P2: It’s just a burning memory – Tianze Shao

Read a patient’s memory on a retro terminal …

Returning to the last day of summer, you will experience a story about sunflowers and fireworks, betrayal and salvation, and MEMORY.

Play the game here (password: cs377gp2)

Note: Autosaves to browser local storage, but re-entering after closing the page may cause unknown bugs. A single playthrough is recommended. (Clearing browser cache can reset progress.)

If you get stuck, you can check the guide here.

After you play it …

 

—————————————————- Spoiler Alert ————————————————–

 

 

Choice map

Figure 1 : Choice Map of It’s just a burning memory (Save to view full size)

Colour: Consistent with the in-game lock status (Blue: unlocked directly, Yellow: unlocked by Q&A, Red: unlocked by collection count)

Name: Consistent with in-game memory/file names

Structure: Single ending, no branches (Options in Memory: Gregor do not affect the ending)

Idea Formation

As a psychology student, I believe “memory” is an excellent theme for a game. Therefore, the starting point of this project was: let the player explore a character’s memory and use game mechanics to show that “memory” is subjective, fragile, and easily manipulated—but it is also these fragile neural activities that make up who we are.

In the 4A class, we used the Snowflake Method to write stories. At that time, I didn’t have a specific plot in mind, but rather a trick: show some memory fragments from the same date, but which actually come from different years. When the player isn’t aware of this, they will think these memories are from the same day, discover many contradictions, and experience suspense. But this seemed more like a puzzle game than an empathy machine. So, using my professional knowledge in psychology, I thought of the symptoms of Alzheimer’s patients, which are very similar to my description—losing the ability to reality-check, trapped on the same day.

Based on this, I continued to research AD: a movie called The Father, where the protagonist sees his relatives as strangers; some videos of AD patients, where I saw a 90-year-old veteran who said he was still 19 and charging on the battlefield. I thought of my grandfather, who always mistook me for my father—which made me feel sad and helpless. Combining this material, I wanted to use procedural rhetoric to let players personally experience the situation of an Alzheimer’s patient, thereby forming empathy for them.

While researching, I remembered a piece of music I had heard before, called “It’s just a burning memory”. This music is from the composer The Caretaker’s Everywhere At The End of Time (a 6-hour musical piece about AD, simulating the stages of the disease’s progression. An absolute work of art!). This music samples Al Bowlly’s “Heartache”. “Popular music” from nearly 100 years ago has now become “nostalgic”. The movie The Shining also cited Al Bowlly’s songs in the ballroom scene, presenting an elegant, ethereal, and slightly eerie atmosphere. “It’s just a burning memory” is a lyric from “Heartache”. This melody looped in my head: “burning memory” is an excellent metaphor for AD patients. The shining days of youth are like a flame burning, and in old age, these memories are burning until they are hollow. This lyric became my game’s title and also established the aesthetic tone: retro, nostalgic, and slightly weird. Therefore, I used the aforementioned music as BGM and also paid homage in the game to literary works I love about memory: In Search of Lost Time (Proust) and The Sound and the Fury (William Faulkner).

Figure 2 : The ballroom scene of The Shining

History version

In the making of this game, the story writing and program development were relatively independent. Therefore, I will explain them separately.

Timeline

Story

10/15 (paper prototype) – 10/18 (complete outline) – 10/20 (detailed outline) – 10/24 (game script) – 10/26 (final script)

Code

10/17 (tech selection) – 10/18 (framework) – 10/26 (features) – 10/28 (implementation) – 10/31 (final version)

Story

10/15 Paper Prototype

This was a very challenging part! From 4A to 4B was less than 48 hours, and I still had other classes and homework. Moreover, my story includes a trick: this means I couldn’t just present a fragment, but had to have a complete version for it to work in the playtest.

Therefore, I used a story I conceived this March: the female protagonist is also named Camellia, also in a totalitarian country, and the male protagonist escapes with her through a sunflower field. But the previous story was very large, involving multiple factions and a long timeline, which is why I hadn’t been able to make it into a game. This time, in a short period, I used Occam’s Razor to trim the original story prototype, limiting it to the scope discussed by Jamie Antonisse in “Paper Tales: A Guide to Narrative Prototyping”: I limited the story to “one day” and simplified the dramatic conflict (but increased its intensity). For example: at first, let the player think the protagonist is capturing Camellia, but later discover he wants to protect her; Billy is the classic “villain who stops the protagonist,” capturing Camellia before the protagonist, causing intense conflict, but without a very complex motive.

