featured image credit: ESO/S. Brunier

Link

Playable link: https://looseleft.itch.io/chronos

CONTENT WARNING: This game deals with themes of psychosis, dissociation, 
schizophrenia, isolation, and suicidal ideation.

Branching Choice Map

My game has… a lot of routes. Each transcript/document has three choices, and for most, all three link to different subsequent transcripts. Every transcript links forward in time—i.e. a transcript from earlier on the timeline is never unlocked by one later in the timeline (see Appendix A for full list of choice links). However, there are still too many possible routes to comprehensively list here. The route tracker I used to check reflections during the drafting process is linked here and will hopefully be useful in grading!

The intention behind having a large number of permutations goes back to the root of the story—everyone perceives history differently. Attention impacts understanding. Whatever you give attention to will unlock additional information that reinforces that bias in some way—if you care about Kira’s family, for example, you might unlock a voice recording from her sister. If you categorize her as psychotic, you’ll unlock an email from her supervisor implying just that. Yet, under it all, the timeline is exactly the same for everyone—nothing deviates from the given record.

It is for this same reason that everyone gets the same ending—Dr. Kira Osei floating off, untethered. The facts of the case don’t change, only your interpretation of them based on which narratives you chose to feed and which you chose to starve. In this way, the choices you make are only impactful insofar as they change your understanding of the narrative, not in how they change the narrative itself.

Overview

History isn’t stagnant—it is shaped by what someone chose to preserve, emphasize, or ignore. Every historical record is shaped by the biases and priorities of whoever filed it. Chronos aims to put the player at the center of that tension: an unreliable narrator, multiple actors with different priorities, an institutional responsibility—all pulling them in different directions.

In Chronos, you play as an archivist reviewing the final recordings of Dr. Kira Osei, a physicist who disappeared from a deep-space research station a few months prior to the start of the game. Following a Freedom of Information Act request by Dr. Osei’s sister, you are tasked with organizing Osei’s transcripts by tagging them appropriately. With every choice, you shape the historical record: tag her observations as “research data” and they are preserved as potentially valid science; tag them as “psychological concern” and you build a case that she lost her mind. There’s no clear, objective truth—only what you construct through what you choose to preserve and how you choose to frame it; an exploration of archival bias, epistemic injustice, and institutional complicity through the act of classification.

Target Audience:

Chronos is for…

• Sci-fi enthusiasts
• Long-form narrative game lovers
• Anyone interested in historiography, archival studies, archeology, or information ethics
• Alternatively, anyone who is ignorant about historical bias
• Thinkers involved in epistemic justice
• Those happy to sit with the ambiguity of incomplete records and unverified truths

Chronos might not be the best fit for…

• Audiences uncomfortable encountering themes of mental illness, isolation, dissociation, and suicidal ideation (see content warning)
• Players looking for a story with clear heroes/villains or action-driven gameplay
• Haters of long-form fiction (there’s a ton of text!)

User perspective has a significant impact on the gameplay of Chronos. As a game designed to function around attention and bias, it is expected that different users will have significantly different interpretations of the events outlined in the transcripts. Working toward that experience was an important goal during the drafting process, and I often had to rewrite passages, reflections, and tag options in order to ensure that I wasn’t endorsing a single interpretation of the narrative.

History

(See Appendix B for full list of playtesters)

Use of Large Language Models in Game Development process

Throughout the drafting, writing, polishing, and design process, I used the support of both ChatGPT and Claude. These models were used primarily to adjust the tone of my writing, especially when taking monologues written as paragraphs and turning them into transcripts (adding in appropriate pauses and repetition of words to simulate someone speaking). They were also helpful for pulling out key points from the transcripts to mention in the reflections and system note summaries. Finally, I used LLMs when building and debugging the UI for the game as a supplement to my initially-limited experience working in Twine.

Version 1:

Playtesters: Butch N.

Playtester demographics: I asked… too few questions before this play through. Does Butch like science fiction? He can answer that as he reads this. Does he enjoy long-form narratives? I’ll leave it to him to fill in the blank! (Whoops, I got better with this as I playtested more!)

