Enemy Design - The Onion Ghost Story
- Giorgio De Medio
- Feb 3
- 4 min read
Updated: Feb 8
The Onion Ghost was the first enemy that I had the privilege to design for Town of Zoz, a 3D RPG Action game by Studio Pixanoh. Starting from the concept art, I designed an enemy that would fit in unique ways the design space of the project, pushing the player to learn new ways to fight foes across the game, which required tight coordination with multiple departments in order to nail the vision that the director entrusted us to deliver.

Enemy Design Pillars in Town of Zoz
Enemies in Town of Zoz aren't designed to just fill the ranks, they're brought to life while being flexible for requirements on multiple fronts:
Narrative:
during production, combat implementations followed the core narrative themes of the game. Cooking and farming are pillars of the gameplay, and foes needed to convey the food aspect with their own purpose and personality
Combat:
with 50+ enemies designed, each enemy was built on the designs and challenges that came before them, and they were building towards future obstacles that would push the player to relearn again and again
Scope:
with limited resources, asking for one-off features was a red flag across development. Asking for features would imply the commitment towards those features and their evolution across the game, pushing for organic variation in content during progression.
Enemy Ideation
Brainstorm
After seeing the concept art and getting the requirements from the directors, I approached the research in Miro from two standpoints:
Internal:
what aspect of combat are missing so far? what weapons needs to shine more? what gameplay elements are leveraged in the levels that these enemies would populate? What's the main challenge of this level and what do we need to foreshadow with these enemies?
External:
how can we communicate the nature of this enemy before the player faces them? What does the onion aspect represent? What about the ghost? How can we support the visual so that we don't backstab the player during this communication?
The main idea was having a sad onion, died of rotness, never harvested, looking for friends in this ghostly state, trying to hug them, just for the target to be damaged and for the ghost to be heartbroken after the failed cuddling.
Paper Design
After trimming the ideas based off the lead's input, I started building the paper design, listing behaviour and actions in terms of AI flow, tech requirements, number of animations and general support needed from the rest of the departments.
The paper designs demanded simplicity, in order to be easily understood by each department involved.

The ghost nature of this enemy pushed me to ask the addition of dithering tech to the programmers, in order to have a way for this enemy to disappear and reappear mid-combat as an escape manouver.
They're tricksters: attacking you on their terms and escaping if you're too aggressive with them.
Coordination
The first thing when implementing an enemy was usually the frame data. Early animation blockouts were crucial to prototype a new enemy, so a lot of time was dedicated to break down the animations required, following guidelines that answered to specific needs:
How long do anticipation, strike and recovery last? What about the pre-strike anticipation and followthrough?
Which are the key poses for every specific section? Is there any reference to follow?
Which notifiers are planned for each section of the animation? When does the character rotate, attack and cancel?
How much does the character move during the animation? How much this impacts the overall spacing during combat?
Which pose conveys better the required hitbox shape that the attack is going to require?


These phases were supported by a Miro board for the breakdowns, and a Google Sheet to track the status of the implementation for animation, VFX and SFX especially.
As soon as the early blockins were ready, the prototyping in Unreal would begin, with combat designers setting up the blueprints, montages, tasks and early AI flows.
Results

The first implementation was a pretty solid one, everyone was inspired enough and the vision was clearly shared across the departments. After the animation blocking and some small iterations on attacks frame numbers, VFX and SFX took over the montages, to build upon what combat and animation already did.

In the meantime, programmers implemented character dithering, which was reused for other enemies, bosses and even in game cutscenes.
Everything was strictly tracked on Jira and Miro boards that I created to keep everything in one place, with such a big roster of enemies, bosses, weapons and companions to overview as well.
After everything fell into place, level and encounter designs included, this was how the main character played against the Onion Ghost:

Conclusion
Working on this enemy was such a blast for numerous reasons, mainly because it was a precious time of my career, and this was the enemy that kickstarted the confidence in my enemy design skills.
Aspects of its design became the root for so many other enemies and ideas across development, from archetypes that re-used attacks and abilities, to narrative sidequests implemented later on.

This was just a small insight of the process that gave birth to a lot more enemies and bosses down the line, with cross-department collaboration on each step of the way and a lot of lessons learned along that same path.

Comments