P2: The Empathy Machine-Paprika

Only chapter 1 with no visuals/music

Overview

Test version: Version 1.0: Only chapter 1 with no visuals/music

Player: Young Male

Recording Link: (Didn’t record)

Testing method: In class

Feedback received

• The player described the story as highly subjective, with narration that feels like memory recalling.
• The player praised the evocative language and strong visual imagery several times.
• The player felt that the choices were not yet meaningful enough: they had little noticeable influence on later events, making the experience feel closer to reading a book.
• However, players also felt this approach was reasonable for an opening that establishes the setting. They wanted to see whether the choices would become more meaningful and engaging afterward.
• The player thinks I should add more sentences mentioning “the time is going fast”. The Visitor’s inner voice might occasionally express exhaustion or even think, “I can’t wait to leave” .

Changes After Playtest 1

Completed all chapters without visuals/music/UI.

Added three conflicting inner voices for the player character:

Added a calendar countdown to the game UI and a page-turning effect between chapters.

Overview

Test version: Version 1.1: All chapters with no visuals/music

Recording Link: (Didn’t record)

Testing method: In person

Player: Young Male, never played IF before, Asian

Feedback received

👍 What’s Working

• Great writing — The prose is beautiful and natural. Especially the chapter in the cornfield(the one that I polished the most).
• A complete story — The story feels well-rounded.
• Fun mini-games — Lots of cool ideas with potential, though most aren’t implemented yet.
• inner voice system works — The protagonist’s inner inner voices fight each other and can provide contexts for players.
• Easy to get immersed in — The writing feels real and pulls you in.

🔧 What to Improve

• Make choices harder — Most choices feel too obvious. The player said, “This isn’t a real choice for me.” Each option should come with a meaningful trade-off.
• Build the setting more — The poverty context needs more detail early on to help the player better understand and connect with the protagonist’s situation.
• Resolve the bullying storyline — Bullying comes up in Chapter 1 but never gets addressed again. Give it some follow-up or closure.
• Additional Feedback — The player also gave plenty of feedback on specific plot points and dialogue wording, which I won’t go into detail on here.

Overview

Test version: Version 1.1: Only chapter 1

Recording Link： https://stanford.zoom.us/rec/share/bzLh7XdUcKzQKuYs8UAgV_XEIThP5UbSmUXd_2I5S8fZMHIvO3_eK52IFyIhzKR4.ChlS8nH_vqXgvs_J

Password： Fum.ra2Y

Testing method: In class

Player: Young Female, has teaching experience

Feedback received

👍 What’s Working

• Custom naming: The player loved being able to type their own name instead of picking a preset. Choosing “Poppy” made them noticeably more invested.
• The moment of hesitation before a choice: The narrator’s internal struggle over whether to step in and ask Mai what happened landed well (“I do love the urge”).
• Mae’s characterization: The player felt empathy for the ostracized girl. The detail of the other girls quietly shifting away from her worked.

🔧 What to Improve

• The name input isn’t obvious: At first the player didn’t know how to enter a custom name.
• The conversation gets cut off: Mai starts explaining why she’s upset, then the scene jumps to after dinner. The player wanted to hear the rest.
• Time skips feel abrupt: Jumps like “one day later” and “six days later” felt awkward.
• Not enough context on poverty: Both the player and Yifan couldn’t picture how hard life in the area really is. The environmental details (straw mats, sheets, the village shop) didn’t convey the feeling you intended. Instead, a player with related real-life experience filled in their own associations and said they “don’t know what I’m supposed to be feeling.”
• The phone breaks immersion: Mae using a phone clashed with the player’s idea of poverty. The details that would explain it (a lunch of just two steamed buns, climbing trees to pick pears) aren’t in the text yet.
• The country and culture are unclear: The story never says where it takes place, so players judge “poverty” by their own cultural standards.

Changes After Playtests 2 and 3

Spent more than 25 hours on this iteration version.

Visuals and UI
I added substantial visual content, including background images, movable character figures layered over them, and new UI designs. I will discuss the design process and visual iterations in more detail in the Part Visuals below.

Dialogue and Usability
I revised many conversations based on playtest feedback and made several adjustments to the interactions and interface. These included adding a “Return to Main Menu” option to the pause menu accessed with Esc and providing clearer instructions on the name-entry screen.