I used a spreadsheet to manage the story, and at this time, the prototype for the clue-searching gameplay was already taking shape. In the end, I almost directly used the outline as the paper prototype without polishing it. Despite this, according to player feedback, this version of the story was concise and clear, with no comprehension barriers.

Figure 3: Story outline spreadsheet

Figure 4: Paper prototype

10/18 Complete Outline

According to the paper prototype test feedback, players were a bit confused about the story’s “background”: the setting of the totalitarian country, the General, how they oppressed the people and soldiers, and why the protagonist wanted to oppose them.

This kind of problem reflects a disconnect between the story and the game narrative: I had a core trick design about AD, mainly shown through the game system. At the same time, I was telling a story about pursuit and escape, mainly shown through text. But in reality, these two were independent: even if I told another story, it could still carry the AD design. To fix this, I needed to tightly connect the story with AD.

• Original background setting: The General banned neuroscience and promoted phrenology.
• After iteration: The General researches memory editing technology—using memory deletion to create soldiers without humanity, which is why Camellia defected and sparked a revolution; using memory implantation to create a false memory for Gregor (of killing Camellia himself), thus worsening his AD symptoms.

This change merged the “textual narrative” with “procedural rhetoric,” making the game’s narrative self-consistent and compact.

10/20 Detailed Outline

Based on testing and discussion with my screenwriter friend, I finalized the complete outline. This complete outline includes three layers of narrative:

• First layer: Gregor as a 25-year-old military policeman on August 31, 1945, is on a mission to capture Camellia, but actually helps her escape. He is finally captured and implanted with the memory of “killing Camellia”.
• Second layer: Gregor as a 75-year-old man with AD, doesn’t recognize his wife Camellia or his son. He goes out every day searching for the Camellia of the past. The implanted memory further deepens his condition, causing Gregor to accidentally fall into the water and fall into a coma.
• Third layer: Camellia, as a neuroscientist, has spent her life researching memory technology. She tries to use an instrument she designed to explore Gregor’s memory, deleting the implanted memory to alleviate his condition. In the end, Camellia decides to delete herself from Gregor’s memory, but Gregor refuses to forget Camellia and awakens from the coma.

The player’s experience is like “peeling an onion”: at first, they experience the first narrative layer, personally experiencing the tense and chaotic day of Gregor the military policeman. Then they experience the second narrative layer, realizing Gregor is an AD patient and developing understanding for him. Finally, they experience the third narrative layer, where the player realizes they are playing as Camellia, re-examining the story from the perspective of an AD patient’s family member, and personally making the painful choice (to make Gregor forget Camellia), elevating from understanding to empathy.

The full detailed outline is here.

10/24 Game Script

Going from outline to script took more time than I expected. First, I split the outline, as shown in the figure. Compared to the paper prototype, this split involved a longer flow, adding a 15-20 minute sequence after all memories are unlocked (the second and third narrative layers).

Figure 5: The flowchart I used (Chinese version, similar structure to the choice map, but nodes are plot points instead of numbers)

Additionally, I designed a dual “memory” and “file” system. The former is subjective, the latter is objective. The former required vivid and immersive text. I adopted Faulkner’s technique from The Sound and the Fury (Benjy’s perspective): short sentences, frequent line breaks, focusing on psychological reality rather than physical reality, and even omitting punctuation at moments of high emotion (23-5pm-death), with a certain “stream of consciousness” style.

As for the latter, the file system includes various forms of information: newspapers, embassy telegrams, the son’s diary, technical documents (Camellia’s notes to herself). The game’s guidance system is also based on files. Writing different forms of files required different styles, which took me a lot of time.

