Game
My game can be accessed here. I would recommend playing the game prior to reading the write up.
You can access the original version of my game here and its write-up here.
Overview
As someone studying computer science who hopes to use technology to improve the world, I’m acutely aware of the potential of technological advancement to harm specific people and communities, even when its original purpose was to do good. Thus, Erosion was born from my desire to explore tech ethics, a history of exploiting the bodies of women of color in the name of scientific progress, and complicity.
Players take on the role of June Weng, a 24-year-old software engineer at Cerebra, a fast-growing startup that enables people to re-experience their memories in VR. June joined Cerebra right out of college two years ago and is the prodigy of its founder, Robert Fuentes. Cerebra’s goal is to help people: those grieving, those suffering from dementia, and those needing to process past experiences. However, one night, June meets Elena, a Department of Defense (DoD) analyst who claims to have troubling information about Cerebra. As players go through the game, they’ll learn more and more about this wrongdoing, before it is ultimately revealed that the DoD aims to use Cerebra’s technology for mind control and is testing prototypes on teenage girls of color. June is pulled between a desire to do better for the world, not hurt it, and the job she loves.
Erosion‘s objective is to explore the unintended consequences of technological advancement through establishing an emotional connection with players. First, it develops players’ understanding of June through flashbacks, internal reflection, and seeing her navigate the world: she is not just a role they play but a person with hopes and fears. Erosion also puts players in June’s shoes for both inconsequential choices (what she’s wearing to a party) and consequential ones (if she should work with Elena or not), building up their empathy for June as a person and with the decision she needs to make. Erosion also employs detailed testing logs to force players to see firsthand the physical toll that the testing takes on subjects—in other words, how advanced technology could affect real people. Through these narrative choices, the game invites players to care deeply about the ethical issues it explores.
In the second iteration of Erosion, I shifted my focus to visual design and expanded functionality in response to feedback on the first iteration. I had originally used ChoiceScript, which did all the UI/UX design work for me, but as a result, did not allow for much customization. For the second iteration, I switched to Twine, designed my own UI, and added new functionality (e.g., saving and loading games, logging choices). I also made narrative changes, including adding more choices and fleshing out the endings. Each ending gives the player a satisfying close while keeping the door open for future story (see Figure 1 for the game’s key choices and possible outcomes; note that while there are five main outcomes, depending on the paths taken, the actual narrative, internal reflection, and circumstances differ).
Figure 1: Branching Map
Testing and Version History
The game’s target audience is people who enjoy long-form fiction, especially sci-fi or dystopian works. It also helps (but is not required) to have an interest in tech ethics: I found that both players who were interested in it and who weren’t enjoyed the game. All playtesters fit the target audience, more or less.
Version 1
For my first version, I made a number of changes in response to P2 feedback (from peer reviewers and Butch). The rest of the feedback I addressed throughout the later versions.
I transferred all content to Twine to enable custom UI. This enabled me to select colors and fonts that aligned with game themes and, as suggested in feedback, create starting and ending screens to separate them out from the rest of the content. I also added functionality for users to load and save progress and navigate forwards and backwards. Content-wise, I proofread the story, delineated content by day instead of chapter (which had caused confusion with the first iteration), and added a content warning.
I playtested Version 1 twice.
My first playtester, Angela, is a CS coterm who enjoys long-form fiction, and although she hasn’t taken tech ethics classes specifically, she has interest in and familiarity with the topic from studying CS. She said that the game made her think about tech ethics, specifically when thinking about her own future. She explained that a lot of CS students, like her, may end up at tech companies that may or may not be evil, and as an employee, they would actively have to think about things like this. I was glad that the game evoked her preexisting relationship with technology and plans for the future, in a way that made the game applicable to her life.
She empathized both with the girls who were being tested on as well as with June and felt that the design lined up with the theme of the game, indicating that the game and my design revisions successfully achieved their outcomes.
While she felt her ending made sense, she suggested adding a sentence hinting at why the user got the ending that they did. I realized that this could tie in well with something I already do and that Angela really liked which is ask users if they’d like to go back and change a specific choice they made (since the implication is that that choice strongly influenced their ending), so I made a note to make this change when I got around to fleshing out endings.
She also felt that in June’s initial interactions with Elena, Elena is not actually that mean and June overreacts. I realized that this disconnect between June’s actions and what Angela expected could harm the building of empathy between player and main character, so I added to this of content changes I would make.
Evelyn, my second playtester, is a CS coterm whose favorite genre is dystopian works that engage with future ethical dilemmas, and she likes books. She’s also pretty familiar with tech ethics from her CS classes.
Right off the bat, she mentioned liking the design of the game (0:00). This was very exciting for me to hear, particularly because I know she has a good eye for design from seeing her work. It also indicated that my second iteration was stronger than my first already since she mentioned enjoying the design even unprompted, while design was never a selling point of the first iteration. In particular, Evelyn felt that the minimalistic design aligned with the game theme, and she liked the action buttons (in the new design, they’re integrated into the text instead of just being a “next” button) (36:36).