A Stronger Opening Hook
I observed that two playtesters were not particularly engaged during the opening. The narrative felt too much like a straightforward sequence of events, without a strong reason to keep playing. To address this, I added a cold open with an element of mystery, giving players a clearer motivation to uncover the truth and continue the story.

Expanding Maizi’s Story
One player wanted to hear the complete story during the scene in which Maizi is first bullied, so that the interaction would feel more rewarding. I expanded the scene to include the full story.

Another player wanted to see some follow-up to the bullying, so I added a small moment later in the narrative in which Maizi and her classmates gradually begin talking to one another.

More Player Involvement in the Rainy-Night Confrontation
The Playtest 2 participant felt that the rainy-night confrontation—the story’s largest and almost only major conflict—did not give the player enough involvement. I added timed choices and expanded the scene into a series of consecutive decisions to create a stronger sense of back-and-forth negotiation.

A Timed Braid-Making Interaction
Participants in both Playtests 2 and 3 wanted more interaction and a stronger sense of influence. In response, I added a timed braid-making interaction to give players a more active role in the scene.

The hair-braiding minigame serves two purposes.

• It adds variety to the long stretches of clicking through dialogue and making choices.
• It connects gameplay with the narrative, allowing players to experience firsthand how hectic it can be for a teacher to care for several students at once—and how meeting one student’s needs can mean compromising on another’s.

The minigame received consistently positive feedback in every subsequent playtest and became one of the most memorable moments for many players.

Clearer Context Around Poverty
Two players felt that the story did not provide enough context about poverty in the village. I added more specific details to the opening chapters to better establish the characters’ living conditions. One player also mentioned that the presence of a mobile phone disrupted her understanding of the setting, so I added context explaining the phone and its use.

Pacing and Playtime
I noticed that several players took longer than intended to reach the core of the story. To improve the pacing, I removed the early dormitory-building and straw-mat sequences, which contributed less to the central themes and setting.

Overview

Test version: Version 2.0: All chapters with visuals but no music

Recording Link： https://drive.google.com/file/d/1n2hXUHaiwRrJtD55acJXJAK8XETW97q4/view?usp=sharing

Testing method: Online and using the Chinese language version

Player: Young Male, Asian

Feedback received

👍 What’s Working

• Emotional Response: The player commented on the hardships the children faced at several points in the story.

🔧 What to Improve

• Chapter Transitions: The jumps of ten or more days between chapters felt somewhat abrupt and jarring.
• Ending Feedback: The ending lacked clear feedback and a strong sense of accomplishment.
• Suggestion: Add an ending collection system to make completing different endings feel more rewarding.

Changes After Playtest 4

A Personal Photo Collection as an Ending Reward

The player suggested adding an ending collection system. Since the game follows a linear storyline, I decided against introducing a collection system for endings.

However, their other comment—that they were unsure what the ending meant—suggested a need for a stronger sense of accomplishment and something meaningful to take away from the experience.

To address this, I replaced documentary filmmaking with film photography, making this part of the game more interactive and immersive while giving players more freedom. In different scenes, players can choose whether to take a photograph, and a minigame determines whether the shot comes out clearly or fails.

At the end of the game, after returning to the city from the village, the player develops the film. The resulting photographs form a personal collection shaped by their choices and interactions, giving them a tangible record of their experience and a more meaningful conclusion.

This design received consistently positive feedback in subsequent playtests.

Sounds (ambient musics & sound effects)

Added 8 ambient musics to fit both indoor and outdoor scenes.

Added 10 sound effects to important interactions and events.

Overview

Test version: Version 2.1, only tested Chapter 1 – 3 in the limited time.

Recording Link： https://stanford.zoom.us/rec/share/zrsr4crHxw6dF9aTRj8wqAXD5Rzue5ZMfyGX0ngffbM_o9W0CgEMTa8ibC-WJVTw.ejpphnT_Vb2Hm0Wj

Password: J5EMF#Mn

Testing method: In class

Player: Young Male, has teaching experience, Asian

Key moments

– **[09:18–09:26]** “I really like the way you put the sound, the background noise, which is related to the scene, to the character, to what’s happening.”

– **[10:02]** “It’s like in my real life as well.”

– **[10:29]** “Many things happen at the same time.” → refers to the braiding game.

