Knowledge Base
The Normal Distribution Explained
The bell curve is the most important shape in statistics. It describes everything from human height to measurement error — and it is the foundation for percentile rankings.
When you see your percentile on this site, a normal distribution (or a closely related model) is often behind it. Understanding the bell curve helps you interpret what your ranking actually means.
Am I tall or short for my age?
Human height is one of the most perfectly normally distributed traits. NCD-RisC data across 200 countries confirms the Gaussian fit.
📏 Body & Appearance — Check your percentile →Is my blood pressure normal?
Systolic BP follows a roughly normal distribution in healthy adults (mean ~120 mmHg, SD ~15 mmHg), though it skews right with age.
❤️ Health — Check your percentile →Do I sleep enough?
Sleep duration clusters around 7-8 hours with a standard deviation of about 1 hour — a tight bell curve confirmed by NHANES data.
❤️ Health — Check your percentile →What is your Big Five profile?
All five personality dimensions (OCEAN) are designed to produce normally distributed scores across the population.
🎭 Personality — Check your percentile →What Is a Normal Distribution?
The normal distribution — also called the Gaussian distribution after mathematician Carl Friedrich Gauss — is a symmetric, bell-shaped probability distribution defined by two parameters: the mean (μ) and the standard deviation (σ). The mean determines the center of the bell, and the standard deviation determines how spread out the values are. It was first described by Abraham de Moivre in 1733 and later formalized by Gauss and Laplace.
The normal distribution arises naturally whenever a variable is influenced by many small, independent factors. This is the essence of the Central Limit Theorem: the sum (or average) of a large number of independent random variables tends toward a normal distribution, regardless of the original distribution of each variable. This is why human height — influenced by hundreds of genes plus environmental factors — follows a bell curve almost perfectly.
The 68-95-99.7 Rule
The most useful property of the normal distribution is the empirical rule: approximately 68% of values fall within 1 standard deviation of the mean, 95% within 2 standard deviations, and 99.7% within 3 standard deviations.
For example, adult male height in the US has a mean of about 175.3 cm and a standard deviation of about 7.6 cm (CDC/NHANES data). This means roughly 68% of men are between 167.7 cm and 182.9 cm, 95% are between 160.1 cm and 190.5 cm, and a man shorter than 152.5 cm or taller than 198.1 cm is in the most extreme 0.3%.
Standard Deviations and Z-Scores
A z-score expresses how many standard deviations a value sits from the mean. A z-score of +1.5 means the value is 1.5 standard deviations above average. Z-scores convert any normally distributed variable into a standard scale, making it possible to compare across different units. An IQ of 130 (z = +2.0) and a height of 190.5 cm for a US male (z = +2.0) represent the same rarity — both sit at approximately the 97.7th percentile.
This is exactly how percentile calculators work: convert your raw value to a z-score, then look up the corresponding percentile in the standard normal cumulative distribution function (CDF). A z-score of 0 corresponds to the 50th percentile, +1.0 to the 84th, +2.0 to the 97.7th, and -1.0 to the 16th.
Real-World Examples of the Bell Curve
Height: As noted above, human height is one of the cleanest examples. The NCD Risk Factor Collaboration (2016, published in eLife) measured 18.6 million adults across 200 countries and found height distributions closely approximating Gaussian curves within age-sex groups.
Test scores: The SAT is explicitly normed to produce a near-normal distribution with a mean around 1050 and SD of about 200 (College Board, 2023 data). IQ tests are normed to a mean of 100 and SD of 15.
Measurement error: Repeated measurements of the same physical quantity — weighing the same object 100 times — produce errors that are normally distributed. This principle underpins all scientific measurement theory.
When the Bell Curve Does NOT Apply
Not everything follows a normal distribution. Income and wealth are right-skewed — a few extremely wealthy individuals pull the tail far to the right (Pareto or log-normal distributions fit better). Waiting times between events (like time between earthquakes) follow exponential or Poisson distributions. Binary outcomes (pass/fail) follow binomial distributions. For these cases, percentile rankings still work perfectly — they just cannot be calculated using the simple z-score method and instead use empirical (rank-based) percentiles, which is what this site does for non-normal variables.
Why It Matters for Personal Benchmarking
Understanding the bell curve helps you calibrate expectations. Being at the 84th percentile for a normally distributed trait means you are exactly 1 SD above average — clearly above typical, but not rare. Being at the 97.7th percentile (2 SD) means only about 1 in 44 people exceed you. And reaching the 99.9th percentile (3 SD) puts you in a group of roughly 1 in 741. These intuitions help you decide whether a result is worth celebrating, worrying about, or simply accepting as solidly normal.