This is the fourth and final entry in a series of guest posts by Martha Attridge Bufton, member of the Reacting Consortium Board and author of "Do We Take Shelter? Evaluating 'High Stakes' Information."
By Martha Attridge Bufton
Round 4: Playtest in Safe Spaces
Playing a game to test the mechanics, dynamics, and aesthetics is critical to ultimately creating a game that works, for the designer and the players alike. Player feedback is a rich source of insight into the affective experience of playing a game as well as how well the combination of rules and features supports the achievement of learning outcomes. As Mark Carnes (2014) points out, games get better when people play and modify (p. 11).
Game designers Anna Anthropy and Naomi Clark (2014) discuss playtesting in a way that resonates for me: testing a game is “iterating to fun and beyond” (p. 113). They argue that fun is a reasonable goal for testing (and writing a game in the first place), because fun is traditionally the social purpose of games (p. 113). Clark Abt (1970) would agree: some games are designed just for amusement. The term “beyond” captures the idea that games have additional goals, including pedagogical or educational goals. Plus, the idea of iteration permits us to write a game that is not perfect the first time and will improve over time as we play and revise it – axiomatic perhaps, but worth repeating.
In my experience, sharing our “intellectual property” with others can be beneficial and stressful. As a form of peer review, our peers’ evaluations can highlight where a game works and where it needs revision (Proctor, 2014, p. 15). This information sharing is a good thing. However, as much as we want our projects to be successful, receiving constructive criticism well requires a certain amount of foundational confidence as well as an ability to detach a bit.
I’ve learned through my own playtesting experiences that I am more open to the ideas of those folks I trust. I’ve play-tested in three contexts: with my own students; at the Reacting Summer Institute; and with students in several History courses at Carleton. My most comfortable and fruitful experiences are with engaged students and colleagues who have had fun teleporting back to Manchester, 1940, and actively play the game, however imperfect it is.
The inspiration: My Carleton community and the Reacting community of practice
The mentors: Engaged faculty, students, and colleagues
The design task: Receive, filter, and apply feedback
Playtest 1: Wrote It, Played It
I tested my game the first time by flinging myself off a pedagogical cliff by writing and playing it for the first time within four months.
As mentioned in my last post, I wrote this short game to be played in the first-year seminar for arts and social sciences undergraduates that I teach at Carleton University (Ontario). And note that I am defining “short” as a game designed to be played over two to four classes (i.e., six hours). This timeframe is slightly longer than may be recommended for Reacting to the Past Short Games, and there is some flexibility in the time needed to debrief.
The opportunity to teach the first-year seminar came suddenly, the result of a generous research grant from eCampusOntario and the support of the Dean of the Faculty of Arts and Social Sciences at Carleton. With colleagues from two other institutions, a fellow Carleton librarian and I spent approximately a year developing the curriculum, which is grounded in the Association of Research and College Libraries (ACRL) framework for information literacy. This framework is designed around five threshold concepts, one of which is “authority is constructed and contextual.”
When it became clear in early 2023 that we would teach the seminar that fall, I seized the chance to take my nascent ideas and turn them into a game that I would play in that first term. I shared my preliminary materials with Bill Offutt and Nick Proctor, who were kind and generous in their feedback on the initial version of the game. And then I just went ahead and wrote the first draft before playing the game in the sixth and seventh weeks of the course.
Terrifying. What if it bombed? (Sorry, a very bad pun.)
But this first playing of the game was not a flop. Despite being in the first half of the first term of their first year, the students had gelled enough as a group to work together and embrace what was a new experience for most of them.
I was lucky because three of my colleagues – co-teacher David C. Jackson, Pamela Walker, and Linnet Whiston (who voices one of the characters in the game), stepped up to play characters at the last minute when some students did not show up to class. Having a few more confident players seemed to steady the students. And I appreciated the moral support.
I received helpful feedback from the students at the end of the game – they didn’t even wait for the official debriefing. When the air raid siren went off at the end of Round 4, the game ended. Students were outraged! “Do we live or die? You gotta let us know!”
And I learned the importance of an ending or resolution that is dramatic and satisfying to players. They need a win of some kind. With the help of my colleague Ryan Tucci, I developed a fifth-round mechanic that included a random dice throw and answered the “life or death” question. This new mechanic was integrated into the second version of the game.
A critical element of game design revision is to ensure that the mechanics lead to dynamics or interactions that support the achievement of cognitive learning goals. Do we take shelter? is designed to teach students to at least remember, understand, apply, analyze, and evaluate information in depth. Their feedback on the game experience contributes indirectly to co-creating a better game.
Playtest 2: More Seasoned Gamers, More Questions
The next three times I play-tested DWTS, I played with more seasoned gamers. As a result, more informed feedback on mechanics and certain historical aspects of the game came my way.