Screenshot: An example file from V1 prototype (Google Docs)

The early stages. Version one of Chronos, was a whisper of the final product—far away from even its name at this point. While still focused on the bias of historical preservation, this game had big aspirations in some of the wrong ways. Some context: Originally, I planned to record… all of the transcripts. Using text-to-speech software and some simple audio editing with free background noises (Pixabay, Freesound, etc.), I hoped that the majority of the game would be made up of listening, not reading.

This first draft was written with that end goal in mind. No UI, only written transcripts linked to one another in a Google Doc, each hoping to be peeled off the page and stashed in a .mp4 file—which is how they were labeled in this first run-through. During this playthrough, I was happy to see that Butch was able to gain a solid understanding of the basic world-components, even from a tiny prototype like this. I got good feedback on basic design elements, like switching around the order of choices so that players wouldn’t be able to start clicking through out of habit once they identified patterns between the choices.

Following this playtest, most of my questions still went unanswered. How much time would it take to create each audio file? How quickly does world-building need to happen in the first few files? What can get left for later? The story still felt solidly in its ideation phase.

Moving on from this version, I decided to put off messing around with audio files until I had a more cohesive story. I also officially chose Twine as my engine, and set a goal to begin learning the platform over the following weekend.

Version 2:

Playtesters: Nikhiya S., Angela M., Christina W. 

Version two was very similar to version 1, just with more content. I expanded on the narrative I had written for the original draft, which was set on a space ship full of climate refugees from Earth. The spaceship was a classic utopian-disguised-dystopia where socioeconomic status ruled over all, even on a ship where every passenger had paid exorbitantly to get a seat. My focus moving from version 1 and version 2 was mostly on ideating mechanics that could be used to emphasize the message of historical bias. I settled on using static, where the further away from the “truth” the player strayed, the less comprehensible the recordings would become, and the more clear the “propaganda” messages would get. I also spent time messing around with Twine during this time.

This version was tested with three people: Nikhiya, before class on Monday (10/20), and Angela and Christina in class on Monday (10/20). Here, I will focus on my play test with Nikhiya.

Nikhiya is a coterm student who consumes a significant amount of scientific fiction media, including one of her main shows being Rick and Morty. She does not frequently play online games and is not entirely familiar with interactive fiction. This puts her in my intended demographic in terms of media interest, but not game literacy.

Nikhiya gave a lot of good feedback on the content, including pointing out which transcripts used a tone that stuck out in the context of the game and which components of the world she was still unclear about (social structure, status of people being recorded, reason for recordings, role of playable character, etc.). This was extremely helpful when going back and writing a comprehensive timeline later on; even though I changed the story line, these questions guided by writing in this transition as I realized which components players were searching for.

Version 3:

Playtesters: Julia K.S., Julia B.

Version 3 is when Chronos came to life. Between Monday and Wednesday, I entirely re-ideated and rewrote my story. While some key details remained (the setting in space, the history framing, etc.), the story itself shifted entirely. This was motivated by my own minimal engagement with science fiction literature—while I do enjoy and consume some science fiction, it isn’t the main genre I enjoy. However, I felt as though it fit the premise of my story extremely well, so instead of entirely changing genres, I simply reframed the story into something much more character focused (character building is something I love in creative writing).

Playtester Julia K.S. navigating through the file system UI of V3

The first person to see this new version was Julia Kao-Sowa, a coterm student with experience playing long-form fiction and online games and with an interest in science fiction media—a perfect fit for my game. Julia gave a significant amount of feedback on the UI, commenting on which screens should be draggable, how much text should be visible at one time, how to break up paragraphs, and more. This was incredibly helpful for future iterations of the Twine interface.

Version 4:

Playtesters: Jess C., Leyth T., Krystal L.

Version 4 (completed by 10/29) was the biggest lift. Between version 3 and 4, the amount of written material had quintupled, going from the bare bones of a space sci-fi into a full timeline of 18 files. These playtests were some of the most influential for the final product. Here, I will outline the key takeaways from two, including their audio recordings:

Audio: Play test with Jess C.

