Night Haul

The idea

A hauling game where the interesting choices happen before the drive. Which load, which route, how much risk you're willing to carry for the fee — and then living with it in the dark, where you can see the shape of a problem long before you can see what it is.

The dark isn't a mood. It's the mechanic that makes every other decision cost something.

Silhouette noir

The art direction is a constraint dressed as a style. Objects read as shapes against light rather than as textured models: headlamps, signage, a window, a fire. It gives the game a look that's genuinely its own, and it means a solo developer isn't competing on texture budgets he doesn't have.

Lighting does the work that art assets normally do. What you can see is a design lever — pull light away and tension appears for free; put a lamp somewhere and you've made a landmark.

Routes

Choice with teeth

Faster is riskier, and the map is a set of trade-offs rather than a shortest path.

Cargo

Load as constraint

What you carry changes how you drive, what you can afford to lose, and who's interested in it.

Risk

Pressure, not jumpscares

The threat model is being seen and being slow, built out of light and time rather than combat.

The engine call

It started in Unreal 5. Unreal is superb at photoreal fidelity, which is precisely the thing this game had decided not to compete on — and the cost of that fidelity showed up as iteration time, build weight and a lot of machinery I was fighting past rather than using.

I moved it to Godot 4. Smaller, faster to iterate, and its lighting and 2D/3D hybrid tooling sit much closer to what silhouette noir actually needs. It cost real weeks to switch.

I include it here on purpose, because it's the kind of decision I'd want to see from someone I was hiring: I noticed the tool was wrong for the design, checked that I wasn't just avoiding a hard part, and paid the switching cost while it was still cheap rather than defending the original choice for another year.

Where it is

Pre-production is done: the design document, the core loop, the art direction and the vertical-slice targets are written down and settled. Current work is systems — routes, cargo, risk, and the night itself — in Godot 4, in a repo with the design doc versioned next to the code.

What it demonstrates

I finish the thinking before I write the code. Concept source, design doc, then systems.
I can change my mind expensively. The engine pivot was the right call and I made it early.
I design inside constraints. Solo scope, turned into a look nobody else has.