– **[10:49–12:08]** Braiding minigame: “Try again…”, “Oh, it’s reverse!”, “Oh my gosh!”, “right, left, right, left… Finally!” → lots of emotions and reactions.

– **[17:46–18:03]** “I love the visualization… you use both the real pictures and the cartoon stuff… not too real, not too fake, in the middle. I love it.”

– **[18:03–18:25]** “The sound… helps a lot for me as a player to understand what’s going on. Like in the classroom you can hear the students talking in the back, and it goes along with the captions.”

– **[18:32]** “The visualization is perfect, the sound, perfect.”

– **[18:39–19:13]** “I can relate. As a teacher I understand the context… teaching students in a rural area, far away in the mountains.” Said it reminded them of volunteering at a mountain school, and that the “it’s not that far” line really captured it.

What worked

– Sound design is a big win. Background noise matching the scene (like kids chattering in class) helps players understand what’s going on.

– The art style (real photos + cartoon stuff) lands in a sweet spot: “not too real, not too fake.”

– The story feels authentic and relatable, especially to someone with teaching / volunteering experience. Small details like “it’s not that far” hit hard.

What didn’t

– The story is real but not interesting enough (only tested the beginning chapters).

Overview

Test version: Version 2.1

Recording Link： https://stanford.zoom.us/rec/share/xTlcZvICfKyRjvD1C2AagCDUMKG-1v7HFfwXTAGwtO4Js3A6v_YDKwu_RFSbU6vB.f_EFfPahsL-Joc3o

Password: Wmp57*Q@

Testing method: In Office Hours, in person

Player: Young Female, Asian

Key moments

– **[15:05]** On the photo-taking feature: “Oh, that’s so cool!”

– **[15:41–16:10]** Favorite scene: “The braiding scene stays in my memory… it was the only place where it’s not choosing dialogue, but actually braiding the hair.”

– **[16:10–16:16]** “I was thinking the favoritism was gonna have more impact on the outcome… it was only that the other kids didn’t really talk to me afterwards.”

– **[17:06–17:26]** “What’s the significance of the dog killing the chickens? Does it impact any of the outcome?” → the player couldn’t see the link between that event and the ending.

– **[19:25–19:56]** “Through the game I feel like I built this relationship with Mai… over like 2 months… with the days ticking, it creates that sense that a lot of things are happening.”

– **[20:02–20:29]** You asked about the time skip (10 days in one second) feeling weird. Player: “I thought it was perfect… when you look back on memory for 2 months, some moments stay in your memory and some don’t. It’s very reflective of that.”

– **[21:10–21:30]** On the photo review at the end: “I feel like this should be in every branch… it creates empathy… it actually feels like you spent that time and now you’re looking back on it.” Suggested using film pictures if the player didn’t take any of Mai.

– **[22:03]** “I didn’t know you could do that [in Ren’Py]. I thought it was more limited.”

What worked

– The braiding scene is the standout moment because it’s the one physical, non-dialogue interaction.

– The photo mechanic, and especially looking back at your photos at the end, feels personal and builds empathy.

– The bond with Mai comes through. Time skips feel like real memory (some days blur, some stick) instead of feeling weird.

What didn’t

– Some choices don’t feel like they matter enough.

– Some story events seem disconnected from the outcome (dog killing the chickens → how does that affect the ending?).

Overview

Test version: Version 2.1

Recording Link: (Audio Only) https://drive.google.com/file/d/1E2UQZEPmTApm_enDrXsFg5ayFqMK1Kxd/view?usp=sharing

Testing method: In Office Hours, in person

Player: Young Female, not a big fan of IF

Key moments

– **[00:05]** “I feel like it’s a very complete story. One thing is the passage of time feels kind of arbitrary, like the amount of days.”

– **[01:26]** “I just can’t imagine someone being like, ‘Oh, I’m just gonna take my pictures.’ I thought the focus was more on how can I get to know this person and help them in some lasting way before I leave.”

– **[01:53]** “We saved her dog, which was nice, but it would be cool to know I can, like, give her confidence or something.”

– **[01:53–02:24]** “I don’t know that it needs to be 60 days… every time the days pass I don’t really know how long is gonna pass.” Suggested a month, with **consistent** jumps each time.

