You are small, the land is vast.
Your goal is simple…Stayin’ Alive!
Play now! https://lwcoding.itch.io/stayin-alive
Team: Lucas Wang, Leyth Toubassy, Krystal Li, (and in spirit: Ngoc Tran 🙌)
—————————————————————————————————-
Overview
Stayin’ Alive is a survival-strategy ecosystem game where players inhabit the role of an endangered kangaroo rat and navigate the changing seasons with predators and scarce resources in the Southern California shrublands. Players explore, collect materials, build and manage dens, recruit worker rats, and maintain a sustainable food system for themselves and their colony. At its core, the game is a dynamic system: grasses grow and spread, predators hunt, seasons shift resource availability, and player choices shape emergent outcomes.
Our primary target audience is the same as P3: individuals who like nature, animals, and learning about natural systems. Stayin’ Alive is specifically designed for players who enjoy strategy experiences that blend atmosphere, survival tension, and ecological thinking. This includes fans of resource-management games, creature-based sims, and slow-burn survival loops where thoughtful decisions matter more than fast reflexes!
For P4, our primary goal was to balance the game and remove any game-breaking bugs, improve onboarding and keep the ecosystem grounding clear and intentional. We also wanted to add polish to our game and web page to share with more players. During this project timeline, we conducted 8 playtests and our final game reflects significant iteration informed by playtest data and applying the course concepts.
System Representation
Stayin’ Alive is grounded in a real shrubland food web, inspired by the habitat of the endangered Stephens’ kangaroo rat in Southern California. In this ecosystem, kangaroo rats are preyed on by predators like red-tailed hawks and coyotes, while desert cottontails act as additional prey in the same landscape. We also include pollinators/insects (monarch butterflies, bees, and worms) to reinforce the broader ecology and “living world” context.
Several mechanics are directly rooted in real ecosystem dynamics. For example, after successfully feeding predators will back off for a period of time after a meal. Moreover, plant growth is shaped by environmental conditions similar to real life, with access to water and the presence of pollinators/insects supporting more vegetation, which we represent through faster grass growth near water.
Our resource loop reflects what kangaroo rats actually consume. Kangaroo rats eat grasses and seeds, and they also consume insects; we abstract this diet into two clear, legible food sources: grass (primary plant food) and worms (insect/bug stand-in) distributed across the world. You can read the ecosystem as feedback loops: a reinforcing loop where food → survival → more workers → more gathering supports growth, and a balancing loop where predators + winter scarcity push back against huge populations.
The first and biggest conceit of Stayin’ Alive is the existence of the Kangaroo Rat god. We cannot say with any certainty that there isn’t a Kangaroo Rat god, but if there is it’s a happy accident. The reason for adding this narrative element is that we wanted to ensure we had some kind of in-world reasoning for the player rat having the intellect of the human player.
Another intentional abstraction in Stayin’ Alive is that the world is always daytime. In real life, time-of-day can shape animal behavior and modeling a day and night cycle would require more complex additional systems and explanation of shifting rules. We chose perpetual daylight to keep the core ecosystem understandable and learnable without introducing too many overrealistic layers. Similarly, the game runs on a compressed timeline. Seasons, growth, starvation, and population change occur much faster than they would in real life so players can observe cause-and-effect and adapt strategies within a single session.
Overall, our goal is not a perfect simulation of nature, but to model a system where real ecological relationships (predator pressure, food scarcity, seasonal stress, and habitat constraints) engender different unique strategies for players, so we felt the abstractions were appropriate.
Progress and Revisions
Playtest Media Folder (Use Stanford email to access!)
Phase 1 (System Understanding):
1. Sam Jett 12/3/25
For our first playtest, we playtested with Sam Jett, who is a CS Coterm that has previously taken 377G and enjoys making games. Sam was not fully in our target audience of animal lovers and casual game enjoyers, but their feedback was helpful for improving system feedback overall.
During the playtest, Sam spent most of his time reacting to immediate threats rather than planning ahead, which led to repeated deaths without a clear sense of cause. This surfaced an issue we had seen in other playtests: while the ecosystem systems are functioning, the feedback explaining those systems is too subtle, making it hard for players to form strategy early. Sam did eventually begin to associate safety with dens and food-rich areas, but only after several failed attempts.
