At the boards, someone fits a curve through every point of a braking car’s position data — six radar readings, a degree-five polynomial that misses nothing — and announces “perfect fit”. Is it a good model? Differentiate it. The velocity that falls out of the perfect fit wiggles: it claims the car sped up twice in the middle of braking, something no driver did and no brake can do. Meanwhile the group beside them has a plain constant-deceleration model that misses every single point by a little — and its velocity says the one thing that must be true: steadily down, through zero, then stop. Between “passes through the points” and “tells the truth about the motion” runs the gap this conversation is about, and calculus makes the gap visible in a way nothing before it could: differentiation amplifies wiggles. A model can hide its lies in its values and still confess them in its rates.
Questions worth arguing about:
- What exactly has the perfect-fitter shown, and what have they not? Is hitting every data point evidence of anything?
- The plain model’s parameters mean something: one is the speed the car was doing when the brakes bit, the other is how hard they bit. Why does a model whose knobs have names deserve more trust than one whose knobs do nothing but fit?
- Every model has an expiry: the braking model dies the moment the car stops — trust it past that and it predicts the car reversing into the distance. Whose job is it to say where a model stops being true — the equation’s, or the person wielding it?
- When is simple-but-slightly-wrong the better choice than complicated-but-close? Would you rather the court weighing a speeding ticket trusted the first kind or the second?
- “All models are wrong, but some are useful.” Prosecute or defend this claim — and decide what “useful” has to mean for it to survive.
This stops being talk at The Speed Camera and Smooth Landing, where your group must choose a model, defend every parameter, and say out loud where it stops deserving belief — with its derivatives called as witnesses. The habit of asking what a graph’s shape claims starts small, every morning, in Graph Talks.