The first playtest of version 4 was with Jess C (timestamps of recording provided throughout for reference). She began by commenting on the helpfulness of the system summaries and notes throughout the narrations (00:36). She then commented that she was unclear who the scientist was who was recording the transcripts (3:32), which led to me adding a mandatory character background file at the beginning of the game. Next, Jess noted confusion about the tagging options—she wanted more cohesive tags throughout instead of three unique options for each transcript (4:10, 6:53, and 11:48). She noted that the orange summaries of the impact of each choice were helpful and that they might be moved to before the player selects a tag (7:56). She also noted confusion about whether clicking an option selected it or whether you could go back after reading the impact (4:51). All of these comments were integrated into the revamping that I did of the tagging system, which is now made up of seven tags whose impact is clearly defined and which you select from a drop-down menu.

Finally, Jess noted that there were inconsistencies as to whether the playable character is listening to or reading the transcript (9:54). This was adjusted to be consistent throughout—the character is also reading the transcripts and does not have access to the audio recordings.

Audio: Play test with Leyth T.

The next playtest with Leyth T. offered similar feedback. Leyth does enjoy long-form narratives and is familiar with interactive fiction. They expressed similar confusion over the tagging system (2:57, 16:56, 18:17), noting that they didn’t have a good grasp of how to tag the first few transcripts. They were also confused about who Maya was (5:10). which has now been added to the onboarding file. By the end of the playtest, Leyth noted that they got sucked into the story enough to forget to offer comments throughout, which was a great sign (17:10)!

Picture: Krystal L. playtesting V4

Version 5:

Playtesters: Ryan L., Anthony C.

Version 5 mostly included small fixes—revamping the tagging system to be more cohesive, proofreading all of the transcripts, ensuring that the reflections made sense no matter their order or the route they appeared on, etc. Many of these fixes were caught by two of my final playtesters, Ryan L. and Anthony C. Both playtesters play online narrative games and enjoy long-form narratives. Nothing big changed in this time outside of the tagging system, which they both agreed was now cohesive!

Picture: Anthony C. testing V5

Final:

Playtesters: Josh M.

The final playtest was with Josh M., who frequently plays online games and is familiar with many different styles of interactive fiction. He gave some feedback on the tagging system, including requesting the ability to view descriptions of the tag types throughout, which is now an added feature.

Picture: Josh’s feedback on V5+

Josh was the last person to play through the game in its entirety—Chronos was officially live!

Components of Final Version

Empathy

The final game builds empathy for Dr. Osei not only through her personal narrative—the isolation, loneliness, and ambiguous fate that players bear witness to in her recordings—but also through the representation of her story itself. Players are intended to feel the injustice that comes from reduction: how a complex person gets flattened into bureaucratic categories and how subjective experiences are dismissed as unreliable. Given that the only mechanic players are given to engage with the game is sterile tags trying their best to be objective, the juxtaposition of personal narrative and internal strife with the bureaucratic background hopefully allows empathy to emerge not only from Kira’s personal narrative, but also from a recognition of the violence inherent in how her story is told, preserved, and inevitably distorted by the act of archiving it.

Ending

The ending was left intentionally vague. While I played around with several different ending options, I ultimately decided to leave the players with the final recording of Dr. Osei departing the station alone rather than allowing them to file a final report. This keeps the focus on Dr. Osei’s lived experience and personal history rather than shifting the narrative toward institutional corruption or bureaucratic resolution. History only unfolds one way, and there is always a limited number of artifacts remaining from any historical event. The design of the game—including its ending—was deliberately structured around this understanding: you engage with what exists, interpret it through the lens of your choices, and are left off with a sense of incompleteness.

Reflection

In this project, I took a long time to start genuinely fleshing out my narrative—I spent more time playing around with interesting game mechanics than I did writing, and in the end it resulting in having to drop the game mechanics in order to get the narrative written in time. So, I learned that all first drafts are shitty—so just write one. You’ll be better off the earlier you do. Next time, I will confine myself to a set narrative earlier and give myself longer to play around with world building and mapping character profiles. This part was really fun, it just felt rushed in the end!