This playtest reinforced that our onboarding needs to do more work in explaining system relationships (hunger → food → breeding → population) and long-term strategy.
What went well:
- The survival tension and predator pressure are engaging
What went wrong:
- System feedback around hunger, death, and population needs to be clearer
- Players need stronger early guidance to discover strategy and not feel like they’re just surviving moment-to-moment
[Image of Sam playtesting.]
2. Louis Lafair 12/3/25
In our second in-class playtest, we were able to playtest with Louis, who is a seasoned game designer and enjoys systems games especially. However, he primarily plays board games and was a bit unfamiliar with digital games, so while Louis did not represent our target audience he provided a lot of useful information about how to improve the system feedback and overall gameplay.
At the start, Louis immediately tried to understand the core systems and controls. He questioned basic mechanics like hunger and breeding, asking, “does hunger just go down over time?” [01:12] and “to breed, do you need to be in the den?” [01:45]. While he grasped teleportation between dens quickly, “Oh, okay, that makes sense” [02:30] — several elements about the environment confused him. As he explored the map, he repeatedly bumped into water tiles and commented that the map felt tricky to read, saying “I keep running into water” [05:02]. When attempting to build a den, he momentarily stalled due to missing resources: “how do I build the den? … Oh, I need one more” [04:21]. This was a success as it showed that the feedback saying “need 1 more stick to build” was useful feedback for the user.
During early predator encounters, Louis really felt the tension of surviving as a rat and was reacting audibly when threatened: “ahhh!” [03:58] and later, watching a coyote linger, “is he just gonna stay there?” [07:10]. He understood that rabbits were abstract entities but wasn’t sure how much attention they deserved, asking “Do I care about rabbits, or are they just to distract me from predators?” [08:42].
As the population grew, clarity around breeding and death mechanics became an issue. The dying log wasn’t immediately legible to him, and he struggled to connect deaths to causes. He later suggested the log should be more prominent and tied to MVP progress. Mid-game, Louis began questioning scalability and balance. He said “50 feels really far. Babies keep dying and I don’t know how to sustain it” [15:22]. While he felt seasons made the game harder, he couldn’t clearly articulate why, noting “It feels harder now, but I’m not totally sure what changed” [14:05].
Despite the friction, Louis consistently described the experience as engaging. He enjoyed movement, visuals, and overall cohesion, saying “It’s fun just moving around” [09:30] and later calling it a “fun challenge” [16:10]. He appreciated the thematic consistency and said the abstraction of animals “made sense” once he understood the system relationships.
What went well:
- Louis immediately understood previously confusing features like controls in the den and teleportation between dens
- Felt that the challenge of moving around and stocking up food was engaging
- Felt the differences in seasons
- Had strong emotional reactions to predators and danger, showing we were able to create real empathy for the rat in the player.
What went wrong:
- breeding mechanics were unclear (requirements, cost, scaling)
- hunger and death feedback lacked clarity
- the map navigation was frustrating for Louis due to water tiles but since other players that better fit our target audience didn’t have this problem, we decided to ignore it.
[Image of Louis playtesting.]
3. Jackson 12/4/25
In our third playtest, we playtested with Jackson, a college-aged player who enjoys games casually and represents our target audience in terms of comfort with digital controls and liking light-hearted games.
[Image of Jackson playtesting.]
Jackson began by probing the basic interaction loop, asking clarifying questions about how to pick up and store resources: “How do you pick up the grass?” [02:19] and “How do you put the food in the den?” [04:26]. As he explored, Jackson reacted positively to small moments of feedback. He laughed when items dropped into the world saying “I like that the items fall into the… that’s funny.” [06:15] and later explicitly noticed visual fidelity, saying “Oh, I like that animation, that was cool.” [09:33]. These reactions suggested that the feedback helped reinforce engagement even when mechanics were still being learned.
Over time, Jackson was beginning to pick up the system relationships. When predators approached, he connected movement rules to danger, saying “Oh, and it only can move when you move.” [08:31]. He also started forming hypotheses about population strategy, asking “So you want some of your babies around to protect you…?” [08:02]. These moments demonstrated early understanding of emergent strategy, even if not yet executed optimally.
However, feedback around hunger and death was too subtle. Jackson became aware he was failing only after the fact, reacting with “Oh I’m starving.” [10:55]” and later “I don’t know when my guys died, though.” [10:51]. This indicated that while the system was functioning, the game was not clearly communicating cause-and-effect in real time.
