Run! Jump! Dodge! Blast those Ninjas!
Designed and built by Rafael Pavon
Unreal Fellowship: Games 2025
Enter Ninja Ambush, a fast-paced runner meets slow-motion shooter, built while learning Unreal Engine from scratch.
The Unreal Fellowship is an intensive four-week training programme run by Epic Games, designed to teach professionals how to build games using Unreal Engine, one of the most powerful tools in the industry. The challenge? Learn the engine from scratch and create a complete game from concept to playable build in just four weeks, based only on a single prompt: "Arcade."
To do that, you need to wear every hat in game development, from game designer to coder, tech artist, art director, producer and QA tester. I joined with zero experience. No idea how Blueprints worked, how to handle visual effects, or where to even start.
That’s where Ninja Ambush started – not with a grand plan, but with a blank screen and a ticking clock.
I set out to create a love letter to the arcades of my childhood, a game that felt fast, flashy, and impossible to put down. My guiding lights were Sayonara Wild Hearts for its jaw-dropping visuals, and Rez for its rhythm-first approach to shooting. I didn’t just want to make a game; I wanted to learn coding to trigger a feeling.
The core loop came fast: sprint through a neon-lit city collecting time shards, then leap into slow motion to unleash a wave of shurikens. It's a game about timing, flow and precision, a heartbeat that goes run, slow, aim, strike. Miss a beat, and you're out.
From the start, the goal was simple: make something so fun, so punchy, you can’t resist one more go. That arcade spirit – easy to learn, hard to quit – guided everything.
Week one was all tutorials and headaches. I was learning Unreal on the fly. Piecing together runners, triggers, collisions, desperately trying to make the thing move. And then, something clicked. I prototyped an "ambush moment": Initially, I designed the ambush as a turn-based mechanic where you chose your attack. It worked… technically. But it felt wrong. The arcade rhythm was not there.
So I made a snap decision: no more turn-based. Just a slow motion shooter. Pure flow. No menus, no interruptions. Just that split second of control before the chaos resumes. That pivot unlocked the soul of the game.
To raise the stakes, I introduced a risk/reward system: grab more time shards during your run, and you earn more slow-mo seconds later. Push your luck, or play it safe. It gave every choice weight.
Week two was all about iteration – targeting systems, aim assist, custom projectiles, visual effects, material tweaks, camera behaviour. Each element helped refine the core loop and strengthen that feeling I was looking for.
Week three was all about the vibes. I leaned hard into neon noir – glowing fog, pulsing outlines, abstract geometry. Unlit materials gave everything that arcade glow. Post-process volumes cranked up the drama. I wanted it to look like you’d stepped into a fever dream of a Tokyo back alley inside a 1980s cabinet.
And somehow, in four weeks, I finished a game. A proper one. It’s tight, snappy, and always teasing you to beat your last run. The kind of thing you can pick up in seconds, but want to master for hours.
