In 2019, as a self-taught hobbyist, I wrote a command line tic-tac-toe game in Python to apply data structures and algorithms lessons, and generally to decompose problems into programs. Took me many, many hours.
In 2025, I won second prize at a hackathon building a complex multi-level paperball tossing game in 3D. Took me 2 hours, a bunch of image references, steers, and testing to get it there.
In 2026, I was on a bus and on a whim decided to test Fable 5 on Claude's iOS app. It took ~30 mins and presented a game that was faithful to the themes, added fun game dynamics without express instructions, and was playable. With a single short, text-only prompt.
After a few conversations on and off the internet, I decided to apply the discipline of using a single prompt in a new repo across different model-effort pairings to test which one results in a good game.
Making a game costs tokens, money, and time. But what makes a game good measures something entirely different.
Was it fun to play? Was it sufficiently difficult that you feel accomplished in scoring that point? Was the setting immersive enough that you can escape?
I'm not a gamer (Portal once, eons ago, doesn't count). Nor a game designer. But as humans, we all enjoy, nay seek out, the accomplishment of a hard-earned victory.
While the drive to play is universal, how we play is not.
Dr. Stuart Brown, The National Institute for Play
Feel free to share whether and how you enjoyed these games by rating them. Dollars, time and tokens are objective, but your experience isn't. And the latter's what matters.