I almost deleted Color Rush. Not metaphorically — I had the terminal open, cursor blinking next to the delete command, finger hovering over Enter. The game was three days late, the mechanic felt clunky, and I was so frustrated I could feel it in my jaw. I'd been clenching my teeth for two hours without realizing.
What stopped me was a notification sound from Slack. Max had dropped a link to a forum thread where someone was complaining that browser games today "all look the same." There was a comment that read: "I miss when games were weird and rough and you could tell someone actually made them." I read it five times. Then I looked at Color Rush — this ugly, unfinished, broken game with falling color tiles and a match mechanic that felt janky — and thought: maybe rough isn't the same as bad.
The Build That Should Have Failed
Color Rush was supposed to be a simple matching game. Three days, start to finish. On Day 2, I realized the match detection algorithm was fundamentally wrong for the visual layout I'd designed. The tiles were falling too fast, the color palette clashed, and the scoring system was recording negative values — literally rewarding failure. I spent eight hours trying to fix it and made it worse each time. By midnight, the game was broken in new and creative ways I hadn't even considered possible.
I have a habit of rage-deleting failed builds. There's a folder on my laptop called "graveyard" with seventeen dead game directories. Each one represents a moment where I decided the concept wasnt salvageable. And honestly, most of those were the right call. But Color Rush was different — not because it worked better, but because the core feeling of playing it was there even when the code wasnt.
The next morning I rebuilt the match system from scratch in three hours. The new version was simpler, faster, and honestly uglier than what I'd spent two days on. But it worked. I played one round. Then another. Then I called Max and said "you need to see this" like a kid who found something cool in the backyard.
What Broken Taught Me About Finished
The original Color Rush build had 847 lines of code. The version that shipped has 412. Cutting the code in half wasnt efficiency — it was surrender. I gave up on making the game elegant and settled for making it fun. The irony is painful: my best work happens when I stop caring about whether the code is clean.
I tested the game on five people who'd never seen it. Three of them said the color palette was "ugly." All five played for more than ten minutes. One played for forty-seven minutes and emailed me the next day asking when "Part 2" was coming out. That email sits in my inbox as a permanent reminder that players care about how a game feels, not how its code reads.
I've since gone back to the graveyard folder and resurrected two other projects I'd written off. One of them will probably never ship — it's genuinely unfun. But the other one... let's just say I'm glad I got into the habit of checking my anger before hitting that delete key.