Spinning
1/4
On the Way
2/4
Don't Panic
3/4
Mutant Sea
4/4
Spinning is my MFA thesis project at SCAD (Interactive Design and Game Development) and my most recent game. It focuses on a 3D web-building system, treating the web as a deliberate construction tool—movement and building are decoupled so players can author their own routes and spaces in the world.
The prototype combines node-based placement of continuing and terminal segments in 3D, traversal on and around player-built webs, and strong visual feedback for placement, snapping, and removal. It was refined across several major versions with formal playtesting, with a custom-rigged spider character (“Jenny”) and environment and materials built in Maya, ZBrush, and the Substance suite.
Fig. 1
Spinning gameplay screenshot.
The hardest technical work was pathfinding over graphs that change as the player builds. Blueprint-only approaches proved insufficient across early versions; I moved to a custom C++ A* implementation suited to dynamic, player-authored connectivity, and AI assisted adaptation of reference code where it sped integration. Camera and usability also improved through iteration, including a co-design survey.
Special thanks to my thesis committee Dr. Potanin, Professor Cookson, and Professor Cody, and to all the playtesters and colleagues who took the time to share their thoughts.
Fig. 1
On the Way screenshot.
On the Way is a soapbox derby racer where players compete against ghost versions of their previous runs to unlock cars with unique abilities. I served as one of two Gameplay and Technical Designers on our team working to bring the physics-driven racing mechanics to life.
I initially prototyped the core movement and the ghost recording system, to ensure that previous attempts meaningfully impacted the current course layout to fulfill the “loop” mechanic. My efforts were focused on the feel of gravity-driven acceleration, the car selection look up table, and ghost playback which was done by storing prior vehicles and times and animating the vehicles movements along the spline. I created and populated a comprehensive LUT, including separate mesh data for car bodies and wheels, per car player animation data and player offset, offset data to assemble to "assemble" the vehicles and statistics like speed which were used in wheel animations. This table in addition to being critical for car selection functionality, enabled more convenient tuning of different cars to create a different "feeling" when driving each; William refined movement and developed car abilities and worked on the car selection menu.
Play on Itch.io: On the Way (itch.io)
This was a 10 week long collaborative project during which we created a third-person, action firefighting video game. The project can be downloaded and played on itch.io.
Fig. 1
Early fire system demonstration.
I primarily worked on the implementations of the menus, several interactable objects within the game, and the fire spreading system. We used a cube grid based design for the spreading with random odds to spread to adjacent cubes for a randomly selected space. The basic functionallty of this we got working pretty quickly.
Fig. 2
Spoofed physics based hose bouncing.
In addition to doors, collectables, he hose connecting the firefighter's pack to his sprayer and the enemy's ability to start fire's in the environment, my primary responsibility was to implement the fire spreading system.
Fig. 3
Demonstration of attempt to bake the building structure into data which could be accessed by the fire.
There were many issues in making the fire spread work. We wanted the fire to behave in a particular way with respect to the structure of the building, so we attempted various ways to give the fire the information of the building structure. It turned out that the memory accesses from the data structure which held the information about the structure of the building were taking too long, so we switched to a real time approach. This enabled us to have the fire "stick" to the walls rather than spreading through the air.
Fig. 4
Demonstration of different patterns of "sticking" tested.
Other issues arose with some of the blueprint fuctions I wrote to manage the fire. Since this was a fire fighting game, the player needed to be able to put fires out, and also we had an enemy which needed to be able to start fires and in order to do that they needed to interact with the fire grid system. As an example, one helper function I hastily wrote, WorldLocationToGridLocation, was found through debug trace to be a performance drag and so I analyzed the function and rewrote it.
Fig. 5
Analysis of improvement to WorldLocationToGridLocation function.
Another thing I worked a little bit on for this proejct was the civilian NPC carrying system, which used a character that I had originally created for another class, but for which I needed to adjust the mesh for this project.
Fig. 6
Screenshot of player carrying NPC from final game.
For much of the project I also served as a producer, hosting stand up meetings and checking in with other students about progress. One component of that responsibility was helping with the version control for this project for which we used SubVersion.
Fig. 7
Screenshot of gameplay.
Special thanks to Professor Cookson for his guidance throughout this project.
Fig. 1
Image of dice rolled in Mutatant Sea.
Fig. 1
Mutant Sea full game board during play.
Mutant Sea is a board game where players play as mutant sea creatures who gather resources to spawn new mutants and harvest nuclear waste to mutate existing mutants into larger creatures. Players have ten rounds to claim the most land tiles with their mutants and/or kill other enemy mutants thus eliminating those players from the game to win.
Fig. 2
Mutant Sea first concept playtest.
During each round each player takes a turn drawing cards and using their mutants’ movements and actions. Land tiles can be claimed either directly for resource points or by defeating enemy mutants and claiming tiles as a reward. This game was created by a collaborative group, evolving through multiple iterations.
Fig. 3
Intermediate versions of the game during playtesting.
The game uses six-fold symmetry on the game map and characters and game pieces were designed and modeled in 3D. A rulebook, task cards, resource cards and mutant cards were designed, and the board and other materials were printed in color.
Fig. 4
Table showing expected damage for attacks with a given attack and defense.
The characters’ statistics and abilities for offensive and defensive capabilities were initially specified but then were modified and balanced based on feedback from game play in order to make play fair. Rankings for prototype and mutated characters and combat were modeled based on statistics from a heat map showing expected damage from any combination of possible attack rolls and defense rolls. These statistics were derived from a table generated by a Python program.
Fig. 5
Mutant card versions.
The statistics set ranges for possible power and defense statistics. They also impacted how health values were set. The game was play tested with a variety of stats for the mutants producing the final prototype, a fourth iteration. This game won as the Best Tabletop Game at Entelechy 2024, a SCAD game design competition.
Fig. 7
Game board set for play and Entelechy Award shared with co designers Samm Davis, Yifan "Evan" Lu, and Yinhe "Justin" Wang.
Early and intermediate versions of this game used art which was generated by AI. The final version of the game includes no AI generated art.
Spinning
1/4
On the Way
2/4
Don't Panic
3/4
Mutant Sea
4/4