In the summer of 2024, I was invited to the Reacting Summer Institute (SI) and play-tested the game with ten volunteers. Then, over the course of the 2024-2025 academic year, I played the game in two graduate seminars taught by Carleton colleague Marc Saurette (a medievalist in Carleton’s Department of History) about the history of games. These playtesting experiences were also fruitful and notably different from the initial testing in my first-year seminar. Playing with first-year students who were not gamers and not history majors yielded useful mechanical feedback. Playing with historians who were also seriously into games yielded feedback that was detailed, creative, and often focused on enhancing the historical aspects of the game.
For example, Marc did an extensive search in the historical newspaper databases licensed by the Carleton University Library and seemed to find twice as many potentially relevant newspaper articles as I had. I do feel the allure of the historical aspects of the game, and more research has yielded some sources that could make the game more interesting. For example, now I have a copy of a small leaflet that was distributed to households in 1940 that detailed how to winterize a home air raid shelter. However, I wonder if more history will enable students to understand the core concept and tasks more clearly.
These sessions at the SI and with Carleton historians (both emerging and seasoned) also pinpointed mechanical issues that have been important to address. In particular, I have had to revisit elements that seem good (e.g., an interactive map) and yet are not used as purposefully as they could be in gameplay. Plus, I still do not have one major mechanic, such as a vote, which would further engage students by adding a goal over which players could have more agency and experience some rivalry. Currently, the explicit goal of Do we take shelter? is to decide where to take shelter for a night. The implicit goal is to stay alive, and the latter has limits to player power – the bombs fall where they fall. Players need to come to consensus on where to take shelter (no one gets left behind), but is consensus a satisfying-enough goal, especially because it offers limited opportunities for rivalry (or subversion)?
I’ve been reading Joe Bisz’s and Victoria L. Mondelli’s The educator’s guide to designing games and creative active-learning exercise. Joe and Victoria have written a guide to game design that is accessible, in part because they clearly define and explain core concepts in simple, non-jargony ways. I appreciate the explanation of “give simple game mechanics,” with the contrast between mechanics to determine pace (random, rapid) and mechanics to achieve goals (rival, reward, role) (Bisz & Mondelli, 2023, p. 11). These distinctions are bringing some clarity to my thinking about rivals, rewards, and votes.
Equally important to the redesign process has been the process of reflecting on the aesthetic or affective element of the game experience: are students having fun and learning? Participants seem to have fun while playing Do We Take Shelter? However, participants consistently ask for more opportunities for competition rather than collaboration and a bit of subversion. A new hairdresser (potential spy) role and a revised household vote might meet these needs for more fun.
Playtest 3: Keep Focused on the Goal
Ultimately, right now I remain focused on the nature of my game and the ultimate teaching and learning goals as I playtest and revise. By nature, Do We Take Shelter? is a cooperative rather than a competitive game. As a result, mechanics that support rivalry and competition may not work optimally in a game that relies upon players reaching consensus within individual small groups. However, a solution might be a vote within each household – it is possible (and plausible) that family members may not agree initially on a solution to this problem.
This summer, I have a research grant to work with an undergraduate student to revise the central mechanic (a vote based on points?), play test the revised game with volunteers, and then interview a sample of players for feedback. The revisions will also include a new player (maybe a potential spy who spreads propaganda and disinformation).
My goal (and hope) is that the points/voting mechanic that the student and I develop will support the “evaluating information goals” by acting as a form of assessment for student learning and achieve some mechanical goals, such as making the game more “alluring” because there is at least a stronger reward for doing the work of the game.
Three design takeaways
Clarity with regard to learning outcomes is essential during game development. While Do We Take Shelter? is grounded in an historical event, the main purpose is to allow students to apply the concept that “authority is constructed and contextual” to a genuine problem. As an historian, I enjoy discovering interesting, quirky, or unexpected facts that enhance the complexity and interest of a story. However, the more historical detail that is added, the more players could be distracted or confused by information that is not central to the desired pedagogical outcomes. Clear learning outcomes allow a game designer to say “Yes” to some suggestions and “No” to others.
Play test as often as possible (and take good notes). I’ve been fortunate that I am teaching a course in which I can play test (which means I have some control over the frequency of playtesting). Plus, I am fortunate that I have a Carleton colleague (Marc Saurette) who has been interested and supportive of my game and the opportunity to play the game at the RTTP Summer Institute. As a result, I’ve had seven opportunities to watch and learn as diverse groups of players have had fun (and struggles) playing the game. Each time, I learn something valuable even though I have also felt overwhelmed at times and unsure how to use the feedback I’ve received. In addition, taking good notes during debriefing sessions has helped me track players’ experiences and reflect on potential changes over time.
A game does not have to be “finished” to be playable, so relax. This thought may seem axiomatic to seasoned game designers and yet is helpful to emerging designers who may feel disappointed that a game does not work perfectly the first time. Each time I run my game, I hope that I will finally solve key mechanical problems, so the game development process is finished, and I still feel anxiety that I haven’t arrived at the best working version of the game. However, I am more comfortable now with the imperfections, which frees me up to keep iterating towards fun and beyond and to relax more into this process.