– **[03:23]** “Maybe go through each line and decide if you want to keep it… some description wasn’t super necessary. And if people get tired, they skip and miss the important part too.”

– **[06:51–07:16]** On the “all this corn is the mountain’s children” line: “I honestly did not get this reference… it’s poetic, but I don’t understand poetry.”

– **[08:57]** “Because the mother seemed happy, I was like, oh, they’re not actually down… maybe it’d be more compelling to just see them from the back.”

– **[09:27]** “When she was turned around, I loved the details of the braid on her… I thought, I should have taken a picture there.”

– **[10:00]** “I did empathize with her on the fact that all her teachers leave and she never has a consistent relationship.”

– **[10:00–13:06]** Confused about the setup: “Are they teaching English?… I honestly thought this was a summer English exchange… because the name was like Ginger.” Also didn’t get why teachers use nicknames, or what subject they’re teaching.

– **[13:25–13:45]** Suggestion: “Just short context: OK, I’m here to teach for a month, then next month they get a whole new set of teachers, and I’ll probably never see them again.”

– **[14:08–14:14]** “My understanding was, I want to make the most of my two months… I want to be as helpful as possible… maybe if I make an impact, I can meet them later.”

– **[14:42]** “It’s a very cool concept. I love the art. It’s more about clarifying… what is it exactly? The mother? The family situation? That she doesn’t get a stable education?”

– **[15:33–16:01]** “I kind of feel like maybe I didn’t choose the right choices to fully try my best as a teacher… I could have done better, and maybe that would’ve led to a better outcome. So that hopelessness isn’t fully there yet.” (You said the core feeling is “I can’t do much, I can just observe.”)

– **[16:52–17:26]** “I wasn’t clear on why the guy gave us the mosquito thing… out of nowhere.” And “why did he beat the dog?” → the neighbor’s motivation didn’t come through.

– **[17:58]** “It’s very enjoyable. I didn’t feel like it was really long.”

– **[18:51–19:11]** “I actually think [the big family] is positive, so I didn’t see it as a negative for her… I just felt bad that I couldn’t do more for Mai.”

What worked

– It feels like a complete, enjoyable story, and about 20 minutes didn’t feel too long overall.

– The art is loved, and small visual details (like the braid on Mai’s back) stand out.

– The player empathized with Mai’s core situation: teachers always leave, so she never gets a steady relationship.

– The player felt the “I couldn’t do enough for her” feeling, which is close to the intended theme.

What didn’t

– The passage of time feels random. Day jumps aren’t consistent, and 60 days feels arbitrary.

– The setup wasn’t clear: who you are, what you teach, why the nickname “Ginger,” and how the volunteer program works. The player assumed it was a summer English exchange.

– The intended sadness / hopelessness didn’t fully land. The family looked happy in the images, so the player didn’t feel their life was bad.

– Poetic lines (the corn / “mountain’s children”) went over their head.

– Some long descriptive passages → skipping.

Overview

Test version: Version 2.1

Recording Link: https://stanford.zoom.us/rec/share/aUoyjVwA4tjJMgl_gapHUU4tV5q9aZC-Gr4E5crfU4pXVFobC6_8psS7IWyQS8X2.HBWZ5c9tR4amjAE_

Password: UY@6PN1=

Testing method: In Office Hours, in person

Player: Young Male

Key moments

– **[00:05] / [00:20]** “Oh, I love the art!”

– **[02:42–03:15]** Confused by the speaker labels: “A teacher inner voice? A passerby inner voice? Who is the passerby?… it said ‘the teacher,’ so that’s me?”

– **[07:17–07:23]** On the bedsheet choice: “This is cool, because it makes me feel like my choices did something.”

– **[08:01–08:07]** On “Her earlier choices have left their mark on how this moment unfolds”: “That text could be shorter. It feels like it’s explaining to me.”

– **[08:36–08:40]** “There’s 59 days left? If I have to go through this 59 more times, I’d be crazy.” → worried about length.

– **[09:29]** “I don’t remember what choices I made that could have impacted this moment.”

– **[11:43–12:09]** “I like how each chapter is named after an object, like the lens… but it takes a while to figure out what ‘lens’ means. If that was the first sentence, would it be more impactful?”

– **[12:16–12:24]** “The lens is talking? What?… Wait, the camera is thinking inside our brain?” → the voice of the camera / inner voice was confusing.

– **[13:24]** “The visuals go well together.”

– **[15:33–15:45]** Refused to take a photo: “It’s like going to a poor country and taking pictures of starving kids… but that’s an interesting moment of tension.” → the photo mechanic sparks real moral thinking.

– **[16:09]** “The pictures are cool. I like the idea of choosing what’s important and what’s not.”

– **[17:02–18:04]** On the “pebble” and hidden trust: “I’m wondering how long it’ll take until I understand the pebble… oh, it takes a while for the pebble to come into play.”

– **[19:56–20:20]** “My biggest gripe so far: some parts can be cut… you’ve done a good enough job with the background telling me what the environment’s like.”

– **[20:40]** “I’m not resonating with any of the other students, just Mai, which I actually like a lot, because we’re building this personal relationship.”

– **[21:38]** “I like that you started at the climax.” (the window chapter)

– **[24:44]** On leaving the camera for Mai: “I didn’t know I had these cameras… This might have been nice to know before the choice.”

– **[27:02–27:20]** On the ending: “I see where that was going, I’m starting to lose connection with this person… I wish that went a little longer, it was like 3 days passed… I liked the text format a lot. I wish it could’ve been like ‘I didn’t respond to her for 3 months.'”

– **[30:04]** “I like the story. It’s very memorable in my head. There’s a lot of dialogue you can cut because it’s a little over-descriptive at times, which takes me out of it.”

– **[30:23–30:38]** “The mechanics are really interesting. The polish is beautiful. I like the pictures, I like the braiding… the braiding gives me urgency. The timed choices when the dog is getting beat up are really nice.”

– **[30:38–31:24]** “I don’t know if you should ask the player if they want timed choices or not… it spoils it a little, like ‘stuff is gonna be timed, stuff’s gonna get serious.’ Might be better to remove it.”

– **[31:24–31:52]** “I like that you have different emotions for her. I feel very immersed… the pictures help a lot, I was never lost, I knew what setting we were in and who we were meeting.”

– **[31:52–32:52]** “I was gonna complain about no meaningful choices… but after playing the whole thing, seeing the person who’s nicer to me because I said hi, and the dog being safe… choices-wise, this is fine.”

– **[33:00–33:29]** “Fluffy parts make me want to skip, and then maybe I skip something important.” Specific scenes: walking through the forest, walking through the cornfield.

– **[33:59]** Tip: “Run through the script and ask, if I remove this, is the story still the same? Do that for every line.”

– **[34:13–34:48]** “I didn’t understand the camera part at the end… she didn’t pick it up because she doesn’t trust me? I saved her dog and she doesn’t trust me?” → hidden trust result felt unfair.

What worked

– The art and photo shader look great: “strangely real but gamey enough.” Visuals all fit together.

– The story is memorable and immersive.

– The focused bond with Mai (and not the other kids) works well.

– Mechanics: braiding (with the keyboard option) and timed choices in the dog scene create real urgency.

– The photo mechanic leads to real moral moments.

– Starting the window chapter at the climax works. Chapter names based on objects are a nice touch.

– The text-message ending format is liked.

– By the end, the choices felt like they paid off (the neighbor being nicer, the dog surviving).

What didn’t

– Too wordy. Some descriptive dialogue repeats what the images already show (forest walk, cornfield walk), so the player wants to skip.

– Confusing labels and voices: “teacher inner voice” vs. “passerby,” and the camera / lens “talking.”

– “59 days left” made the game feel daunting.

– Hidden trust felt unfair in the camera ending (“I saved her dog and she doesn’t trust me?”).

– The fade-out of the relationship at the end felt rushed (only ~3 days passed), and the “How do I remember you” meaning didn’t fully land.

Changes After Playtest5 – Playtest8

Dialogue and Pacing: Removed more than twenty lines of dialogue to tighten the overall pacing. However, I retained the “poetic” cornfield passage that participants in Playtests 7 and 8 disliked, because it explores a central theme I want to address: why families in rural poverty may continue to have more children despite financial hardship. Several players in other playtests also specifically mentioned enjoying this passage.

Timeline: Shortened the story’s timeframe from 60 days to 30 days and adjusted the calendar transitions accordingly.