Overall, Jackson’s playtest showed that players can discover systems through exploration and begin to reason strategically, but there was unclear failure feedback that prevented players from adapting before collapse. Jackson really enjoyed the animations and den inventory setup though, and reinforced that the polish we added to both of these things were successful.
What went well:
- Enjoyed the den inventory and thought that it was intuitive
- Understood different behaviors within the system and began strategizing how to move with babies and avoid predators
What went wrong:
- Struggled with some controls after going through tutorial quickly
- Missed some key information from the game log that indicated the babies were dying fast
Phase 1 Changes:
Overall, this update is about making the environment more sustainable and clarify some of the game controls so players can actually notice what’s happening, understand why it’s happening, and adjust their strategy to survive and thrive in the world. We addressed a lot of the friction that was coming up in playtests by surfacing key state information, reducing annoying resource handling, and smoothing the game’s difficulty curve. Here are all the changes that came out of these first three playtests:
Clarity
- Moved hunger bar and seasons counter to the bottom for better readability
- Added text popup that indicates to the player if they don’t have enough sticks
- Add Knowledge menu that shows things you have previously discovered
- Added counters to shot the number of assigned/unassigned workers, and their respective limits
- Assigned Rats now share gathering strategies over the burrows… increased gathering efficiency
- Added blinking outlines to the teleport menu icons
- Gave a fresh coat of paint to the den administrator menu
- You can now properly enter a den while a worker is standing on top of it
Resources
- Kangaroo rat workers can now bring non-grass items back to the den
- Added trees procedurally generated around the map, which spawn sticks around them
- Worms now spawn around water rather than separate “pond” tiles
- Map size increased by 20% to accommodate for more structures
- Added structure (grass patches): Just a patch of grass!
- Added structure (bee trees)… grass spawns 50% faster around bee trees
Balancing
- Increased time for season length by 50%
- Decreased how quick the breeding cost scales for workers
- Grass spreading now spreads 25% faster (when at full tier)
- Grass that spawns or is placed next to water now grows 25% faster
- Decreased drop chance for seeds when harvesting grass by 50% (at both tiers)
- Decreased natural grass spawn (this is made up for in other structures)
- Grass now does not grow at all during the Winter phase
- Amount of grass in grass patch reduced from 9 -> 5
Polish
- Increased number of butterflies that spawn across the map
- Compressed audio and texture sizes for faster load times
- Back-end optimization with object pooling and cacheing that should make the game run a little smoother
- P button now pauses the game instead of Escape, so you don’t exit fullscreen
These changes are also publicly noted in our devlogs linked below:
https://lwcoding.itch.io/stayin-alive/devlog/1134052/stayin-alive-v18
https://lwcoding.itch.io/stayin-alive/devlog/1134832/stayin-alive-v191
https://lwcoding.itch.io/stayin-alive/devlog/1134832/stayin-alive-v191
Phase 2 (Balancing and Affordances):
4. Brian 12/10/25
In our 4th playtest, we playtested with Brian, who is an Astrophysics major at UC Berkeley and does not regularly play games. He largely skipped the tutorial but was still able to piece together the core mechanics through experimentation across multiple runs, which made his playtest especially useful for evaluating whether the system could be learned organically without heavy onboarding.
What went well was Brian’s ability to develop an emergent strategy over time. By paying attention to the death log, he noticed that babies were starving and adjusted his play accordingly, saying “I realized the babies were starving” [29:20] and then was deliberately stockpiling food in anticipation of winter. When asked about seasonal impact, he correctly inferred the underlying mechanic, explaining, “I think the grass stops growing… you kind of have to stock up in Fall to make sure you have enough to sustain the population” [29:50]. This demonstrated a strong understanding of the game’s internal feedback loops.
In later runs, Brian also emphasized the importance of exploration. He realized that different parts of the map had different resource densities and adapted his den placement accordingly: “It was really important to find lusher areas… so I know where to dedicate dens to collect resources” [34:16]. This was a clear example of emergent strategy driven by environmental understanding rather than explicit tutorial instruction.
However, there were a few issues such as tutorial parts being skipped by pressing “E,” and Knowledge entries reset between runs, saying “a little silly to put three question marks when you’ve encountered it before.” He was able to infer most of the system without reading carefully in the tutorial though so this showed that we successfully built in visual cues and information that could guide the player through the real game.
