Broker
Breaker.
A wave-based arcade action game set in Wall Street chaos — designed, coded, and illustrated from scratch by a 4-person team.
We were four people who had never opened Game Maker before. The brief was simple: build a game. No template, no existing codebase. We started where most good design starts — not at a screen, but around a table with sticky notes and a lot of ideas that weren't going to work.
02 — The Premise
What is Broker Breaker?
The concept, the player, and the core design challenge.
Out of 20+ concepts — debt simulators, field trips, city racers — Wall Street won unanimously. It was the one idea that made everyone in the room excited. It had natural game logic built in, a visual world everyone could picture, and something harder to manufacture: nostalgia. The chaos of an arcade game set against the chaos of finance just made sense.
Game Concept
Wall Street chaos.
Wave by wave.
A wave-based action game set on a pixelated NYC street. Battle escalating enemy types, collect XP, upgrade weapons, and survive long enough to face the final boss.
Target Player
Casual gamer.
Ages 8+.
Someone who enjoys arcade-style action but isn't looking for a punishing skill ceiling. Pick up and play — no prior knowledge, no tutorial required.
Design Challenge
Accessible,
but not trivial.
The tension we designed around — easy to pick up, hard to master. Every system had to stay readable for a new player while rewarding skill over time.
The concept came fast. The execution didn't. Building every sprite from scratch with no prior pixel art experience, while simultaneously learning Game Maker's scripting language, meant that the gap between what we imagined and what we could build was humbling. This is the honest version of how we got from a sticky note to a shipped game.
03 — Process
From sticky notes to shipped game.
6 steps. Every one documented.
Before we touched a single sprite or line of code, we needed to answer one question: what does the player actually do, every 10 seconds? The game loop was the skeleton everything else hung off. Get it wrong and no amount of good art or sound design would save it. We mapped it on paper first — kills feed XP, XP feeds weapons, weapons feed survival, survival feeds the boss. Every system had to justify its place in the loop or it got cut.
04 — Game Loop
Every system feeds back into the loop.
Weapons, XP, difficulty — nothing is isolated.
The loop was designed so that every action — kill, collect, upgrade — reinforces the next. No dead ends.
We didn't want enemies to just be obstacles — we wanted each one to teach the player a new behavior. A game that introduces everything at once is overwhelming. A game where each encounter layers in one new rule keeps the player learning without realising it. So we designed four enemy types, each with a distinct mechanical purpose, not just a different HP value.
05 — Enemy Design
Each enemy teaches the player something new.
Not just difficulty increments — designed behaviors.
Here's the thing about Broker Breaker — it looks easy. It isn't. When we recorded our own playthrough for the submission, it took the team 5–6 tries to actually win. The game is deceptively calm in the first 100 kills. Then suddenly you're surrounded by Runners and Tanks at the same time and everything falls apart. Getting that ramp right — gradual enough to not frustrate, steep enough to stay exciting — was the hardest design problem we solved. We did it with probability, not hard cutoffs.
06 — Difficulty Progression
Invisible design. Visible feeling.
Players never see these numbers — they just feel the shift.
| Stage | Kill Range | Enemy Mix | Type 1 | Type 2 | Type 3 |
|---|---|---|---|---|---|
| Stage 1 | 0 – 99 | 100% | — | — | |
| Stage 2 | 100 – 249 | 90% | 10% | — | |
| Stage 3 | 250 – 449 | 80% | 20% | — | |
| Stage 4 | 450 – 699 | 70% | 22% | 8% | |
| Stage 5 | 700 – 999 | 60% | 25% | 15% | |
| Stage 6 | 1000+ | — | — | BOSS |
As the enemies got harder, we needed the player to feel like they were keeping up. The weapon system was our answer to that. Four weapons, each designed to counter a specific threat. But we didn't want to just hand them over — weapons had to be earned through XP, which came from the rats, which required you to survive long enough to collect them. The whole system was one connected loop of risk and reward.
07 — Weapon System
Four weapons. Each with a job.
Visual language reinforces mechanical purpose.
We thought the interface was fine. Then someone else played it. Three things we assumed were obvious turned out to be invisible — which weapon was active, when the area attack was cooling down, and that the rats were even collectible at all. None of these were guessable from watching us play internally. It took fresh eyes to expose them, and each fix was a direct response to a specific moment of player confusion.
08 — UI Evolution
Every change came from observation.
Three problems. Three fixes. Each one from a real playtest moment.
Playtesting is humbling. You spend weeks building something, you know exactly how it works, and then you watch someone play it and do something you never expected. Three sessions. Three people. Each one broke something different. The rule we held ourselves to: every piece of feedback had to result in a specific change, not a conversation about whether the feedback was valid.
09 — Playtesting
Three sessions. Real iteration.
Observation → hypothesis → change. Every time.
We shipped a game we're proud of. But shipping isn't finishing — it's the moment you can finally see clearly what's left to do. The things that worked, we'd double down on. The things that didn't, we know exactly why. Here's the honest breakdown.
10 — Reflection
What worked. What's next.
Honest assessment of the build.
Survive waves of Wall Street enemies. Kill enemies to collect XP. Upgrade your weapons. Reach 1000+ kills. Clear the final enemies — and the boss spawns. Take it down to win.
Fair warning: it looks easier than it is. The team that built this took 5–6 tries to beat their own game.