New Release: Underground Football League!

Ever developed a game where you’ve moved goal-posts mid-way through? And as development slowly inches forward, you’re saying in your head “it’s so close to being released!”? Yet new features keep creeping in, a deluge of minor details and bugs need addressing and this drags on for weeks and weeks?

Well, this is the TL;DR of making Underground Football League, my 26th browser game.

Tech

Made in Unity and C#, with sprinkles of JavaScript to finish it off.

Take 1

The initial concepts for game, UI and characters were done in 2020, as part of a “concepts for the next 7 games I want to make in the future” effort. The basic idea was to make a physics puzzle with a football theme.

Take 2

Eventually, at the end of 2021, I’ve picked this up as my next game to develop. In the initial concepts, I’ve done DragonBones animations for the characters, however, as I planned to learn more of Unity native features, I’ve abandoned it in favour of Unity’s native IK and animations system.

I’ve also chose Unity 2D package, as well as abandoning my custom localisation solution used in Last Ball, Bumper To Bumper and Snake Island for Unity built-in localisation system. I’ve kept the same architecture though, based on Zenject and signals.

A playable version was ready in about 3 months, but as in April 2022 I’ve started work in the mainstream industry, the project was postponed indefinitely.

I assessed, back then, that the game was in an advanced development phase. How little did I know.

Take 3

At the end of 2025, I’ve decided to take another shot at finishing it. Four years have passed and I eventually had an idea of what direction to take with the game. The problem was the existing code and assets were only partially useful for what I had in mind.

I’ve decided to give a more emphasis to the puzzle part of the game and less to the skill/physics. Problem was, there were not enough puzzle elements implemented in the game and it took a few weeks to come up with them (timed doors, buttons/switches, pinball flippers, elevators, saws and so on).

Each level has 4 “modes”: one, simply finishing it. Second, finish it perfectly to get the maximum rewards. Third, to collect the diamonds. Fourth, to finish it in the fastest time possible (do a speed-run). However, this impacted the level design massively, toward the point that each level took 1-2 days to be produced and tested.

Character customisation (the in-game kit shop) was also important and while the original code supported it to a certain extent, it didn’t support volume – lots of items and color variations. This lead to a massive refactoring of both code and kit assets spanning multiple weeks, including coding an internal Unity tool to populate the shop, as manually editing the shop ScriptableObject became untenable.

And, as assets grew in volume, and the built-in Unity search tool is not fit for anything but small projects, this spawned a side-project that I wanted to do for a while: a Unity favorite tool.

I didn’t realise how much work implementing an immersive audio would require and how unfit what the prototype had in terms of audio programming. Took almost two weeks of familiarising with Unity audio mixers, retro-fitting all the physics elements with audio support and searching for suitable sound effects. It’s also probably the first time as a solo game developer when I felt I would have needed custom sound effects made instead of spending hours searching suitable, very specific sfx, e.g. “football hits human, with or without clothes”…

Lastly, monetization, and having exclusive cosmetics felt like a good match. With Coil closing and Web Monetization not being popular enough (and in its current state, rather limiting for the developers), and still wanting to keep my games ad-free, I decided to take another route. After careful consideration, seemed like semi-fungible tokens (SFTs), which are somehow similar to NFTs, fit my game a bit better. Eventually, this spawned another side-project, integrating web3 features into Unity.

In total, I’d say this whole adventure took spanned over the course of about 10 months, bit longer than I originally envisioned.

Take 4

But wait, this game is far from being finished…

Now comes the boring part of doing promotion.

And the TODO list has at least 100 more items, from minor code refactoring to more game modes, from additional cosmetics to puzzle elements. But the future of it depends, of course, on how popular it becomes.

I’d also be tempted to make a more in-depth write on the level design process. Hopefully, soon.