He also found a softlock in the game because it was impossible to walk through the many rabbits in the scene (see Figure below).
[Image of Brian playtesting where he is softlocked by a rabbit.]
5. Kevin L. 12/11/25
In this playtest, we playtested with Kevin (Krystal’s brother) who is very familiar with roguelikes and long progression games like RuneScape.
[Image of Kevin L. playtesting.]
The playtest began in the tutorial, where the player pointed out some inconsistencies like clicking “OK” with the mouse, saying “wait, do I have to click this one? I thought I was using E.” [01:12] He also suggested that the tutorial should consistently respect core controls and visually signal when progression is blocked: “I need something that tells me I can’t go further yet.” [02:04]. As he continued through onboarding, Kevin also pointed out some readability issues as the tutorial text was “hard to read.” [03:10]
Once in the main game, Kevin began exploring the environment and was a little unsure at first. For example, he clicked on rabbits and paused, unsure how they fit into the food system and asked “How do I eat these?” [05:02]. But after experimenting, he concluded they weren’t interactable. As population and inventory grew, the player struggled to track deaths and population loss, only realizing late that his baby rats had been dying steadily, “wait I only have four rats? have they all just been dying?” [08:19] This moment revealed that while death was mechanically present, its feedback in the log was too subtle to indicate that he needed to change his strategy in real time.
After a few playthroughs, Kevin had a much clearer understanding of the ecosystem, listing interactions aloud as he pieced them together: “Okay, so rabbits eat grass, coyotes eat rabbits, I eat grass, bees make grass, birds eat me.” [12:10] Despite this growing clarity, he felt the difficulty curve was punishing without sufficient explanation, especially around death states, saying “it’s too difficult without telling me why I died.” [13:44] He strongly suggested death feedback be explicit and thematic, “If a red-tailed hawk killed me, just say that, and give me a little fact or something.” [14:06] Toward the end of the runs, Kevin discovered a viable strategy of placing dens near reliable food sources. Overall though, he felt the game escalated population demands faster than resources could reasonably support and questioned whether the world generation sometimes produced unwinnable states and suggested that difficulty should scale spatially to reward exploration rather than punish it.
Overall, despite the player’s frustration about the balancing issues, he said the game was “very well made” and really enjoyed the art. From this playtest we saw that we primarily needed to balance the starting resources more and tweak the onboarding and death feedback to better support learning without external explanations. We made an intentional decision to ignore some suggested fixes such as greater vision in the fog of war. This is because we wanted to maintain the limited vision to add to the theme of survival and stay true to the idea that a small rat in the wild may not be able to see its predators coming from very far.
6. Kevin N. 12/11/25
In this playtest, we playtested with Kevin, a CS undergraduate at Stanford who was struggling on his CS148 final project who was roped into playing the game. He plays social competitive games like League of Legends and Overwatch online, but does not play many singleplayer or systems games. He refused to play the tutorial at the start of the playtest because he had a final project due in four hours, but he used his familiarity with games to figure out the system mechanics. He eventually played the tutorial and won the game on his second attempt.
On his first run (ignoring the tutorial), he learned about various mechanics simply through experimentations, such as the knowledge menu (“Oh, there’s knowledge! What is this?” [2:35]) and seasons (“Oh my god, it’s winter. Everyone’s gonna die now.” [5:46]). He expressed some initial frustration at using the number keys to select items in the inventory on multiple occasions, but when asked what he would prefer instead, he claimed that he’d prefer mouse clicking and later retracted his statement after prolonged play. He made a decent amount of progress and explored various interactables in the environment. Eventually, though, he died (“Oh, shit.” [8:51]).
Then, Kevin played the tutorial, and after correcting some small grammar mistakes in the text, won the game on his second run. Throughout the playthrough, he made a lot of comments that indicated that he was understanding the underlying system. When he found a place that was plentiful in grass, he said, “Oh there’s so many … I assume all the wolves down there killed all the rabbits” [18:45]. Then, when he was trying to gauge if the lower-right area of the map was safe, he reasoned, “Is this place safe? … All the rabbits are completely gone, so I’ll take that as a no.” [22:52]. He also experimented with viewing the knowledge menu during his playtest out of curiosity. He found it to be helpful, but also expected to see information about seasons, where we only initially had information about animals and interactables: “I was hoping it would, like, tell me about what the seasons do, but I’m just gonna assume that, like, it just, like, affects grass growth or something” [25:57]. We added these additional entries after the playtest.
