✦ ✦ ✧
Future-Ready · Data Detective

When Averages Lie

“Average” is an umbrella word hiding three different tools. Pick the wrong one — or let one huge number sneak in — and you get a sentence that is perfectly true and completely useless.
🫙The Billionaire Problem
Nine friends count their savings jars: $10, 12, 15, 18, 20, 22, 25, 28, 30. That adds to $180, and 180 ÷ 9 = $20. Then one billionaire sits down. Nobody’s jar changes by a penny.
BEFORE · nine friends, nine jars MEAN = MEDIAN = $20 $0$10$20$30$40 ✓ Point at anyone in the circle and $20 describes them. That’s a typical value. AFTER · a billionaire sits down (every jar unchanged) MEAN = $100,000,018 🚀 median: $20 → $21 $10$20$30 $1,000,000,000 ✗ “People in this room have about $100 million saved.” Perfect arithmetic. Useless sentence.
Do the math yourself: $180 + $1,000,000,000 = $1,000,000,180, and dividing by the 10 people now in the circle gives $100,000,018. Line all ten values up in order, though, and the two middle ones are $20 and $22 — so the median only creeps to $21. One number got about five million times bigger. The other moved by a dollar.
🧰Three Averages, Three Different Jobs
When data is calm and bunched together, all three land in nearly the same spot. When it isn’t, which one you picked is the whole story.
Same word. Three different tools. Always ask which one. 🥧 MEAN add every value up, divide by how many USE IT WHEN… TOTALS matter 12 people × a mean of 2.5 slices each = 30 slices eaten 🎯 MEDIAN line them up in order, take the middle one USE IT WHEN… EXTREMES lurk incomes · home prices wait times · view counts the billionaire-proof one 👟 MODE whichever value shows up more than any other USE IT WHEN… MOST COMMON wins sizes sold: 5,6,6,6,7,9,12 mode = size 6 (the mean, 7.3, fits no foot)
Check that last one: those seven shoe sales add to 5 + 6 + 6 + 6 + 7 + 9 + 12 = 51, and 51 ÷ 7 ≈ 7.3. A shoe store that ordered “size 7.3” would be ordering a size that doesn’t exist — which is exactly why it wants the mode instead.
🧲What One Outlier Does
Five quiz scores: 82, 88, 90, 94, 96. They add to 450, and 450 ÷ 5 = 90. The middle score is 90 too, so mean and median agree — “the average was 90” tells the whole story. Then somebody turns in a blank page.
BEFORE · five scores: 82 · 88 · 90 · 94 · 96 mean = median = 90 0255075100 AFTER · one blank page goes in as a 0 (same scale) MEAN falls 90 → 75 ↙ one blank page = 0 median barely blinks: 90 → 89
Why 89? With six scores — 0, 82, 88, 90, 94, 96 — there is no single middle one, so the median is the halfway point between the 3rd and 4th values: (88 + 90) ÷ 2 = 89. The total is still 450, but now you divide by 6, so the mean lands at 75. One value dragged the mean 15 points away from where almost everyone actually scored. A lonely, faraway value like that is an outlier, and when outliers pile up on one side we say the data is skewed.
📏The Half of the Story Every Average Hides
Two soccer teams, four games each, and exactly the same mean. Would you say these are the same team?
⚽ TEAM A · 3, 3, 3, 3 mean = 3 G1G2G3G4 RANGE = 3 − 3 = 0 ⚽ TEAM B · 0, 0, 6, 6 mean = 3 G1G2G3G4 RANGE = 6 − 0 = 6 Both add to 12 goals. 12 ÷ 4 = 3. Identical means. If you’re the coach planning Saturday, that gap IS the story — and the average erased it.
📍Percentiles: a Position, Not an Amount
A percentile tells you where you’re standing in the line — nothing about how much of anything you have.
