In our penultimate assignment for game design, Cole Gleason and I worked on an interactive fiction (aka text adventure) game. As a part of the assignment we were encouraged to reflect and analyze our playtest data and the experience as a whole. While many of my reflections were specific to the interactive fiction space, I think my biggest and most generalizable takeaways were about how my mental model and the players’ mental models differed and various strategies for aligning them.
If you are unfamiliar with interactive fiction, the Wikipedia definition says, “software simulating environments in which players use text commands to control characters and influence the environment.” So players can typically use commands such as “examine ___” or “go west”. Cole and I are both avid fans of interactive fictions and were really excited to write one together. We put in a LOT of time into brainstorming and writing the story and narrative beats as well as a ton of time learning the Inform 7 system which supports writing interactive fiction. We were really excited and nervous to start playtesting our game, which had had many hours of work poured into it. Our first several playtests were a disaster.
Disaster might be a little strong, but Cole and I had put so much time and effort and care into crafting this world and repeatedly saw players unable to comprehend what they could do and what the game wanted them to do. In the excerpt above you can the see the player is trying to meaningfully interact with the world and learn more information (an intro had previously brought them to this room where their mentor is unconscious and they are encouraged to solve the mystery). However most of their attempts are completely misunderstood by the interpreter. The curtain was pulled back way too quickly and easily, breaking the magic circle of the game.
The beauty of interactive fiction is that we as designers have to do very little to figure out what the player’s mental model is. There’s not a ton of interpretation necessary on our part, just by reading the above it’s clear that this player expected to be able to use all of her senses to examine suspicious objects (e.g., smell, touch, and in other transcripts she repeatedly tries to taste things). In another instance, we had wanted players to find and tattoo themself with a healing spell. We had consistently called it a ‘healing spell’ and ‘healing tattoo’ in the game, but in all of our discussions we had always called in the ‘heal’ tattoo. So in the game the command is ‘tattoo heal’ but every player we encountered tried ‘tattoo healing’ and were frustrated and upset when it didn’t work.
Watching these playtests is tough as the designer when the player struggles and it’s difficult not to jump in with the answer. It’s weirdly some of the joy of an interactive fiction to struggle and try commands until something big and cool and exciting happens that makes you feel clever. But in situations like our early playtests, it’s very clear that our mental model of what is obvious and clear in the system was not true to the player experience.
I love playtesting. Lots of people much smarter than me have written excellent things about playtesting (e.g., Shawn Patton on VR Playtesting, Judy Choi on Playtesting with a Purpose). It was a gift in these playtests to not have to dig in so deeply to uncover what our players thought and felt, but some of the questions I like to ask when their thoughts are much more obfuscated are “what did you think your goals were in the game?”, “what did you think you had to do or want to do with item x?”, and “if you could change anything about this game, what would it be?” Think aloud protocols are another great way to get inside the mind of the player and figure out how they’re reacting to each new piece of information (whether it be characters, dialogue, art, etc.) in the game.
It’s deeply important to confront that what is in your brain is rarely in the brains of your players and that it’s rarely their fault. If I had an infinite amount of time and budget, something I’d love to do in the future is ask my novice players to play and existing and similar game in front of me before playtesting my own to get a better sense of how novice players interact with similar systems.