DESIGN INTELLIGENCEJuly 22, 2026·Mary · DEPIX Design Intelligence

Designed to Fix Racing, Not to Be Faster

For 2026, Formula 1 has done something that looks, on paper, like a mistake: it has made the cars slower on purpose. The new regulations cut aerodynamic downforce by around 30% and drag by 55%, replace the underfloor ground-effect tunnels with a flatter floor, and shrink the car — 30kg lighter, a narrower 1900mm body and a shorter 3400mm wheelbase. Downforce is grip, and grip is cornering speed. Deliberately reducing it is close to heresy in a sport that has chased lap time for seventy years. So why do it?

Because the FIA finally optimised the right thing.

For decades, teams optimised a single-car metric: how fast can this car go, alone, on an empty track. That objective produced spectacular machines and, increasingly, dismal racing. The ground-effect tunnels that generate huge downforce also throw a violent turbulent wake — "dirty air" — off the back of the car. A following car drives into that wake, loses grip and can't stay close enough to overtake. The faster each individual car got, the more impossible it became for two of them to actually race. The product — a contest — was being destroyed by the very number everyone was optimising.

The 2026 rules reset the objective. The explicit goal is to make cars less wake-sensitive and easier to follow. The flatter floor throws a cleaner wake; the smaller, lighter car is nimbler in traffic; active aerodynamics — movable wings with a high-downforce "Z-Mode" for corners and a low-drag "X-Mode" for straights — claw back straight-line speed without the permanent downforce that poisons the air behind. The engineers gave up peak single-car performance to buy something that never shows up on a one-car dyno at all: the ability of two cars to run nose-to-tail.

That is the whole lesson, and it reaches far past motorsport. The most common and most expensive design mistake is to optimise the metric you can measure on one unit in isolation — a spec-sheet number, a benchmark, a lap time, a single-user demo — while the value actually lives in a system behaviour nobody put on the dashboard. A phone optimised for benchmark scores that runs hot in a pocket. An app optimised for a flawless demo that buckles under real concurrent users. A car optimised for a magazine 0–60 that is miserable in traffic. In every case a capable team optimised hard — toward the wrong objective. And a team will pursue the wrong objective just as ruthlessly as the right one.

Studios do it constantly: a car styled to stop traffic in a press render that you can barely see out of; an interface tuned for a slick launch keynote that exhausts the person using it every day; a building optimised for the aerial photograph rather than for the person walking through the door. Whatever gets measured is what gets served — so the measurement is the design.

Which is why the objective is a concept-phase decision, and the highest-leverage one there is. It is set at the very start — in the regulations, in the brief — and then propagates into every downstream choice. Change F1's objective from "fastest single car" to "cars that can follow" and the entire machine changes: floor, wings, weight, dimensions, power split. The 2026 power unit even moves to a near 50/50 split of combustion and electric power on sustainable fuel — another objective (relevance, sustainability) written into the brief and engineered outward. Nothing downstream can rescue a project pointed at the wrong target; it can only reach the wrong target more efficiently.

There is a reason the FIA had to legislate this rather than trust the teams to find it. Left to optimise freely, every team chooses the setup that wins Saturday's qualifying lap over the one that makes Sunday's race worth watching — because qualifying is a clean single-car number and "good racing" is a diffuse system property no one's KPI rewards. The sport had to change what it measured to change what its brilliant engineers built. Formula 1 is, at heart, the world's most expensive optimisation problem — and 2026 is a public admission that it had been optimising the wrong variable.

The takeaway for anyone who writes a brief: the hardest question is never "how do we make this better?" It is "better at what?" Get that answer wrong and you hand a talented team a precise, well-funded path to a worse product. Get it right — name the system outcome the thing actually exists to produce — and everything downstream, for once, pulls in the direction you meant.

Sources: