Ethyn.net

Though I have already shown how I achieve this, I feel it is worth stressing that this style is deceptively time-consuming, and not as lazy or straightforward as it may seem.

We begin by using Microsoft Paint. It is important that we use the legacy version of this software and not the newer Windows 11 version, as there is a drastic difference in the pixelation of the paintbrush tool. With the Windows 7 (legacy) version, we are able to achieve a style that appears more cleanly hand-drawn. To create the first variant of our asset, we gather reference images from a range of sources. The beauty of this strategy is that it results in our own new imagery. By avoiding the recreation of logos, we can avoid the copyright grey areas that I have faced whenever I directly port a range of images from the web.

STEP 2

While developing the art style for the game, a number of rules emerged for us to follow in order to achieve the creative intentions we agreed upon, akin to the Simpsons “no-nos” sheets (Brooks et al., 1989–present).

- ALL ART IN THE GAME MUST NOT CONTAIN ANY SHARP RIGHT ANGLES, PERFECT CIRCLES, OR STRAIGHT LINES.
(THIS INCLUDES 3D MODELS, UI, CHARACTERS, ITEMS, AND OTHER WORLD PROPS)

- DO NOT USE MORE THAN 3 COLOURS FOR AN OBJECT (EXCLUDING OUTLINES AND BLACK/WHITE SHADES) UNLESS NECESSARY TO CONVEY THE OBJECT EFFECTIVELY

- DO NOT USE WINDOWS 11 MICROSOFT PAINT (DISRUPTS LINEWORK)

- DO NOT BEGIN DRAWING UNTIL THE CANVAS SIZE IS THE CORRECT RESOLUTION
(1280 × 720, 1920 × 1080, 2560 × 1440)
(Based on asset size in scene)

- DO NOT USE ANY OTHER BRUSH EXCEPT FOR THE PAINT BRUSH AT THICKNESS OPTION 2
(IF MORE THICKNESS IS NEEDED, REDRAW OVER THE DESIRED LINEWORK WITH THE SAME BRUSH. THIS EMPHASISES “SKETCHINESS”)

ASSETS MUST NOT BE TOO SIMPLISTIC OR APPEAR OUT OF PLACE.
 IF EXTRA DETAIL IS REQUIRED, CREATE AN INNER BORDER USING THE COLOUR DIRECTLY ABOVE / BELOW THE PRIMARY COLOUR OF YOUR CHOSEN ASSET TO EMULATE “SHADING”

E.g. if lime green is the primary colour of an object, prioritise the darker green above it to be its shaded tone as an inner border.
(SHOWN NEXT WEEK)

It feels too rough. It needs more saturated colours and more confident line art.

In this 1 day experiment, I learnt how to utilise AI-assisted scripting for pathfinding for realistic angressive crowd simulation extremely rapidly. This feels trivial now, and I yearn to explore uncharted lands for myself.

The video shows that the terminal was no longer working. When I took it home and plugged it in, it sparked, I smelt burning, and it started making an error sound with “ERROR 2” flashing all over the screen. I was so sad about this, because it was what I wanted to pivot my energy towards, just for a little bit, to have fun. So instead, I made a software that works on all platforms and is an extremely customised version of what I would have done with the terminal.

ADJUST ACTOR WITH CHAT GPT

CONTINUE TO ADJUST UNTIL DESIRABLE

GENERATE ISOMETRIC / TOPDOWN SPRITE SHEET

ADJUST SPRITE SHEET

For my first design of the player’s starting apartment, I used LEGO brick textures to efficiently convey the 3D printed and plastic nature of the space in a way that immediately made sense to the player. Unfortunately, the result was messy, unreadable, and visually incohesive with the other props in the room. The repeating studs of the brick texture created so much competing detail that the room became cluttered, and visual clutter has been shown to make it harder for people to find and recognise what matters in a scene (Rosenholtz et al., 2007).



My first attempts to solve these issues included simplifying the ground to be grey/green, moving the camera further away, and adding more non-LEGO assets to the space, in the hope of making the apartment appear more cohesive.

This didn’t solve the issues in the way I hoped.



For my final design of the space, I used the same floor texture for the walls and made them a single colour each instead of numerous colours. I think this design choice adds to the emotions I am attempting to convey with this space: bright, playful colours, yet cheap and unliveable. I also made the walls tower less over the player in the name of simplicity.

 

RESEARCH FOR INTRO OF GAME / PLOT

Outside of being a pivotal landmark for indie game development, The Binding of Isaac (McMillen & Himsl, 2011) established an emotionally gripping story before the player even started gameplay. I would like to utilise the same family dynamic tragedy to bring players into the world of MFFP efficiently.

I am choosing to do this because I believe the themes and tone of the game require the player to be shifted into a specific lens.

For this emotionally charged introduction scene, I want the player to experience escaping their abusive household as an infant. To add to the tension of the scene, I wanted to use music that was tribal, primal, slow, and trance-like, aiming to simulate an enlightening and monumental moment occurring.

The scene acts as a metaphor for modern youth and their capabilities being greatly repressed by the over-integration of neurologically malicious technology extremely early in their lives. The scene demonstrates the child receiving more information, empathy, and entertainment from the device in front of them than from their parents. Tablets are taking over the parents’ role in this dynamic. These children are the lab rats for entertainment corporations who have the single agenda of maximising profit (at the cost of our future). There is research supporting this concern. A brain imaging study of preschool-aged children found that higher screen-based media use was associated with lower white matter integrity in brain tracts that support language and emerging literacy skills (Hutton et al., 2020).

EXPERIENCE HARSH LIVING CONDITIONS

A simple baseline tutorial that effectively establishes the player’s role in the world, their desire to prevail, and the aesthetic and emotional tone, all while teaching the player how to play the game via point-and-click interaction.

As I repeat this process of fusing gameplay mechanics, world building, and targeted emotions throughout these short-term game projects, I can feel it becoming less draining, overwhelming, and painful to execute. Especially after working with Jonah, narrative-focused scenes are becoming more effortless, often presenting themselves in my mind, with connected elements and efficiency of execution coming naturally.

In other words, I feel I am strengthening my abilities in creative direction through rapid iteration, which has been a huge goal this semester.

   

The first use case I found was when I couldn’t locate an adequate texture for the floor, one which resembled a fast food restaurant or McDonald’s (wood mosaic-esque).

To solve this, I found an image of McDonald’s flooring and prompted ChatGPT (OpenAI, 2026) to turn the reference image into a texture that is tileable in a game engine. The texture above was my first result. Something I notice is that the design wasn’t transferred with perfect accurately. In my case, this actually works in my favour, as it resembles the design without stealing it directly, but it is worth mentioning when looking at the flaws of the technology.

The most effective use case I found for this scene was generating a top-down perspective of the McDonald’s front counter.

Finding existing images I could use to achieve this without image generation would have been extremely time-consuming. With only 1 prompt and the reference image above, it generated an extremely sufficient top-down asset that would’ve been unachievable even through image bashing.

Once adjusted, cleaned, post-processed, and placed in engine around other assets, I had created a scene that otherwise would’ve been impossible to build in such a short time frame. I was also able to use dramatic 2D lighting to add some consistency between the assets in the space.

I am extremely happy with my results, the software gave me full control as if I had edited it myself, while taking significantly less time than it woudlve in premiere pro for example. In addition to this, I can now use E-Sequence for all of the experimental cutscenes in the game as well as any other projects going forward. This is an overall success!

A stick RPG esque layout featuring

Radioactive beach
Poluted sands and green toxic sludge that cant be traveled without a flying car

3D Printed smart slums
Lego esque plastic government housing that is free for all but incredibly hard to get ahead

The church of mcdonalds
The first work place for the player (No qualifications required) and the only religious building in town to raise your karma stat manually.

Disney Amazon Bank
PLayers can invest, purchase property, or put money into the bank for interest.

Futuristic Casino
Standard casino with slots and blackjackgiving players a chance at fast paced wealth without working.

Futuristic Doctors Office
Heal yourself from any physical alements caught by overuse of vice items / injury.

Toyota-Nvidia Dealership
Buy / rent futuristic 

Super Mega Ultra Corp
The final work place for the player requiring maximum stats to become the CEO

Self Driving Convanience Store
A mobile convience store van driven by friendly robot sanjay1000, whenever the player leaves a building, sanjay will be within a short distance to sell players vices.

Dense Towering Apartments.
Boardring and through the map are real estate properties that can be purchased at the bank and have rent claimed from.

View of Industrial District
At the top middle of the map there is a long scenic view of the world crumbling due to over industrilisation based idiocracy.

On foot + Flying car Pedestrians
Optimised Pathfinding AI for pedestrians, with queued destinations, entering and exiting, locations, driving and parking personal vehicles, avoiding other drivers in the air, etc.

WEEK 10 REFLECTION

Going back to Rust and Spy vs. Spy reminded me that the most memorable multiplayer moments come from the people, and from mechanics that tap into very human instincts: wanting stuff, protecting your stuff, and wanting to win. That gave us a clear focus for Alien Rock Game, casual fun with friends, and made it easy to decide what mattered and what could be compromised.

The bigger lesson was how much the technology had changed. Alien Rock Game had already beaten us once. We spent months fighting netcode tutorials and gave up, and this time AI handled the netcode and we had an entire prototype in a week. The only thing we were still doing by hand was the Low Poly assets, because we were sceptical that AI could help with models and meshes. We later found out it is very capable, which shows how quickly my own assumptions about these tools keep going out of date.

IN PERSON PLAY TESTS:

Our first in-person playtest was with people who weren’t game design students. They were from the comms room, and it was really valuable having the perspective of people who don’t play this kind of game, or any games. The biggest thing that stood out was the tutorial telling players to press LMB (left mouse button). One of the players didn’t understand what that meant, so we turned all control hints into icons instead, for example a mouse with the left button highlighted. Every single player was also confused about how to lock their own doors, or whether that was even possible. Instead of adding this to the tutorial and making it more complicated than it had to be, I added a large lock in the team’s colour to all of their doors, so players automatically know their team are the only people with access. We also had to really improve the damage of raiding items/ make the walls weaker.

Outside of that, players got a lot of fun out of it. We noticed that the most fun came from the airdrop items, random variation items that really mix up the game, like a portal gun, a death wand, or a camouflage that turns you into a tree, which amazingly had all the playtesters dying of laughter. This lines up with how the brain’s dopamine system responds most strongly to rewards that are unexpected (Schultz et al., 1997).

ONLINE PLAY TESTS

Our online playtests were with people who are far more inclined to understand the mechanics of games, and their feedback was mostly small changes around balancing. One pretty big thing they noticed was that horses made the same footstep sounds as players, so whenever horses were idle around people’s bases, players got confused and thought there were other players nearby. Another bug came from someone building a base completely out of doors. They placed hundreds of doors and it destroyed the performance of the entire game, so we now have a limit of 10 doors as they eseentially act as 1 way walls. These are only a handful of things. Numerous changes came from watching our playtesters very intently.