For thirteen chapters I’ve argued that theme park queues are an engineering problem, not bad luck. Explaining it wasn’t enough. So I made it playable.
Also available in Spanish▶ Open fullscreen ↗ — free, runs in the browser, no signup.
–>Why a game
Across this series I’ve designed, on paper, a theme park for Asturias. Zones, attractions, capacities, layout, numbers. And in every chapter the same thing kept happening: I could write «if installed capacity doesn’t cover attendance, satisfaction collapses», and the sentence read like a truism.
It isn’t. It’s a truism right up until it’s your problem — and then you discover it never arrives announced. It arrives as a ride breakdown on a Saturday at 13:40, with the car park at 98% and a parade about to cut the central avenue in half.
An article hands you the conclusion. A simulator makes you commit the mistake yourself. That’s why there’s a game.
What’s underneath
It isn’t a building tycoon. You don’t build anything: the park already exists and you run it through the day, in real time. Underneath are four models, none of them decorative.
1. Real capacity, not catalogue capacity
Every attraction has a THRC (theoretical hourly ride capacity) and a train count:
| Attraction | Type | THRC at full trains |
|---|---|---|
| Sobre Asturies | Flying Theater | 1,350 |
| El Pozu | KUKA dark ride | 1,200 |
| El Vendaval | LSM coaster | 1,000 |
| O Sella / La Güestia / 4D Cinema | Log flume / Dark ride / 4D | 900 |
| La Ferrería | Spinning dark ride | 800 |
| El Esbardu | Drop tower | 700 |
Total: 7,750 pph with everything running full. But you don’t start full — you start on half trains, because every train draws from a pool of 18 crews that also have to staff walkways, the gate and food service.
Capacity is not a spec-sheet number. It’s a budget decision you make every single morning.
2. Walkways are a graph, not scenery
The park is a network of nodes and paths. Each visitor group routes itself with a cost of length × (1 + saturation). In plain terms: crowds avoid full paths. Close a walkway and the flow doesn’t vanish — it redistributes, sometimes exactly where you didn’t want it.
3. Demand runs away from queues
Each group weights its destination by popularity / (1 + wait/10). They don’t always head for the most popular ride: they head for the best ratio between what they fancy and what they’d have to queue for.
That single term is the difference between a toy model and one that behaves like a park.
4. The day can be lost
There’s a park pulse that drops with average waits, walkway saturation and a full car park. Leave it on the floor long enough and the day collapses, and you take the reputation hit. Reputation decides how many people show up tomorrow.
Everything you can do in the game is what a duty manager can actually do on a Saturday: open and close paths, move crews, add or drop trains, dispatch mechanics, and choose the queue type. There is no fix-it button.
What to watch while you play
If you’re only going to give it ten minutes, watch these three things.
The first bottleneck is never a ride. It’s the gate, or the car park. Almost always. The people who never got in don’t show up in any queue — but they left all the same. This is the single most under-modelled thing in park discourse: everyone argues about ride throughput and nobody counts the people who turned around in the entrance plaza.
Adding a train isn’t free. It comes out of the same pool as your walkway crews. Max everything out and you’ll watch the park choke on its own paths while the rides sit comfortably under-used. Capacity you can’t feed is capacity you don’t have.
The parade. At 16:00 it cuts the central avenue. Every routing assumption you had in your head stops being true for a few minutes. It’s the cleanest demonstration of why a show is never just a show: it’s a flow decision with a footprint.
And one lesson you only really learn by losing: virtual queue and fast pass do not create capacity. They move the wait somewhere else. Sometimes that’s enough. Sometimes you’ve just handed the problem to the next hour, with interest.
This is a prototype, and I mean that
It’s playable end to end — a full season, with a shareholders’ meeting, investment decisions and a closing verdict — but the balance is raw. Investment costs, breakdown frequency and collapse difficulty are set by feel and tuned by me, who is the worst possible judge of whether this is fun.
That’s precisely why I’m publishing it now instead of in three months.
What I’m asking for
If you play it, tell me in the comments:
- Where you got lost. Which button you didn’t understand, which number you couldn’t read.
- Where you got bored. The exact minute, if you remember it.
- What felt unfair. If something sank your day and there was nothing you could have done, that’s a design failure on my side, not a skill failure on yours.
- And if you work in the industry: which number looks wrong to you. The THRCs, the staffing ratios, the repair times. I would much rather be corrected than stay wrong.
The game runs in English and Spanish — the language selector is at the top right of the start screen.
Ride Intelligence — theme park analysis from an engineering standpoint. Medium · Patreon · The book