10/26 Final Script

My screenwriter friend and I read the full script together. Based on her suggestions, I revised some wording and added the memory fragments in the epilogue: from 1925 to 1995, from childhood sweethearts to Darby and Joan, fragments of Gregor and Camellia’s life. She hoped Gregor and Camellia could have a good ending, or at least let the player feel acceptance and be moved at the end, rather than continuously oppressed. I adopted this suggestion—from an empathy perspective, I want players to understand AD patients. When they encounter someone with AD, they can adopt an accepting and tolerant attitude, rather than seeing them as an outcast or simply viewing AD as an “enemy”. Therefore, in the game’s epilogue, I had Gregor say: Love cannot stop aging, but it can give us the courage to face it.

Code

My Technical Background

• Not a CS major. Studied Python for half a year in undergrad, understand basic data structures and algorithms.
• Can use Godot (basic) and Unreal’s Blueprint system.
• Familiar with HTML, CSS, Javascript, but cannot write independently.
• 4 months of vibe coding experience, did full-stack development for a forum project.

10/17 (Tech Selection)

I used vibe coding approaches for this project with Claude Code. In this part, I will include some of my insights on using AI for non-CS game developers to reference.

Firstly, write Claude.md. I copied the Claude.md from a previous project (including documentation standards, git upload standards, etc.) and referenced this game’s design doc as the “constitution”. (See drive for specific content)

Vibe coding tips: Let the AI ask you questions! Whether the AI can do a job that satisfies you depends on how much context you provide it. However, your first draft might not cover all the information in your head. Therefore, when I provide the design doc (or other important info) to the AI, I first let the AI ask me many detailed questions, and I answer them one by one. After a few rounds of this, the information in my head is fully extracted, and the AI’s implementation is more in line with my expectations.

My initial plan was to reference the game Type help, which is also a terminal-like search system. Its implementation is based on Twine’s Sugarcube v2. I obtained the source code for Type help—just one HTML file, very elegant. I had the AI analyze its structure and generate a detailed technical assessment document.

Afterward, I provided this game’s design doc to the AI, needing to implement search, unlock, and other functions. I asked the AI if it could implement these technologies by referencing Type help. The AI said it could and generated a detailed technology migration document. Therefore, I decided on Sugarcube v2 as the tech choice, installed the dependencies, and asked the AI about its knowledge of this engine. The AI told me its training cutoff version of Sugarcube. After comparing it with the update log on GitHub, I found no obvious method changes in the current version, so I used the latest version of Sugarcube.

Vibe coding tips: When using engines, libraries, or other external tools, always ask the AI about its knowledge of them. A common problem is that the AI’s training data is old, while the external tool has been updated, causing the AI to be unable to use the relevant methods correctly. It’s also possible the external tool is niche, and the AI lacks training on it. Therefore, when making tech selections for vibe coding, be sure to consider this issue and prepare a solution for when the “AI’s knowledge version is too old,” such as having it read the latest documentation, preparing a version comparison document, etc.

But soon, the sad thing happened: I found that the Sugarcube engine hijacks the DOM, making it impossible to implement the UI structure I designed. Also, when configuring the environment, I felt something was off: many of Sugarcube’s features were not what I needed. Compared to the classic IF it supports, my game didn’t involve branching choices but had special functions like search and unlock. This framework was more of a limitation than a help, not to mention the risk of the AI writing incorrect methods. I again asked the AI about the pros and cons of this engine for my expected functions. This time, Claude listed a whole bunch of “cons”. In the end, I decided to abandon Sugarcube and just use native JS.

Vibe coding tips: The AI is unlikely to refute your ideas. When I decided to use Sugarcube, the AI would only argue for its benefits based on that premise. When I decided not to use it, the AI would list a bunch of drawbacks. Therefore, human judgment is very important here. Perhaps a better practice is: conduct a more objective evaluation during tech selection, and have the AI analyze the implementation difficulty of each function in the design doc using Sugarcube vs. native JS, point by point.

Figure 6: Sugarcube’s UI

10/18 (Framework)

After returning to the native JS tech selection, I used the AI’s image reading function to quickly implement the UI’s HTML framework. The result was pretty good.

Figure 7: UI Design Sketch (In Figma) 

Figure 8: UI of the first prototype

Subsequently, I used the Opus model for planning: breaking down the design doc into various features and giving a rough development plan. I adopted an agile development strategy: implement a runnable prototype as quickly as possible, and then gradually add features. Soon, an interactive interface was built, with features currently implemented including: search system, unlock system, list switching, and text rendering.

10/26 (Features)

In the following week, I didn’t have large blocks of time for development, so I used fragmented time to develop various features, including:

• Text import
• New content notification
• Sort command
• Garbled text
• Visual effects
• Animation
• Music

During this stage, the AI generated a large number of test HTML files, as shown. Additionally, I required the AI to generate clear technical documentation, recording the new methods it created for future reference.

Figure 9: Test interface during feature implementation – Text Decay

Figure 10: Test interface during feature implementation – Visual Effects

In the actual development process, I was quite casual and didn’t completely follow the initial plan. I thought this was a small project, so my attitude was “as long as it works,” without focusing on the specific implementation. But this sloppy approach created technical debt (multiple parallel implementation methods), all of which would come back to torture me this weekend.

Vibe coding tips: To avoid technical debt, I have two common methods. I’ll introduce one first—let the AI fully understand the code structure before it starts writing code. Every time I clear the context, I have the AI read Claude.md, then read the work log from just before, then read the existing implementation in the workspace. I add a prompt in the first round of conversation, “Please do not write code yet,” and “ultrathink” (to penetrate multiple code layers). Only when I confirm the AI’s understanding of the workspace is correct (at least consistent with my own cognition) will I let it write code. Otherwise, the AI might hastily skim the background info, start working too early, and ignore some already implemented methods, leading to multiple parallel implementation methods in the system, which is terrifying.

It’s worth mentioning that I referenced screenshots from other games to create some animations and introduced external libraries for special effects. These are all in the AI’s comfort zone and weren’t hard to implement.

Figure 11: Deletion animation reference (Honkai: Star Rail)

Figure 12: Deletion animation implementation (It’s just a burning memory)

10/28 (Implementation)

After finishing the scriptwriting, I implemented the content into the game. I had originally developed an HTML tool for batch importing content. This tool required formatted text, but the text I wrote wasn’t fully formatted and also included some natural language commands (e.g., “Unlock by question, question is xx, answer is xx, hint is xx”). Therefore, I tested using the AI for splitting, formatting, and importing. I first tested with a small sample to establish import standards and common issues. Then I imported the large sample—the results showed the AI did great! It not only completed the mechanical work of formatting but also understood my natural language commands, automatically setting different unlock conditions. (Docs in the drive)

After that, I still had to call the previously developed features (based on the tech docs) to implement the game’s logic triggers and flow. This process was full of challenges: some features couldn’t be combined well. It was like building a bridge from both ends, only to find they didn’t connect. Thus, I fell into a long debugging session. To avoid technical debt during debugging, I used git’s revert function many times.

Vibe coding tips: Another method to avoid technical debt—use revert more often. When debugging, every failed attempt introduces some changes. After multiple rounds of debugging, even if you succeed, you’ve introduced a lot of redundant content, which will harm you someday. Therefore, when I successfully debug, I always have the AI record the core changes, then revert and apply the fix all at once.

Despite this, there was good news: HTML has many debugging tools, like the console. Adding a lot of debug text in the code and pasting the output to the AI is a good debugging technique because it lets it see “what happened”. In addition, I also used the console to directly call certain files or trigger certain events, like playDeletionAnimation1(), avoiding testing from the beginning.

Therefore, for non-CS students / non-game engine experts, I think using native front-end technologies to develop a game is a choice that balances difficulty and freedom: it’s more free than engines like Twine and can implement many custom features; and it’s not too hard, because the AI’s knowledge of it is very rich, and the testing and debugging tools are also plentiful.

10/31 (Final Version)

After a few rounds of testing, based on player feedback, I made the following optimizations:

• I observed that players would rather repeatedly look through “memory” than open the “clue” or “file” tab list, possibly because it was too inconspicuous. So, I moved it from the bottom left to the top left.
• I observed that players didn’t notice the new content notification (small blue dot), so I changed it to flash: based on psychological principles, dynamic objects are more likely to attract human attention.
• Highlighted the title of the content currently being read.
• Removed the backdoor I left during testing. (One player accidentally triggered it and saw the ending right at the start of the game, so embarrassing.)
• Fixed a persistent bug reported by players (some Q&A unlocked items required three consecutive clicks after unlocking—this bug took me a whole afternoon to fix, finally found the root cause was hardcoding).
• Revised the wording of some questions (e.g., the repeatedly mentioned “Type A B” question).