The narrative also resonated with her: she found the erosion concept clever and unique (35:10) and didn’t feel that it or its details were predictable (30:38), despite being someone familiar with the genre. She thought that the DoD would be torturing people with bad memories, but she enjoyed that it was unexpected: the game evoked her preexisting notions of the DoD, technology, and dystopia but subverted them. When we were discussing the game after, she put it in conversation with other works she enjoyed like Black Mirror episodes (33:25), and it was very interesting to see the connections she made. She liked that the game made her think about the unintended consequences of something that can be good, and she empathized with the girls being tested on, finding their situation realistic and worrisome (35:29).
She suggested making a special format for the logs, enabling the user to read through them with June (enacting narratives), instead of just being told what the logs said. I held off on implementing this until Version 3 so that I could introduce key design changes more gradually, drawing on recommendations in “Playtesting Formally” to incrementally test variations.
Both Angela and Evelyn were also confused by the dates, so I switched to only using them for flashbacks (e.g. “x time ago”). I had originally added them for every section to split up the text (same function as chapters), but Evelyn said that she didn’t find that necessary. Evelyn also suggested changing the color of the interface during flashbacks, and I really liked the idea. It gave me the opportunity to take advantage of functionality that works seamlessly in a game but not in other narrative mediums (like writing), inspired by Henry Jenkins’s explanation that games that tell stories don’t do so in the same way as other media.
For the next version, I also started working on a log so that players could see the story that they were constructing (enacting narratives). When they went back to old choices, they could see what they selected, and they could also use the log to go back to old choices (something Evelyn wanted to do). Angela also mentioned that because the game was longer, it was harder to click back with the navigation button to old choices, another motivator for adding the log!
I also implemented autosave, added a restart dialog (asks players if they’re sure they’d like to restart the game), made menu buttons properly line up with text, and made buttons show up properly in dark mode.
Version 2
Figure 2: Abi playing Erosion
Abi, my third playtester, is a CS coterm who likes reading books and is interested in tech ethics. She enjoyed the premise and found that the game presented a cool topic and interesting, realistic ethical dilemmas. She found herself worried about if this could be real and if this were truly happening at a few points during the game, showing a strong interaction between the game and her preexisting beliefs/worries relating to tech ethics.
She felt that the characters were real, and she wanted to get to know Elena and socialize in the party scene. She said she had a vested interest in characters and felt what the main character felt, especially when making choices (enacting narratives to build empathy). I also saw her reflect a lot on the choices she made as if she was the character itself (even inconsequential ones like going home at the end of the party).
While she felt that her ending was accurate, she really wanted to see what happened next, indicating that I did a good job keeping the door open to future narrative and engaging players. I fleshed that out in the next version, ensuring that the game clearly felt like a few chapters since endings were left open-ended, and also added Angela’s suggestion for a why lost/won.
She also felt that the design lined up well with the game content, and she didn’t find the content hard to read even though it was full pages.
She wanted more choices, so I added in some more choices in conversations with Elena (at the party and in the cafe). Responding to peer review feedback, I also added more content before June decides to hear Elena out or not.
She also suggested clearing log entries that no longer applied when players went back to make a new choice. I didn’t have time to implement this for Version 3 due to playtester availability (it needed to be done before finals week), but I kept it in mind for my final version.
For Version 3, I also made language surrounding telepathy more consistent (e.g. didn’t “read” messages, she processed them; used mind-related language), which partially addressed P2 confusion over rules of telepathy. This would make it easier for users to immerse themselves in the game (enacting narratives) if there weren’t inconsistencies.
I also implemented Angela’s feedback to more gradually escalate June’s reaction to meeting Elena and Evelyn’s suggestion to distinguish the text for the testing logs.
Version 3
Ellie, my fourth playtester, is a History coterm—reads a lot!—who is familiar with IFs and enjoys them. She does not have a specific interest in tech ethics but still enjoyed the game. She found that it made her think a lot about tech ethics, and she also empathized with many characters (47:16). She empathized with June and was trying to help her get the best ending. She felt empathy towards the girls being discussed in the logs and also liked how the logs were laid out. She appreciated the inclusion of Robert’s backstory as it helped her sympathize with him more.
She felt that her ending made sense, and she also really loved the explicit option to go back and switch a choice (46:12). She also enjoyed the design, especially the dark mode during flashbacks (45:45), and she liked the amount of choices in the game (45:30), which is something she had asked more for when she played the first iteration.
Given the positive assessment of story, choices, design, and functionality, I felt confident that Erosion was in a much better place than when I started.
For the final version, I made a few more touch-ups: I made the game responsive to different screen sizes, fixed a bug where opening the Choices Log would cause the whole menu to shift to the right, edited story content for clarity and consistency, and cleared log entries that no longer applied (as suggested by Abi + without this change, when Ellie was going back to old choices in the log, the log was very messy). The last change especially aided in the log’s goal to enable enacting narratives because it provided a coherent story that the user had created, instead of just being an audit log.
Future Updates
I loved iterating on Erosion! In the future, I would like to further customize the design (e.g., create an animation when June sends and receives messages with telepathy), strengthen June’s characterization (e.g., resolve confusion surrounding her social anxiety), clarify and expand on June’s relationship with Elena (e.g., are they friends? why do they always meet in person when they could simply telepath?), and add more nuance to June’s relationship with Robert (e.g., show more interactions between the two of them that convey gradations of their relationship).