After distributing dens in safe areas around the map and planting grass near each to preserve the rats, he described his strategy as “[the first] 10 minutes were just me, like, looking for places that would be safe and real dense. And then after that, it was just, like, a matter of, like, setting up, like, grass farms that would, like, keep them from starving to death. … Then it’s just setting up that infrastructure enough to, like, be able to support 50 of them” [30:39]. This indicates that Kevin was able to strategically explore the map (which is a behavior we seek to encourage), find out viable locations based on the interactables around them, and both maintain his current population and expand to hit the worker goal. Ultimately, this playtest highlighted the presence of an emergent strategy to win the game quickly; Kevin had never seen this game prior to this playtest but figured out the system within thirty minutes of playing.
[Image of Kevin N. playing.]
Phase 2 Changes:
Overall, this pass was mainly about making the knowledge system more clear, while cleaning up smaller bugs and issues from the playtests. On top of that, we further adjusted the balance and difficulty curve so early runs don’t turn into an unwinnable situation, especially for our target audience of casual players who don’t want a game where they have to optimize and try too hard to play for fun. We wanted to achieve that flow state that balances challenge vs. ability to keep the game engaging. Lastly, we tweaked the grass and seasons so that multiple viable strategies (explore vs. stockpile vs. den placement) can work. A complete list of changes made during this phase are here:
New key feature: Knowledge system
- Created knowledge menu that shows players information about anything they’ve interacted with, gives tips on what these objects do and encourages discovery fun by showing what is yet to be found.
- Tutorial refactored to mention knowledge and logs
- Added knowledge on seasons gained after seasons pass
- A brain particle spawns on items/interactables/animals when they get added to the knowledge menu to cue to the player that they have gained new information
UI/UX consistency + readability
- UI sprites like buttons and dialogue boxes were updated
- Fixed issues with game UI scaling on non-16:9 screens
- Moved current den label/worker counter to be consistent with the unassigned worker count
- “Bad” logs now shake on spawn but shake less long for improved readability
- Updated Hawk Den and Grass Seed Sprites for visual differentiation
- Tooltip text should spawn in the right spots more consistently
- Fixed menus appearing too small on game launch
- Re-stylized knowledge menu to have accurate UI buttons
Balance tweaks (seasons + food economy)
- Increased grass spawn rate in grass patches and bee trees
- Grass now grows 50% faster in the Spring
- Decreased amount of grass that is killed during Winter
Bug fixes
- Fixed bug where special interactables could overwrite the player’s starting den
- You can now walk through rabbits and worker animals (to avoid common softlocks)
- Fixed bug where no worms would spawn throughout the map
- Fixed bug where worker rats only picked up food and ignored other items
- Disabled horizontal scroll in the knowledge menu
- Fixed predator dens not granting knowledge
- Re-added tutorial text that explains grass growth and spreading
- Fixed bug where logs overlapped with inventory
Our changes are also publicly noted in the devlogs here:
https://lwcoding.itch.io/stayin-alive/devlog/1141099/stayin-alive-v1112
https://lwcoding.itch.io/stayin-alive/devlog/1140905/stayin-alive-v111
https://lwcoding.itch.io/stayin-alive/devlog/1140229/stayin-alive-v110
Phase 3 (no bugs allowed):
7. Renee (12/12/25)
Renee is a CS coterm who enjoys games like Minecraft, Fortnite, and Papa’s Freezeria. She is motivated by emotional stakes and loves animals, which made her an especially good test for whether our game and whether the system could create tension and emergent strategy without heavy hand-holding.
During the tutorial, Renee reacted positively to the playful language, sound design, and the overarching “religious mission” framing, which helped her feel invested in saving the population. She liked the eating sound and described the language as “fun.” She wanted to reread tutorial text and felt frustrated when she couldn’t go back. Several core mechanics were unclear early on: grass growth and regeneration were not explained, which caused her to die during the tutorial, and depositing items one at a time felt limiting. She also disliked the word “breed”, finding it off-putting.
