A distribution curve is a smooth line on a graph that shows how frequently different values appear in a set of data. If you measured the heights of 10,000 adults and plotted how many people fall at each height, the resulting shape would be a distribution curve. The highest point on the curve marks the most common value, and the tails stretching out to either side represent increasingly rare values. It’s one of the most fundamental tools in statistics for understanding patterns in data.
How a Distribution Curve Works
Imagine collecting thousands of measurements of something, like test scores in a large school district. You could sort every score into bins (60–65, 65–70, 70–75, and so on) and stack them into a histogram. With enough data points, the jagged tops of those histogram bars start to smooth out into a curve. That smooth curve is your distribution. It tells you, at a glance, where most values cluster, how spread out they are, and whether extreme values are common or rare.
The vertical axis represents frequency, or how often a particular value shows up. The horizontal axis represents the values themselves. A tall, narrow curve means most of the data clusters tightly around one value. A short, wide curve means the data is spread across a broader range. The total area under the curve accounts for 100% of the data, so you can use sections of that area to estimate the probability of landing in any given range.
The Normal Distribution: The Classic Bell Curve
The most well-known distribution curve is the normal distribution, often called the bell curve because of its symmetrical, bell-like shape. In a perfect normal distribution, three properties hold true: the mean (average), median (middle value), and mode (most common value) are all identical and sit right at the center. If you folded the curve in half at that center point, both sides would mirror each other exactly.
The normal distribution shows up constantly in nature. Human height is a classic example. When researchers analyzed military records for Italian men born around 1900, the age-adjusted average height was about 164 cm with a standard deviation of roughly 6.5 cm. Plotted on a graph, the data closely matched the expected bell-curve shape, with most men clustered near the average and progressively fewer at the extremes.
Blood pressure across a population follows a similar pattern. In a study of adults in Northern Ethiopia, the average systolic blood pressure hovered around 111 to 120 depending on sex and whether someone lived in an urban or rural area. Most people fell near the middle of the curve. Those in the far right tail, with readings at or above 140, met the clinical definition of hypertension. Those in the far left tail had unusually low blood pressure. The curve itself doesn’t define what’s “healthy,” but it reveals how common or uncommon any given reading is relative to everyone else.
The 68-95-99.7 Rule
One of the most practical features of a normal distribution is a pattern called the empirical rule. About 68% of all data falls within one standard deviation of the mean. About 95% falls within two standard deviations. And about 99.7% falls within three. This gives you a quick way to gauge how unusual a value is without doing any complex math.
Standard deviation is just a measure of how spread out the data is. A small standard deviation means values are packed tightly around the average, producing a tall, narrow bell curve. A large standard deviation means values are more scattered, producing a flatter, wider curve. Two datasets can have the exact same average but look completely different on a graph because their standard deviations differ.
To put this concretely: if the average adult male height in a group is 175 cm with a standard deviation of 7 cm, then roughly 68% of men in that group stand between 168 and 182 cm. About 95% fall between 161 and 189 cm. Someone who is 200 cm tall sits more than three standard deviations above the mean, placing them in the rarest 0.3% of the distribution.
Z-Scores: Locating a Value on the Curve
A z-score tells you exactly where a single data point sits on a normal distribution curve, measured in standard deviations from the mean. You calculate it by subtracting the mean from your value and dividing by the standard deviation. A z-score of 0 means you’re right at the average. A z-score of +2 means you’re two standard deviations above it.
This conversion is useful because it puts any dataset into a common language. Whether you’re looking at test scores, blood pressure readings, or factory part dimensions, converting to z-scores lets you compare how unusual a value is regardless of the original units. Once you have a z-score, you can use a standard reference table to find what percentage of the data falls below that point. A z-score of +1.0, for instance, means roughly 84% of values fall below yours.
Skewed Distributions
Not all data forms a neat bell curve. When data bunches up toward one end with a long tail stretching in the other direction, the distribution is skewed. A positively skewed (right-skewed) distribution has most values clustered on the lower end with a long tail reaching toward higher values. Income distribution is a common example: most people earn modest amounts while a small number earn dramatically more, pulling the tail to the right.
A negatively skewed (left-skewed) distribution is the opposite. Most values cluster at the higher end, with a long tail stretching toward lower values. Age at retirement in a stable economy can look like this: most people retire around a similar age, but a smaller group retires much earlier, creating that leftward tail.
In skewed distributions, the mean, median, and mode no longer line up. The mean gets pulled toward the tail because extreme values drag it in that direction. This is why income statistics typically report the median rather than the mean. The median sits closer to what a “typical” person actually experiences, unaffected by billionaires stretching the tail.
Bimodal and Uniform Curves
Some distributions have two distinct peaks instead of one. These are called bimodal distributions. A bimodal curve might appear if you measured the heights of all adults without separating by sex. You’d see one peak near the average female height and another near the average male height. Any time two overlapping groups contribute to the same dataset, a bimodal pattern can emerge.
A uniform distribution, by contrast, has no peak at all. Every value in the range is equally likely. Rolling a fair die is the simplest example: each outcome (1 through 6) has the same probability, so the distribution is flat. A multimodal distribution has three or more peaks, which typically signals that several distinct subgroups are mixed together in the data.
Why the Normal Curve Appears So Often
There’s a mathematical reason the bell curve keeps showing up in nature and research: the central limit theorem. This principle states that if you take large enough samples from any population and calculate the average of each sample, those averages will form a normal distribution, regardless of the shape of the original data. The rule of thumb is that a sample size of 30 or more is usually sufficient for this effect to kick in, as long as the samples are drawn randomly and the population’s variability isn’t infinite.
This is why the bell curve is so central to statistics. Even when individual data points follow a lopsided or unusual pattern, the averages of repeated samples from that data will settle into a predictable, symmetrical curve. It’s the reason researchers can make probability statements about everything from drug effectiveness to manufacturing quality, even when the raw data looks messy. The normal distribution isn’t just one shape among many. It’s the shape that emerges naturally when enough independent factors combine, which is why it appears in contexts as different as human biology, measurement error, and financial modeling.