No One Is a 4.3
A restaurant has 4.3 stars. The number looks objective, precise, trustworthy — it's what the system calculated after hundreds of people each entered a score. But have you ever stopped to ask: who, exactly, gave it a 4.3?
No one gave it a 4.3.
The person who gave five stars thought it was worth five stars. The person who gave two stars thought it was worth two stars. No one in between pressed a "4.3" key. This number is not any single real customer's judgment — it's what you get when you take a crowd of enthusiasm and a crowd of disappointment, add them together, divide by headcount, and squeeze out an average. It describes, with perfect precision, a person who does not exist: the "average customer."
And the average customer has never walked through that door.
The ones who actually walked in — one came loving the restaurant's signature, willing to wait an hour in line. He pressed five stars. Another came for a different dish that turned out to be unavailable, and hated the wait. She pressed two stars. Both experiences happened, both were recorded honestly — and then the average stirred them together, cooked them into a flavor no one has ever tasted, and served it up as 4.3.
This Ghost Doesn't Live Only in the Review Section
Average income is a number no one actually earns. A city's average temperature might not match a single real day all year. A class average might be a score no student in that class actually received. We've grown so used to representing a group with a single number that we forget what's standing beneath it: not a row of identically-spec'd copies, but a crowd of people who are utterly unlike each other.
The closest example may be in your own room.
The air conditioner is set to 22°C. That number on the panel is clean, certain, shared by everyone in the room. But the person sitting directly under the vent is cold enough to reach for a jacket. The person tucked in the corner is still fanning themselves with a hand. The person on the window side, with the afternoon sun coming in, can barely tell the AC is running. 22°C is the panel's temperature — not anyone's actual temperature. Under the same roof, sharing the same number, a whole row of bodies sits at completely different levels of warm and cold. No one is really living at 22°C.
We've Been Asking the Wrong Question
When we encounter an average, we habitually ask: is it accurate? As if there's a correct answer lying somewhere, waiting for this number to converge on it. But for a group of people who are already different from each other, there is no single correct answer to converge on — only answers that are right "for a certain subset of people."
The right question isn't "is this number accurate?" It's: which two groups did this number stir together?
Reading an average shouldn't be about getting the answer — it should be about pulling it apart, and finding the crack it's been smoothing over: who's standing on one side, who's on the other, and why they ended up on opposite sides. The crack is the truth. The average is just a board laid over the crack.
The real danger of an average isn't that it lies. It doesn't lie — it calculates, faithfully and honestly. The danger is that it looks too much like an answer. It hands you a clean, precise, trustworthy number, and gently lets you stop asking "for whom?"
And almost every important question has its answer hidden inside those two words: for whom.