Once in the main game, Renee showed strong emotional reactions to predators, especially coyotes, which she repeatedly described as terrifying (she identifies with the rat!). She learned quickly to hide as soon as she heard predator audio cues and began reacting strategically to environmental signals. She noticed and commented on systemic patterns, such as realizing that building dens near rabbit areas was a bad long-term strategy and that bees dramatically improved grass growth. Her major “aha” moment came when she decided to move her entire operation near a bee tree and away from predators, after which the game “clicked” and felt much more manageable.
Some issues during this playtest were that Renee often forgot about children and population management, which led to sudden losses that felt unfair due to fading logs and subtle feedback. She strongly disliked stressful audio cues like the “inventory full” sound, but loved the physics-based item movement and environmental storytelling. Despite frustration, crashes, and steep difficulty, she entered a clear flow state(!!) stayed invested for long stretches, and repeatedly expressed that she wanted to keep playing and win.
Key takeaways:
- Strong emotional engagement and empathy for the rats
- Strategy emerged organically through environmental understanding (bees, predators, rabbits)
- Onboarding and early system explanations (grass, children, MVP goal) were insufficient
- Feedback around deaths, logs, and audio stress needed refinement
- Other than the memory crash issues after a long time in the game, there were no glaring issues
8. Renee (Part 2) (12/12/25)
In this follow-up session, Renee skipped the tutorial and jumped straight into gameplay, which was important for us to understand replayability and mastery over time. She immediately evaluated her spawn quality, reacted strongly to hawks and coyotes, and began planning around environmental risk.
Renee demonstrated growing system understanding, intentionally saving seeds, checking resources before traveling, and using teleportation to escape danger. She set self-imposed goals like scouting better locations, expanding safely, and placing dens near large grass patches. While she still occasionally forgot mechanics like hiding in bushes, we found that she was checking the knowledge menu less frequently, which showed how players were building understanding of the environment and mastery over time.
She was still highly emotionally invested which was shown by how she celebrated spring, reacted strongly to winter’s impact, and expressed real distress when her children were eaten. While she focused less on the MVP objective and more on “expanding her empire,” this still reflected successful engagement that we wanted!
Some bugs and crashes persisted, but Renee explicitly stated she still wanted to keep playing and win, even restarting multiple times. By the end, she had a clearer understanding of seasons, predator behavior, and optimal den placement, and felt the difficulty curve was challenging but compelling.
Major takeaways:
- Clear evidence of emergent strategy and learning across runs
- Strong replay motivation despite frustration and crashes
- Environmental systems (seasons, predators, bees) were increasingly legible
- MVP goal and progression framing still needed reinforcement
Phase 3 Changes:
This phase was about stress testing and making sure there were no game-breaking bugs for all situations. It involved a lot of internal playtesting, and then two playtests with our playtester Renee to confirm that everything worked and figure out any last needs for polish. The changes made are listed here:
- Re-increased Winter penalty for grass (50% death rate -> 70% death rate)
- Added tree description in Knowledge menu
- Knowledge menu buttons now scale to full width
- Fixed bug where pressing ‘E’ outside of den would transfer items
- Fixed bug where time would start immediately when the scene loads
- Fixed bug where grass patch would prevent people from building dens on top of it
- Reduced chance of memory issue happening by severely optimizing the game
- Re-stylized UI for hunger bar and season bar
Polish/Fidelity
UI Changes
To add polish to our existing game, we drew all of the UI assets (buttons, scroll bar, keys, etc.) to keep a consistent style. For example:
[Image of the drawn buttons.]
[Image of the drawn “E” key letting players know they can interact with this grass.]
Based on playtesters needing more information about the world beyond the tutorial and wanting to promote discovery fun, we also added a new mechanic with the “knowledge” menu that lets players know information about items and environment objects. This new mechanic is visually consistent with our existing elements.
[Image of the new knowledge menu.]
To attract new players outside of class to our browser game, we also added polish to the itch.io webpage, adding a drawn banner, background, story background, and thumbnails.
[Image of itch webpage.]
[Image of our new gif thumbnail that appears when “stayin alive” is searched.]
After doing research on how to push games on itch, we also added game tags and a dev log. These help people find our game on public feeds. For example, when a new dev update is posted, it gets pushed to the top of the dev feed that people check to find new games that are actively being updated.
[Example of one our earlier devlogs, logging all the changes for that iteration.]
