Histograms

A histogram answers a question no other everyday chart handles well: how is one set of numbers spread out? Take a single continuous measurement collected many times over — and a histogram shows you where those values cluster, how wide they range, and whether they pile up at one end. It does this by chopping the number line into equal slices called bins and counting how many values fall into each one.

How a histogram works

Start with one column of numbers — many measurements of the same thing. Divide the full range of those numbers into a series of equal-width intervals (the bins), for example 0–10, 10–20, 20–30, and so on. Then count how many of your values land in each bin and draw a bar whose height is that count. Bins that contain many values produce tall bars; sparse bins produce short ones. The outline of the bars traces the distribution of the data.

Because the bins are consecutive slices of one continuous scale, the bars sit directly against each other with no gaps. That touching outline is not decoration — it signals that the horizontal axis is an unbroken number line, where the space between bars would itself be meaningful.

The crucial difference from a bar chart

Histograms and bar charts look similar, but they answer completely different questions, and confusing them is the most common mistake people make.

• Axis meaning. A bar chart's horizontal axis holds discrete categories — names or labels with no numeric order. A histogram's horizontal axis is a continuous number line split into ranges.
• Gaps. Bar charts leave gaps between bars to show the categories are independent. Histogram bars touch because the ranges are continuous.
• Order. You can freely reorder a bar chart's bars — by size, alphabetically — without losing meaning. A histogram's bars must stay in numeric order; rearranging them would destroy the distribution.
• What height means. A bar's height is the value of that category. A histogram bar's height is a count — how many observations fell into that range.

Choosing the bin width

Bins are the one setting that changes how a histogram looks, and getting them wrong can hide or invent patterns. There is no single correct number — the goal is a clear, stable shape.

• Too few bins (very wide). The data gets lumped into a handful of tall blocks. The chart looks tidy but smooths away detail, and a second peak can vanish entirely.
• Too many bins (very narrow). Each bin holds only a value or two, so the outline turns into a jagged comb that reflects random noise rather than the real shape.
• About right. Enough bins to reveal the genuine peaks and tails, few enough that the outline is smooth and steady. Trying several widths and keeping the one that stays consistent is the practical method.

One more rule: keep every bin the same width. