Designed for No One: What the Air Force's Vanishing 'Average Pilot' Reveals About Who You're Really Designing For

In 1950, the US Air Force had a problem it couldn't explain. Its planes were among the most advanced in the world and its pilots among the best trained, yet they kept crashing - sometimes many in a single day - with no mechanical fault and no clear pilot error. The cockpits had been built to fit the "average" pilot, using body measurements last updated in 1926. The obvious fix was to re-measure and update the average, and a young lieutenant named Gilbert S. Daniels was assigned to do it.

He measured 4,063 pilots on 140 dimensions, then asked a question nobody had thought to ask: how many pilots were actually average? He took the ten dimensions that matter most for cockpit design - height, chest circumference, arm length and so on - defined "average" generously as the middle band on each, and counted how many of the 4,063 fell inside it on all ten. The answer was zero. Not one. Even on just three dimensions, fewer than 4% qualified. The "average pilot" the entire cockpit had been designed around was a person who did not exist.

That is the trap, and it is far bigger than aviation. When you design for the average, you are not designing for the middle of your users - you are designing for nobody in particular, because no real person is average on every dimension at once. The average is a statistical ghost. A cockpit sized to it fits no one properly, and in a fighter jet, not quite reaching a control at the wrong moment was fatal.

The Air Force's response is the part worth stealing. It did not compute a better average; it threw the average out and demanded adjustability - seats, pedals, straps and stick that moved to fit the individual. Performance and accident rates improved dramatically, and the idea escaped the cockpit: the adjustable seat in your car is a direct descendant of that 1950 crisis.

Modern car interiors are, in effect, a formal answer to Daniels. A cabin is packaged not around one dummy but around a range, typically from the 5th-percentile female to the 95th-percentile male. The whole geometry is anchored to a reference called the H-point - the hip pivot of the seated occupant - and engineers define an H-point "window," a polygon that must contain the hip positions of the small woman, the median man and the large man across a range of comfortable joint angles. Sitting height sets headroom; leg length sets pedal reach and knee clearance. The seat and wheel then adjust across that whole window. The car is not designed for a person; it is designed for a population.

Here is why this is a concept-phase story and not an ergonomics footnote. Adjustability cannot be added at the end. The H-point window and its adjustment ranges are decided in the earliest package drawings, because they dictate the hard points - where the floor, the cowl, the roof and the pedals physically sit. Get the window wrong at the concept stage and no amount of nicer seat foam later will stop a tall driver's head hitting the roof or a small driver stretching for the pedals. The accommodation of human variety is a first-move architectural decision, made long before a single surface is styled.

And it generalises to everything with a user. Every product is quietly designed around an imagined "average" someone - an average hand, an average level of ability, an average context of use - and that someone is as fictional as Daniels' average pilot. The design that fits real people is never the one optimised for the mean; it is the one that decides, at the concept phase, to accommodate a range - through adjustability, through flexibility, through inclusive design that anticipates diversity of body, age and ability instead of averaging it away.

So the question worth asking at the very start of any design isn't "who is the average user?" It's "what is the range of people I'm actually building for, and have I decided, in the architecture, to fit all of them?" Design for the average and you design for no one. Design for the range, from the first sketch, and you design for everyone who was never average to begin with.

Sources:

• ●Lt. Gilbert S. Daniels and the myth of averages - LessWrong
• ●When the US Air Force discovered the flaw of averages (PDF)
• ●No Such Thing As an 'Average' Pilot - World War Wings
• ●No one size fits all - Continuous Delivery (Medium)
• ●How the US Air Force Ditched the 'Average' and Saved Lives - Mann Howie
• ●US Air Force Finds Averages Can Kill - Austraffic
• ●See Beneath the Average (cockpit design) - FullStory
• ●Basic Ergonomics in Automotive Design - Dsource
• ●Automotive ergonomics in human-vehicle systems (review) - Springer
• ●Automotive Ergonomics Design Guide (H-point) - Car-Chassis
• ●Survey of Auto Seat Design Recommendations - UMTRI / Reed (PDF)
• ●Automobile seat comfort vs anthropometric accommodation - ScienceDirect
• ●The Average Man - dlm-econometrics