Gamedev.js · 2025 / Case study

This Game Is Unbalanced!

For Gamedev.js 2025, I wanted to join the game jam, learn Phaser and make extensive use of GitHub Copilot. At the end, I also needed a finished game.

The English home screen of This Game Is Unbalanced!.
Learning a framework and shipping a game within a game jam's deadline.

Quick to generate, not always quick to understand.

Learning Phaser with Copilot helped me experiment quickly. But generating code did not remove the need to understand what it did, or how its parts fitted together.

I reworked sections and refactored generated code. That review was part of development, even when it took longer than generating the code in the first place.

Choosing what I could actually finish.

The game jam deadline meant making choices. I cut scope so I could finish and ship a complete game on time.

The experience taught me as much about using AI as it did about game development. Moving quickly is not enough if the result becomes too hard to understand or too large to finish.

A finished game, and two first places.

The game took first place in both Innovation and Vibe Coding.

What stayed with me was the combination: learning a framework, reviewing generated code critically, and accepting a smaller version of the original idea to make it playable.