Game test

Note: I conducted three game tests in class—but unfortunately, I forgot to take photos of my classmates! The tests outside of class were all online, so there are no photos either. To make up for this, I have included audio recordings, questionnaires, and videos.

First Round: Paper Prototype (10/15)

Details here.

Participant: Ngoc Tran (student of CS377G)

Main Suggestions:

• Add some pictures to present the story more visually.
• Provide more detailed background information. The “big picture” such as the revolution, the protagonist’s and Camellia’s positions, etc. is not clear enough.
• If designing questions, pay attention to the convergence of information. (For example, the 40-year-old man that appears somewhere; my setting is the protagonist’s son, but the player might think he is just a passer-by.)

Figure 13: Desktop prototype (player sorted manually)

Second Round: Video game Prototype & Storyline (10/18)

Details here. 

Participant: Joyce Zhang (screenwriter, and my girlfriend)

Main Suggestions:

• Regarding UI: Don’t have two flashing cursors at the same time.
• Regarding interaction experience: Try to unify keyboard and mouse input as much as possible.
• Regarding Story: Enhance the emotional impact, strengthen the empathy segment.

Third Round: Paper Prototype (10/22)

Participant: Marielle Zheng (student of CS377G)

Observations:

• Without guidance, deduction is very difficult. (From the early story fragments to the truth of AD.)
• When the mystery was revealed to be AD, I felt her experience was not “Aha!” (surprise), but “Okay, so that’s it” (expected). This might be because Marielle is also a psychology major and knows about it.
• Therefore, I shouldn’t package the AD reveal as a “big mystery” (because many people might guess it), but as the beginning of a new stage. AD is not a gimmick, but the topic the player truly needs to empathize with.

Fourth Round: Complete Video Game (10/29)

Participant: Ryan Esteban Loo (student of CS377G)

Recording Link here. (Main discussion in the last 4 minutes and 33 seconds)

My Observations

Guidance friction points:

• The “File” area was “not very conspicuous,” causing the player to “search for a long time” to see it.
• The player “didn’t proactively use clues,” preferring to “review previous mems” (re-read memories). This caused him to spend more time “looking for words.”
• In the deletion phase, he entered /delete #18, indicating the # prompt was misleading.
• Bug: The gun keyword was not highlighted.

Puzzle-solving Process:

• When solving puzzles, the player relied heavily on “re-reading, iterating, and going in chronological order,” showing great patience.
• When answering “Type A B,” he tried answers like “make up” and felt the question was misleading.

Ryan’s Feedback

Highlights

• He repeatedly emphasized the UI was “very cool.” (“I think the design, the ui. It’s really cool.”)
• He really liked the combination of “fragmented memories” and the “sorting mechanism.” (“I like how it’s like the fragmented memories and then how you introduce the sorting mechanism.”)
• He believed the game’s biggest appeal was the sense of mystery at the beginning. (“I like how you don’t really know what’s going on at first because everything’s contradictory… I think that’s like obviously a big allure of the game. Like that’s the big draw…”)

Suggestions

• Adjust the timing of the “sort” function’s introduction. He suggested introducing the “sort by time” function after all “base memories” are unlocked.
• Highlight the currently read file. When a player clicks on a file in the file list, that file in the list should be “lightly highlighted.”

Empathy

• He clearly understood the game’s narrative layers. He mentioned that after sorting, he could “more understand what’s happening,” and through the “files,” he could understand the story from an “outside perspective.”
• He thought it was a “very clear and well-organized” story, especially “the way it all comes together at the end.” He concluded: “I think you do a really good job narratively.”

Figure 14: The version Ryan played

Figure 15: The optimized version based on Ryan’s feedback and my observations

Questionnaire

Subsequently, I conducted broader online testing, uploading the game to itch.io and distributing it to my friends with a questionnaire. As of 10/31, 4 people had completed the game, and I also conducted phone/text interviews with the players.

