likely someone on the team will check how the team performed compared with others and there's a good chance that player will report the team status to the team. Some players may also use the team score and individual scores together to get a general sense of what percentage of the team's progress they themselves have contributed. This method isn't foolproof, as some scoring elements go into the individual scores that are not counted for the team score (like positive ratings from strategies contributed on the message board). Overall, scoring in Labyrinth is an additional, optional way for players to track their progress.
Can we generalize from these design decisions that promote personal visibility of mastery to identify general design heuristics? Scores are good, but are abstract and can be ignored. Something tied to story like freeing pets is better, because the number of freed pets is quantifiably tied to player action in an obvious way. Finally, the game should not just notice but also celebrate player accomplishments. In this case, showing pets being freed (scurrying away) would have been an improvement over simply showing the number of freed pets increased.
Social Visibility in Labyrinth
How does Labyrinth create social visibility of increasing mastery? First, the game is designed to be played with a group of students from the same class in school. So, likely the whole school is aware of the game. Second, teams are aware of members' progress and contributions on the message board. If one player contributes a strategy on the message board that the rest of the team finds helpful, that player's mastery is clearly visible within the team. That player's reputation, visible on the message boards, becomes a lasting reminder to the team of her contributions and her understanding. Also, the whole class playing the game knows what it takes to overcome its challenges. So, when one player progresses significantly, her team will see, and when one team progresses significantly, the class will probably find out.
The mobile and online accessibility aspect of the game again increases social visibility. If people play the game on the bus, in the library, in the classroom, and at lunch, they will also be seen playing the game in all of those situations and by different people. The game becomes a focal point for conversation, and examples of mastery will more frequently receive attention. Groups of people playing together become even more visible socially and more interesting for passersby to come investigate.
Choosing an avatar also creates visibility for a player and her mastery within the team. Like the player's reputation from contributing strategies, the avatar becomes a visual shorthand for identifying the player and her accomplishments. Conversely, if a player contributes unhelpful strategies or doesn't pull her weight, her avatar and reputation identifier become easy tools for her teammates to remember and generalize from those negative details. Of particular concern is any negative judgment attached to a player who asks questions that teammates may consider ignorant or unintelligent. In those instances, we would prefer not to help these team members easily remember and generalize from these perceived negative details. On the other hand, if a player is contributing willfully unhelpful strategies and comments to the message board, a negative reputation is more useful and appropriate.
Leaderboards for individual and team scores also create social visibility for mastery. Players can see their own performance measured against the performance of others. They can say things to each other, if they so choose, like, "I'm number 236 on the high score list. What number are you?" Teams can measure themselves against other teams. Players and teams can always play the game again to try for a higher score, if that interests them sufficiently. All of this activity takes place within a socially visible environment, increasing social relevancy.
Too much social visibility can intimidate players, discouraging self-exploration. The fear of failure only increases with social visibility. Games should provide a range of activities with less or more social visibility, so that players can choose their preferred level of risk based on their moods and personalities. Additionally, any opportunities the game provides for failure should, if possible, have reduced social visibility for that failure. With many players, highlighting a winner generally doesn't embarrass all of the players who didn't win. Labyrinth's leaderboards should publicly only show the top players, while allowing every player to see their true ranking privately. Games like labyrinth can more easily keep failure (or lack of success) private, because social visibility is generally asynchronous. In MMOs with real-time interactions that could result in failure, it can be harder to hide.
What design heuristics can we use to increase social visibility? Create easy ways to compare performances, and make that comparison easily visible and accessible in a public place that most players will visit frequently. In this case, leaderboards for individual and team scores serve this function. Create ways for players to contribute to the greater good and then help players remember those contributions over time. In Labyrinth, writing strategies for teammates on the message board and receiving ratings to add to a reputation serve this function.
