Building Forge Rating

What it is

Forge Rating is a browser draft game built around legendary characters from comics, fantasy, anime, and more. You spin universes, fill roles, build chemistry, and chase a Forge Rating after five picks. It looks casual. Underneath, I wanted five picks to feel like building a team, not trivia with a scoreboard.

How I built it

I’m not a career game developer. I built this by experimenting in Cursor, putting something friends could open in a browser, and learning from what they told me after they played.

One mode shows the stats. Another hides them. Same draft, different kind of thinking. That split taught me more than adding extra modes. People who like spreadsheets and people who draft by gut can share the same chase without me building two games.

I also kept the first play easy: no account wall, open it in the browser, and start drafting right away. Sharing a result with your friends had to feel natural. Leaderboards were not an afterthought, though. They were an essential part of the chase. People come back because they want to climb, beat a friend’s score, or take another shot at the daily board. If the first two minutes aren’t fun, nothing else matters. Once they’re hooked, the leaderboard is what keeps the competition alive.

What I learned

What I took away wasn’t textbook game design. It was how to ship a browser game with AI.

I learned how to build a frontend people can open without installing anything. I learned how to invent a scoring system in code and keep it consistent so a shared result means the same thing for everyone. I learned how to make the core draft work in the browser first, then stand up a backend so leaderboards and saved teams could power the chase. I learned tools by needing them: React and TypeScript with Vite, Cloudflare Pages for hosting the site, Railway and PostgreSQL when the API and database became real. I learned how to use Cursor to turn a fuzzy “five picks should feel like a team” idea into something friends could actually click.

I also learned to put the game in front of live users early, ask what worked and what didn’t, and bring the impactful feedback back into the build. Friends and players telling me where they got stuck or what made them come back taught me more than guessing alone. The craft is listening to real people, then using Cursor to fix what doesn’t hold up.