Key questions:

• What is the first word or emotion that comes to your mind after finishing the game?
• What was the most impressive plot point or memory fragment in the game for you? Why did it touch you?
• In the early stages of the game, how did you feel when you were faced with those chronologically disordered and contradictory memory fragments?
• When you finally understood the protagonist’s predicament (Alzheimer’s disease), did your feelings change?
• If you have someone with Alzheimer’s disease in your life, did this game give you a new perspective on them?

Fifth Round: Complete Video Game (10/30)

Participant: Zijian

• Gender: Male
• Age: 28
• Frequency of playing games in the last year (0-5): 5
• Favorite genres: MMORPG (like FFXIV), Social, Indie
• Playtime: 1 hour

Zijian live-streamed his playthrough, so there is a game recording, which is very valuable! (Though he used Chinese for commentary.) The file has been uploaded to the drive.

Figure 16: Zijian’s livestream screenshot

Highlights

• The player liked the game’s guidance, finding it “progressive” and well-integrated with the narrative.
• During the game, the player noticed the “three narrative layers” and called it an “aha moment.”
• The player believed that when the theme of “Alzheimer’s disease” was revealed, the game’s “social value and play value both rose,” praising this concept for “transforming the game from an entertainment experience into a value output.”

Suggestions

• At the very beginning of the game, the player did not read the instruction file and instead directly typed “test” into the terminal. This action “accidentally” caused him to see the ending (as seen around 8:22 in the first playthrough recording). This confused him, and only after studying it for a long time and restarting the game did he find the correct starting file. This was a serious bug caused by a testing backdoor I left in the game.

Empathy

• The player explicitly mentioned, “I was actually a bit moved by the end.” He described this emotion as “a kind of feeling where I don’t know why, but the atmosphere just hits you,” finally summarizing it as “a great and long-lasting love.” He said the epilogue (daily life fragments) served as a good emotional release, making him feel it was romantic and “sweet.” He likes good endings.

Sixth Round: Complete Video Game (10/30)

Participant: Dophin

• Gender: Female
• Age: 30
• Frequency of playing games in the last year (0-5): 4
• Favorite game genres: Text-based games, real-time strategy, and games that don’t require much dexterity.
• Playtime: ~50 minutes

Empathy

• First emotion after finishing: A bit scared” (startled by the bug), but regarding the game itself, “quite relieved, relaxed.”
• Most impressive plot point

• • The garbled text after “killing” Camellia for the first time (91l-8m-obodp).
  • The scene running through the sunflower field (9-9pm-sunflower).

• Reason it was touching
  • The sunflower scene reminded her of a childhood memory of “digging for mushrooms and playing happily” with her mother in a forest park, giving a sense of nostalgia.
• Feeling about the initial memory fragments: Felt the protagonist was quite in pain, “both sad and angry, and also pathetic.”
• Feelings after understanding Alzheimer’s: First feeling was that the protagonist was “especially pitiful,” “having suffered so much when he was young, only to be bullied by this disease in his old age.”

• New perspective on Alzheimer’s patients

• • Believes that if the patient was “someone close,” it would indeed change things, leading to feelings of heartache or pity.
  • Described the feeling as “like one’s own memory is a black hole… with things constantly burning and falling into it,” feeling it was “quite tragic.”

Figure 17: Garbled text (Zalgo Text) after “killing” Camellia

Seventh Round: Complete Video Game (10/30)

Participant: Chang Shan

• Gender: Male
• Age: 24
• Frequency of playing games in the last year (0-5): 5
• Favorite game genres: Many types, including anime games, ARPGs, MOBAs.
• Playtime: 30 to 40 minutes.

Empathy

• First emotion after finishing
  • Thought of some novels. No strong emotion, felt more like reading a story.
• Most impressive plot point
  • Son’s Diary
  • Camellia’s Death (18-8pm-blood)

• Reason it was touching
  • (Son’s Diary) Felt like the first time the perspective clearly shifted to another character, creating an “oh, so that’s how it is” feeling.
  • (Camellia’s Death) He had been reading casually before. After realizing she died, the player seriously re-read the previous information.
• Feeling about the initial memory fragment: Curious. Wanted to know the reason behind these memory fragments and how to sort them. This was his motivation to keep playing.

