Leo+DadMade for Leo
Scientific Models
Rung 4 of 4 · Mastery

Models, Out in the Real World

Where models stop being a school idea and start running the actual world — and where you learn to reach for the right one on demand.


Apply Each scenario hands you a real question. Pick the model someone would actually use, then reveal where it breaks.
🎧
Audio WalkthroughComing Soon
Video ExplainerComing Soon

This is where models earn their keep — running things you rely on every day, and then the real test: being handed a question and reaching for the model that fits, limits and all.

The Models That Quietly Run Your Day

That weather forecast on your phone is a computer model of the atmosphere, grinding billions of sums about pressure and temperature to predict tomorrow. Climate models do the same over decades, which is how scientists warn about warming. When a new disease appears, public-health teams run disease-spread models to predict the case curve and decide what to do. Engineers crash a car a thousand times in a simulation before they bend a single real bumper. And the humble particle model you met last year is still doing its job, explaining everything from a melting ice cube to the pressure in a tyre. None of these is the real thing. Every one of them is a stand-in we trust exactly as far as we know it holds.

The Real Skill: Reaching for the Right Model

Rung 2 went forwards — here's a model, here's how to use it. Mastery is going the other way: you're handed a question and you reach for the model that fits, then immediately name its limits. "How tall will this tree be in ten years?" — that's numbers changing over time, so reach for a mathematical model, a growth curve; its limit is that it assumes nothing unusual happens, so a drought or a chainsaw breaks it. "How will smoke spread if this building catches fire?" — too many moving parts for your head, so reach for a computer simulation; its limit is that it's only as good as the floor plan and air-flow data you feed it. "Show me how the parts of an eye fit together." — that's shape in three dimensions, so reach for a physical model; its limit is that it won't tell you anything about how the eye actually sees.

Say it plainly: mastery is two moves — match the question to a kind of model (shape → physical, flow → diagram, numbers → mathematical, too-many-parts → simulation), then say out loud where that model would break. Never hand over a model without its limits attached.

A Depth-study Thread

This is a great launch pad for a Year 8 depth study (the scope sets aside time for one): build and test your own simple model of something and find where it breaks. Model how fast a paper aeroplane's flight distance grows with wing length, or build a tiny "predator and prey" tally you step forward turn by turn. Make a prediction with it, test it against the real thing, then write up honestly where your model was right and where it lied. That's real working scientifically (SC4-WS-04, SC4-WS-07) hanging off the model idea.

Us, Thinking Out Loud

The weather model is wrong often enough that we joke about it — so why do we still trust it for tomorrow but not for six weeks away?

Pick a question you actually care about. Which kind of model would you build, and where would you expect it to break first?