[Our game tags, carefully chosen to represent our game with the most popular tags.]
The polish that we added was useful, as a random stranger on Itch found our game and thought it was worth playing!:
[Comment from a random stranger who found our game on itch.io.]
After checking itch analytics, we also found that quite a few people were randomly finding our game outside of direct links. The image below shows the unique user visits from each link. As you can see, some people were visiting from the dev log, so this was successful in getting some new players!
[Image of itch.io analytics for how users are getting to our game as of 12/11/25.]
Aspect Ratio:
This was a major coding task, the game now works for different screen sizes (and namely Butch’s wacky screen size 16×10).
Other Experimentation:
While adding features, we went back and thought about all of our design decisions thus far. One change we experimented with was a brown floor to make the interactable items more clear and differentiated from the ground, however this changed the vibe of our game more than we expected and ultimately we pivoted from this “desert” style, though different environments could be something explore in the future.
[Image of the rat in the southern california desert.]
In case you absolutely love our game and can’t get enough of it. We’ve also created these Slack emotes that are globally available on all Stanford Slack workspaces! This was to help foster the community of people love Stayin’ Alive! iykyk 🙂
[Image of Stayin’ Alive slack emotes.]
Onboarding
Initially, we thought of having our complex system game by a figure-it-out-as-you-go sort of game, similar to the exploration in Minecraft or Don’t Starve Together. However, we found that our playtesters generally found this confusing, so we started to optimize for explaining foundational mechanics in a brief tutorial before allowing the player to go off and learn on their own. Although most of this tutorial was already implemented before our refinement process, we spent some time adding new foundational mechanics to this tutorial such as the explanation of worker logs and the Knowledge menu (see next section). We also revamped the looks of various buttons and changed the sizing of different blockers for clarity.
[Image of tutorial explanation of workers.]
Additionally, we figured that there were more things that could be better explained throughout the game. We found that many players were left guessing at various mechanics, such as the spawn behavior of worms or the purpose of rabbits in the ecosystem. Because one of the main types of fun we wanted to evoke in Stayin’ Alive is discovery fun, we wanted to incentivize leaving the den to explore more exciting areas of the world. Thus, we created a Knowledge menu that is accessible at any point in the game, which explains the purpose of different animal/interactable/world behaviors once they are discovered. According to Nielsen’s heuristics, this satisfies H10: Help and Documentation, allowing the player to validate their assumptions by viewing a reference at any time.
[Image of the explanation of rat knowledge accumulation in the tutorial.]
Through playtesting, most playthroughs of the tutorial required little to no moderator intervention. In some instances, we even noticed that playtesters who paid little attention to the tutorial (Brian or Kevin N) could understand simple game mechanics based on exploration and button cues (ex: “E to close”). Note that although this is the case, skipping the tutorial is not a behavior that we are expecting from our game’s target audience (we expect our players to have a basic interest in understanding the game). However, to circumvent players accidentally missing portions of the tutorial, we have an overt popup to skip the game on the title screen, as well as “OK” buttons that must be pressed to progress the tutorial. One playtester (Kevin L.) noted that this was annoying, but it was an intentional feature designed to block user flow until the mechanic is understood.
The knowledge menu was also reasonably well-received as an optional source of information. Brian claimed that he forgot about the knowledge menu (at 34:43) in his playtest, but this did not overall impact his experience and development of a strategy throughout the game. On the other hand, Kevin N. spent time scrolling through the knowledge menu during a point in his playthrough in an attempt to learn more about seasons (at 25:57). These two experiences were great in showing that it was successful as an optional feature for players to clear up confusion.
Future Thoughts
Given more time, we’d love to add more elements (that are out of our current scope of priority for this project) for long-term replayability to the game. Such as:
- achievements like “alone in the world” or “plucked all the grass” that make exploration more engaging
- leaderboards and high scores to encourage players to try and optimize their resource management to the fullest
Acknowledgements
Thank you to all of our awesome playtesters for their time, honesty, and feedback! Sam, Jackson, the Kevins, Brian, Renee, and everyone involved in making our P3 iterations better: Ryan, Brydie, Madison, Butch, Amelia, Abram, Amaru, Evelyn, Daniel, Ari, Angela! <3
And super special thank you to the CS377G teaching team for all the guidance throughout the development process this quarter!