• Feelings after understanding Alzheimer’s: Felt it was okay, as he had started to guess it towards the end. Also had an “oh, so that’s how it is” feeling at the final reveal.

• The animation after deleting the memory made the player “a bit nervous.” The nervousness came from his experience with similar games. He was worried the game had multiple endings and was afraid that when deleting (inputting the name):

• • He didn’t choose the right name to trigger the “best ending.”
  • He missed some clues, which led to him choosing the wrong name.

• The player mentioned that as a native Chinese speaker playing an English game, his energy was focused on “understanding the text,” making it difficult to form an immersive emotional experience. He was using his “rational brain” more for puzzle-solving.

Eighth Round: Complete Video Game (10/30)

Participant: NNN

• Gender: Male
• Age: 25
• Frequency of playing games in the last year (0-5): 2
• Favorite game genres: Action-adventure, ARPG
• Playtime: ~1 hour

Highlights

• The player found the game’s guidance “very well-done,” “very gentle and reliable,” and well-integrated with the story (file system).
• The player believes the core gameplay system has “a lot of development potential” and “many expectations for the further presentation methods this system could offer.” The player even joked that from an investment perspective, this could be made into a Script Editor for scalable profit.
• Music: “Very on-point,” “has a great sense of the era,” “the part at the end where the vocals come in was also very impactful and fit the theme.”

Suggestions

• Missing a restart button. The player wanted to play a few more rounds to study the game’s special effects and animations but “couldn’t find a way to restart the game.”

Empathy

• First emotion after finishing – “Longing”
  • The player mentioned in the questionnaire that the first word that came to mind after finishing the game was “Longing”. He interpreted it as: for Gregor, Carol, or their son, the 50 years from 1945 to 1995 were, to varying degrees, “worn down and taken away.” He “felt quite deeply for Gregor, whether it was the trauma from his time as a military policeman, his guilt towards Carol, or his regrets about his son.”
• Experience of AD
  • The player believes the core of Alzheimer’s (AD) is a change in the perception of time. He thinks AD is a “shell” here, and the deeper exploration is about “how time, based on one’s perspective, can become accumulated wealth or a constantly depleting principal.”
  • The player clearly felt “the unsettling feeling of an AD patient being out of sync with the time of others and their own body.”
  • The player compared this “sense of temporal disconnect” and “sense of loss” to the scene in the movie Interstellar where the father is trapped in the tesseract (behind the bookshelf) watching his daughter grow up. He felt both explored the themes of “memory” and “loss” and brought him “quite similar feelings.”

Figure 18: Scene from Interstellar about “memory” and “loss”

Reflection

I have some excellent reflections, but this space is too small to write them all. I’ll write the most important insights here, and the others will be in tomorrow’s Reflection assignment:

• Less is More: To complete the work in a short time, I didn’t design overly complex task motivations or plots. This is also related to the non-linear narrative method—the work already contains a puzzle, and the player needs to read fragments to piece the puzzle together. This design goes from divergent to convergent (unlike other classmates’ works that go from convergent to divergent – multiple choice branches). The beginning can’t have too much narrative freedom, otherwise, it will be difficult for the player to form their own understanding of the story. Combining the use of procedural rhetoric and textual narrative, and letting the two fully integrate, was a key thing I learned this time.
• Method Library: This course provided many opportunities for practice and experimentation, as well as theoretical frameworks and scaffolding to support that practice. For example, the GMC (Goal, Motivation, Conflict) framework and the Precise techniques provided in this course are effective frameworks for writing stories, helping me overcome the “difficulty of starting.” The reference games provided in this course were also very enlightening. Based on my experience, DDLC‘s narrative techniques led me to form the “three-layer narrative” framework, while “Buddy sim 1984” made me fully explore the expressive forms of a terminal, creating glitch animations and other special effects. The application of procedural rhetoric in these games also inspired me to implement designs like “using gameplay to simulate an AD patient’s memory” and “letting the player personally delete themselves.”