Appendix

Appendix A: Choice Routes

0 – Day 28 – Personal Log

• Choice 1: Unlocks: Day 43
• Choice 2: Unlocks: Day 50 (Patel’s email)
• Choice 3: Unlocks: Day 43

1- Day 43 – Personal Log

• Choice 1: Unlocks: Day 64 (first temporal event)
• Choice 2: Unlocks: Day 50 (Patel’s email)
• Choice 3: Unlocks: Day 88 (Maya’s message)

2- Day 50 – Patel’s Email

• Choice 1: Unlocks: Day 64 (first temporal event)
• Choice 2: Unlocks: Day 71 (Rivera discovery)
• Choice 3: Unlocks: Day 64 (first temporal event)

3-Day 64 – First Temporal Event

• Choice 1: Unlocks: Day 81 (conversation with future self)
• Choice 2: Unlocks: Day 71 (Rivera discovery)
• Choice 3: Unlocks: Day 78 (replicating experiments)

4-Day 71 – Rivera Discovery

• Choice 1: Unlocks: Day 98 (Patel’s memo to Medical)
• Choice 2: Unlocks: Day 78 (replicating experiments)
• Choice 3: Unlocks: Day 88 (Maya’s message)

5-Day 78 – Replicating Experiments

• Choice 1: Unlocks: Day 94 (intense experiences)
• Choice 2: Unlocks: Day 103 (encrypted log)
• Choice 3: Unlocks: Day 98 (Patel’s memo to Medical)

6-Day 81 – Conversation with Future Self

• Choice 1: Unlocks: Day 98 (Patel’s memo to Medical)
• Choice 2: Unlocks: Day 94 (intense experiences)
• Choice 3: Unlocks: Day 88 (Maya’s message)

7-Day 88 – Maya’s Message

• Choice 1: Unlocks: Day 103 (encrypted log)
• Choice 2: Unlocks: Day 98 (Patel’s memo to Medical)
• Choice 3: Unlocks: Day 94 (intense experiences)

8-Day 94 – Increasingly Intense Experiences

• Choice 1: Unlocks: Day 109 (video log to archivist)
• Choice 2: Unlocks: Day 103 (encrypted log)
• Choice 3: Unlocks: Day 115 (communication with anomaly)

9-Day 98 – Patel’s Memo to Medical

• Choice 1: Unlocks: Day 109 (video log)
• Choice 2: Unlocks: Day 115 (communication with anomaly)
• Choice 3: Unlocks: Day 103 (encrypted log)

10-Day 103 – Encrypted Personal Log

• Choice 1: Unlocks: Day 115 (communication with anomaly)
• Choice 2: Unlocks: Day 109 (video log to archivist)
• Choice 3: Unlocks: Day 122 (final research log)

11-Day 109 – Video Log to Archivist

• Choice 1: Unlocks: Day 115 (communication with anomaly)
• Choice 2: Unlocks: Day 122 (final research log)
• Choice 3: Unlocks: Day 122 (final research log)

12-Day 115 – Communication with Anomaly

• Choice 1: Unlocks: Day 127 (Maya’s desperate message)
• Choice 2: Unlocks: Day 122 (final research log)
• Choice 3: Unlocks: Day 122 (final research log)

13-Day 122 – Final Research Log

• Choice 1: Unlocks: Day 127 (Maya’s desperate message)
• Choice 2: Unlocks: Day 133 (final log)
• Choice 3: Unlocks: Day 127 (Maya’s desperate message)

14-Day 127 – Maya’s Final Message

• Choice 1: Unlocks: Day 133 (final log)
• Choice 2: Unlocks: Day 133 (final log)
• Choice 3: Unlocks: Day 133 (final log)

15-Day 133 – Final Personal Log (All paths converge here)

• All choices unlock: Day 133 Incident Report (auto-continues)

16-Day 133 – Incident Report

Appendix B: Full List of Playtesters