Skip to content

Percentiles Explained: What 'Top 10%' Really Means

Percentiles are everywhere, from growth charts to test scores, and most people read them slightly wrong. What a percentile does and doesn't tell you.

Math···4 min read

What a percentile actually is

A percentile tells you what share of a population falls at or below your value. If you're at the 70th percentile for height, you're taller than about 70% of people in the reference group, and shorter than the other 30%.

That's it. It's a ranking within a group, not a score out of 100 and not a percentage of some maximum.

The 70th percentile is a rank, not a scoreWhat a percentile says about where you stand in the group

70thpercentile: at or above 70 of every 100 people in the reference group, below the other 30

70 of 100
Definition of a percentile rank

The three things people get wrong

1. "90th percentile means 90% as big." No. It means bigger than 90% of people. The 90th percentile and the 50th percentile might be only a few centimeters apart, because most people cluster near the middle.

2. "Higher is always dramatically different." In a normal (bell-shaped) distribution, values bunch up around the average. Going from the 50th to the 60th percentile is a small step. Going from the 95th to the 99th is a much bigger jump in actual value, because you're out in the thin tail. Percentiles are not evenly spaced in real units.

Ten percentiles in the middle are a smaller step than four at the topLength of 12-month-old boys on the WHO growth standard, in centimetres
WHO Child Growth Standards, length-for-age, boys at 12 months (LMS values as used by our kids growth calculator)
Show the numbers
Values
50th to 60th percentileP50, P60 75.75 to 76.35 (+0.6 cm)
95th to 99th percentileP95, P99 79.66 to 81.28 (+1.6 cm)

3. "My percentile is a fixed fact about me." It depends entirely on the reference group. The 50th percentile for a metric among teenagers is different from among adults. Always ask: percentile compared to whom?

The bell curve, in one picture

Most human measurements (height, birth weight, and yes, the ones behind our percentile calculator) follow a roughly normal distribution:

  • ~68% of people fall within one standard deviation of the average.
  • ~95% fall within two.
  • The extremes (top and bottom ~2.5%) are genuinely rare.
Most people sit within two standard deviations of averageShare of a normal distribution within 1, 2 and 3 standard deviations of the mean
NIST/SEMATECH e-Handbook of Statistical Methods, the normal distribution
Show the numbers
Share of people
Within 1 SD68.3%
Within 2 SD95.5%
Within 3 SD99.7%

So "average" isn't a single magic number. It's a range where most people sit. Being at the 40th or 60th percentile is, statistically, completely unremarkable.

Why "1 in N" framing helps

Percentiles can feel abstract, so it's often clearer to flip them into rarity:

  • 90th percentile → about 1 in 10 people are larger.
  • 99th percentile → about 1 in 100.
  • 50th percentile → you're right in the middle; half above, half below.

This is why good calculators show both: a percentile and a "1 in N" or "in a room of 100 people, about X would be larger." Same fact, easier to feel.

Percentiles you'll meet on this site

The same math powers several tools here:

The healthy way to read your own percentile

A percentile is a description, not a verdict. It tells you where you sit in a distribution, and nothing about worth, health, or outcomes on its own. For growth charts, doctors care far more about a child tracking a consistent curve than hitting a specific number. For most adult metrics, anything in the broad middle is simply normal.

When you check your own percentile, hold onto that: the interesting information is usually "am I in the normal range?" (almost always yes), not "which exact percentile?"

mathstatisticspercentiledataexplainer

Share this post