Mother’s Character Image: Changed the mother’s image to a side view that keeps her face out of sight. This is to give players a space of imagination and interpretation. (Changed from left to right)

Opening Choices: Removed the time limit choices at the beginning of the game.

Opening Context: Added a brief explanation that the player character is staying for only one month and teaching English. The village has a shortage of teachers, so university volunteers take turns teaching during the summer.

Trust Feedback: Made Mei’s reactions more noticeable when her trust increases or decreases, including changes in her facial expression. Added brief text notifications, such as “Mei trusts you more.”

Dog Rescue: Significantly increased the trust gained from rescuing the dog.

Dialogue Clarity: Clarified which chicken the villagers are discussing in the transition scene, making it clear that the animal that died was a chicken.

Photography Tutorial: Added an option to photograph the classroom without inviting Mei to be the subject. If the player chooses this option, Mei’s character image is removed from the scene. Both choices lead to the photography minigame, ensuring that this scene consistently introduces its mechanics.

Choice Notifications: Temporarily removed the “Your earlier choices affected this moment” notification.

Background Music: Added several background music tracks.

Overview

Test version: Version 2.2

Recording Link: https://stanford.zoom.us/rec/share/fEQ5sMF6B8WDKja_yfU-2e4ade-bCaE2B9M2YPXmsT5EW9cRKCDm9991bDknbO2y.iEc8WEoNuVpxLJ9m

Password: s4CwP8r@

Testing method: In class

Player: Young Male

Key moments

– **[13:34]** “Maybe I didn’t fully follow… why was I, as a teacher, taking photos?” → the photo contest setup didn’t stick.

– **[14:10]** “It’s hard to take photos… easier with a mouse than the trackpad.”

– **[14:36]** “Not sure you need both the act structure and the day structure.” If you keep the calendar, show the days flipping by during jumps.

– **[15:50–16:24]** “Honestly, not much empathy for Mai. As a teacher, I want to care about all my students equally… saying no to other kids felt weird.”

– **[16:25–17:55]** “I’ll save the dog, but there’s no longer-term fix (like keeping it tied up at night).” Didn’t want more power. The villager’s “you’ll leave and it’ll happen again” line just felt a bit forced.

– **[20:05]** “I still liked the story. It seems polished.”

Worked: the story and polish overall. The player was fine with the “you can’t do much” theme.

Didn’t work:

– Favoring Mai read as bad teaching, which lowered empathy.

– The photo contest goal was unclear.

– Photo-taking was hard on a trackpad.

– The dog plot has no follow-up.

Overview

Test version: Version 2.2

Recording Link: (Sadly I forgot to turn on the camera and was also muted)https://stanford.zoom.us/rec/share/YQbbI8jv_3nWnXPMOSnAg29C_Zoh8DbOiDohfTWTF_ggTo4fd1C_4HDRdu8nCFZ9.b1eDtwPFLI06xAKj

Password: e6e=8*a&

Testing method: In class

Player: Young Female

Feedback

Overall Impression: Reflecting on our conversation before Playtest 1, the player said, “I was already excited to see how this project would turn out, but I didn’t expect it to become this good!”

1. Hair-Braiding Minigame: The player found this section frustrating but successfully completed it on her second attempt.
2. Photography Moments: She was excited by the opportunities to capture spontaneous moments, especially when the dog chased the geese. As soon as the geese appeared, she anticipated a photo opportunity and repeatedly exclaimed, “OMG, pictures, pictures, please!”
3. Emotional Ending: At the end, she remarked, “Oh that was such an emotional ending.”
4. Narrative Progression: She appreciated the sequence of revelations: learning that Mei’s younger sister had drowned, discovering that her mother was pregnant again, and then reaching the discussion about the corn harvest. She felt this progression was smooth and thoughtfully designed.
5. Cornfield Scene: She particularly liked the poetic language in this scene. On seeing the mother’s updated side-profile image, her immediate reaction was, “Oh, is she pregnant again?” The subsequent dialogue about children and corn left her feeling sad.
6. Background Music: She felt that the newly added music strengthened the emotional atmosphere.

Changes After Playtest9 & Playtest10

Fixed a few bugs that happened.

Changing the calendar flipping effect based on the feedback.

Dialogue Picture: Added a selfie of Mai and Maizai in the texting part, to give players more rewarding on saving the dog.