I vibe coded (With Claude Code) a billiards/pool game using Godot 4.4 with a slightly existential adventure mode + Player vs Player and Player VS CPU arcade modes.
No one really knows what's on the other side of life. It could be anything, but I'm pretty convinced it's just you and an endless loop of playing Billiards.
In this game, enjoy Arcade mode for local CO-OP PVP or PVC play styles, or for the more daring try out Adventure mode complete with story driven existential angst while enjoying a classic game of pool. There's fun for everyone if you're willing to wait for it, but I mean what else are you going to do. This is the place between life and death. This is the place outside time and space, a place for purgatory pool. This is it; this is all, forever and eternity, hitting cue balls, scratches, and perfect pocket play.
- Art: Assets adapted from Open Game Art
- Music: Self made
- Development: A mix of my own abilities augmented by Claude Code
After taking some time to develop the concept and learn enough of Godot on my own to setup a rough outline of assets, scenes, and nodes; I was pretty much able to turn over the rest of the development cycle to Claude. Claude handled adding physics, menu options, sound, event triggers, and the framework for the adventure/story mode after building up the capabilities in the arcade mode (PVC + PVP). I was able to accomplish all of this in about 2 - 8 hour days, with about 4 extra hours to understand exporting and finishing touches. If I had stuck to just myself developing this I estimate it would have easily taken a month or longer.
Now the game is far from perfect but it does works the way I imagined it from the beginning. However a Billiards game even with 3 styles of play is relatively small in the scale of indie gaming and I'm not sure due to context size if a larger game would be possible to pull off. Claude was definitely beginning to struggle towards the end with the size of some of these files. When that occurred, I could alway kill the current session and have it review the CLAUDE.md file to refresh it's memory and get it back on track fairly quickly.
The biggest draw back I would say is that it allows the developer to reject learning. Sure, it did all the things for me, but once I published the game I certainly didn't feel like I had a deep understanding of how my game worked or if I could repeat this success, because it let me avoid learning difficult topics.
I'll also totally admit that the act of asking Claude to do something and watching it develop a plan and execute that plan over and over became it's own mini-game that felt like a slot machine/loot crate, where every pull had a chance of success or back sliding. It was addicting and I found myself having fun just asking it do more and more. Player be warned.