100 kids the same age, lined up shortest → tallest 25th 50th = the MEDIAN 90th shortesttallest 90th percentile = roughly 90 out of every 100 kids your age are your height or shorter. ✗ It is NOT a grade, and NOT a percentage of anything. 25th percentile ≠ “25% as tall.”
🕵️The Three Questions of a Data Detective
Next time someone hands you an average, these take about four seconds to ask.
1Which average is this? Mean, median, or mode? If the person quoting it can’t say — that’s information all by itself.
2How spread out is the data? Ask for the range, or the highest and lowest. Team A and Team B had the same mean.
3Is anything extreme hiding in here? One billionaire, one zero, one record-breaking day can shove a mean somewhere nobody actually lives.
🎢 Try it on a sign at an amusement park: “Average wait: 15 minutes.” Which average? Measured when — Tuesday morning or Saturday afternoon? And is that a steady 15, or half the day at 5 minutes and the other half at 40?
🔑Key Terms
☂️AverageAn umbrella word for any single number used to stand in for a whole group — it might be the mean, the median, or the mode.
🥧MeanAdd every value together, then divide by how many there are. The “fair share” number.
🎯MedianThe value sitting exactly in the middle once the numbers are lined up smallest to largest.
👟ModeThe value that shows up more often than any other in a set of data.
🧲OutlierA value sitting far away from all the others — and often pulling the mean toward it.
🛝SkewedWhen the extreme values pile up more on one side, dragging the mean away from the middle.
📏RangeLargest value minus smallest value — the quickest way to see the spread an average hides.
📍PercentileWhere a value stands in the lineup: the 90th percentile means about 90 out of 100 are at that value or below.
Two more worth knowing: a data set is the whole collection of numbers you gathered — every row of the spreadsheet — and a typical value is a number that genuinely describes most members of a group, instead of just landing in the middle of the arithmetic. The billionaire mean was a real average, but it was never a typical value.
🌍Where You’ll See This in Real Life
🩺Growth charts at the doctor’s officeChildren’s growth charts are built out of percentiles, not averages, so a family can see where a child sits in the whole lineup of kids the same age. And that 50th percentile line running down the middle? That’s simply the median.
📰Newsrooms and statistics agenciesReports on incomes, home prices, and rents lead with the median almost every time, because a handful of enormous values would haul the mean far above what an ordinary household experiences. It’s the billionaire problem at national scale.
📜The word median comes from the Latin medius, meaning “middle” — the same root behind “medium” and “mediator.” It has meant “the thing in the center” for roughly two thousand years.
🧮Mathematicians actually use several different kinds of mean. The everyday add-them-up-and-divide one is the arithmetic mean; the geometric and harmonic means are built for questions where straight addition gives the wrong answer.
🏠Try the detective questions on a real headline: if a town reports a mean household income of $95,000 but a median of $48,000, that gap is doing the talking. A small number of very high incomes are pulling the mean up — so the median is the number that describes a typical household.
📌Remember This
1“Average” is an umbrella word for three different tools. The first question about any average is always: which one is it?
2A single outlier can move the mean enormously while barely moving the median — which is why the median describes “typical” better whenever extremes are in the room.
3Every average hides the spread. Ask for the range or the percentiles too, because two groups with identical means can be completely different.
🤔 Think about it
If a company reports the mean pay of its workers instead of the median, is that lying — or just choosing? Where exactly is the line between selecting a number and misleading with one?
Grades, sports stats, and video view counts all get squeezed into single averages. What gets thrown away when a whole year of your work becomes one number — and what would a fairer summary look like?
⭐ Remember: averages aren’t villains. A summary’s whole job is to throw information away so a human can hold it. The skill isn’t refusing to trust averages — it’s knowing what got thrown away, and asking for it back when it matters.
✏️ ClickClass Anchor Chart · When Averages Lie: Mean, Median & the Billionaire Problem
From ClickClass — hundreds of free printables at clickclassedu.com/printables