TECHNICAL DESIGN

I worked hands-on in Unreal Engine 4 throughout the project’s lifespan to deliver functional concept proofs, finished mechanics, and game balance changes.

Snippets from a Controls Design Doc I drafted at the very start of development. This remained the game's primary control scheme until the introduction of single-joystick orientation in Update 2.

CHARACTER MOVEMENT

As a physics-based party game, our challenge was to craft fun and intuitive movement for the Broombot while ensuring balanced collision interactions between players and level elements.


During our concepting phase, I drafted a series of collision reaction diagrams depicting different impact scenarios and their results based on factors such as impulse location and velocity.


After making adjustments to my designs based on team discussion, I collaborated with our lead programmer to develop an early movement prototype in Unreal Engine based on my diagrams, using primitive geometry as a visual placeholder.


We used this prototype to tweak variables such as velocity, acceleration, turn speed, and camera distance. This test map also allowed us to determine the Broombot’s scale and proportions, which I passed along to our 3D artists for use in developing final character designs and cosmetics.

The test map I created would see continual use as a playground for further implementation and testing of systems. Partway through development, for instance, we decided to introduce the Jump, Brake, and Quick-Turn abilities to increase movement utility and diversify combat further. Thanks to our Agile production strategy, iterating our game around these new systems was relatively simple, and allowed me to add more variety to my level designs.


Post-launch playtesting also revealed frequent player requests for a single-stick control scheme, which we eventually implemented as an option in our second free content update.

GAMEPLAY PROTOTYPING

As programmers continued to polish our movement and combat mechanics, I began working to set a technical foundation for each of our three core game modes.


By this point, our engineers had already established a framework for our Timed Battle and Last Bot Standing modes, which utilized the combat mechanics we had built in previous sprints. My focus, therefore, was on our game’s premier Clean of the Hill mode, which centers around dust collection rather than racking up kills.


Similar to my approach with character movement, I created diagrams to visually outline dust particle patterns across various spawning scenarios, including annotations of finer design and technical considerations. Images like the one shown above helped me more fluently communicate my ideas to the team during stand-up meetings, and provided engineers with a wealth of practical reference materials.


I additionally used my concept diagrams to create functional prototypes in-engine (as seen below), allowing for practical demonstration of my design concepts in action. These concept-proofs established a framework for further implementation by our dedicated programmers.


This video highlights the evolution of our “Clean of the Hill” mode. I kept in constant communication with our programmers as we turned practical concept proofs into fully implemented features.

My documentation-centric design approach proved valuable later in development as well. When we introduced power-ups, for instance, I suggested having them be obtained via popping special balloons that spawn throughout the map – an action that meshed well with our popping-focused combat loop.


To this end, I created a full design document with detailed specifications for how Power-Up Balloons could spawn, where they spawn in each level, and what factors to consider in their implementation (shown here).

Most recently, I was responsible for developing Oddbot mode as part of our third free content update (shown here).


Working closely with our engineers throughout the project’s lifespan helped me learn our existing frameworks inside and out, and I was able to leverage this knowledge to implement the new gamemode on my own while programmers handled higher-level tasks.

LEVEL SCRIPTING

When developing battle maps for the game, I was often required to prototype and polish modular level elements that would be utilized in a variety of different contexts. I typically began by sketching interaction diagrams and annotating potential avenues for technical implementation.


From there I’d usually move to Unreal Engine to create blueprints based on my documentation, though I sometimes passed my schematics along to other programmers for them to implement instead, such as with the “Sewer Monster” mechanic seen below.


A quick sketch I did while working the “Sewer Monster” mechanic out with another programmer during a live dev session.

The “Sewer Monster” in action. Definitely don’t go swimming here.

I designed this “Sewer Monster” mechanic as a way to dissuade players from dwelling in the upper lane of the Sewer map, as much of that area is occluded by foreground elements. The monster also doubles as a stage hazard that players can knock each other into!

Once I or another developer had created a suitable prototype and thoroughly tested the mechanic to ensure stability in all documented scenarios, I passed along design specifications to our artists for final asset creation.


I frequently utilized Unreal’s Sequencer tool as well as blueprinting to achieve the desired effect, for example with the Spiderweb mechanic in Attic (shown here).

BALANCING AND TUNING

As we entered late development, I became more involved in iterating and refining the various mechanics we had implemented. Much of this was driven by player feedback and heuristic data we gathered at showcases and festivals. Each gamemode posed unique design challenges for me to consider, particularly when trying to appeal to a wide variety of player motivations.


Clean of the Hill mode, for example, places emphasis on collecting dust over taking out opponents, which meant we had to find ways to further incentivize fighting between players. To address this, I suggested having the debris from exploded Broombots be vacuumable and contribute towards dust score, making hunting players over dust a viable strategy for players who prefer to play offensively.


My most significant balancing contributions have involved changes to movement mechanics, collisions, gamemode systems, dust spawning in Clean of the Hill mode, and power-up buffs/de-buffs such as those shown in the video below.


Playtesting data allowed me to identify problems in our game systems and develop elegant solutions to balance those issues out. This video shows how I approached issues with our Hyper-Dash and Teleport powerup.