An AI tycoon game maker for the simulation underneath

Every tycoon game is a set of rates. Customers arrive at some frequency, spend some amount, have some patience, and leave a rating that changes the arrival rate. Staff cost money per hour and reduce queue time. Upgrades cost capital and change one of the rates. That is the game, and everything else is presentation.

The presentation is not optional though, because the reason people play a tycoon game rather than open a spreadsheet is that they want to see the thing they built. A park, a hospital, a railway, a restaurant, a space station. Watching the little figures move is the reward the numbers pay out.

Stage Engine splits the work along that line. You build the place. Director 1 writes the simulation that makes it behave.

You lay out the site directly in the builder: terrain, paths, buildings, props, lighting, weather. This is hand work and it should be, because the layout of a tycoon site is a design decision with consequences. A long path past the shops is a revenue decision. A bottleneck at the entrance is a difficulty decision.

You can build a fixed site and treat the game as a scenario, or build modular pieces and have Director 1 write the placement system so the player constructs it themselves. Both are ordinary requests. Neither is a mode you switch into.

Because the world tooling handles terrain, water and weather, an outdoor tycoon game is not a stretch. A park on a hillside where rain closes the outdoor rides is a thing the underlying world supports rather than a feature you would have to fake.

Ask for customers who arrive in a daily curve with a lunch peak, queue at the nearest till, get impatient after ninety seconds and leave a review that shifts tomorrow's arrivals. Ask for staff who can be hired, assigned and who quit if unpaid. Ask for a loan system with interest. Director 1 writes the GDScript for all of it inside your project.

Then you play it, which is where the real information is. Almost every economy is wrong on the first pass. Either money is trivial by day four, or the opening hour is so tight that nothing interesting happens. You find that out by playing an hour, not by reasoning about it.

When you find it, you say so in plain terms and it gets changed. That back-and-forth is the process, not a failure of the first attempt.

Tycoon games get heavy in a specific way: a lot of simple agents, all updating, all pathing, all making decisions. A hundred visitors is unremarkable for the genre and demanding for anything running in a browser tab.

Stage Engine is a desktop app for Mac and Windows with Godot bundled inside it, and the game runs on your machine. There is no streaming and no render queue. Whatever your computer can simulate is what your game can simulate.

Building needs an internet connection because Director 1, asset generation and world data come over the network. The game you built runs without one.

Someone with a specific tycoon idea and no intention of learning GDScript. The genre is unusually well suited to this arrangement, because tycoon design is mostly judgement about rates and layout, and almost none of it requires you to have written the code that applies them.

It suits you less if you want a finished game by the end of the evening. Balancing an economy is real work and it is the part nobody can do for you, because "is this fun" is not a question with a computable answer.

Most tycoon games are careful about how you win and vague about how you lose. Then the player builds an efficient loop in hour two, the money curve goes vertical, and the rest of the game is watching a number rise with nothing to spend it on.

Designing against that is the real craft of the genre. Costs that scale, staff that get harder to keep, demand that saturates, maintenance that eats the surplus, competitors who take a share. These are all rules Director 1 can write, and all decisions only you can make.

Play until it breaks, and pay attention to exactly when. Hour two is a different problem from hour twelve, and the fix is different in kind, not only in size.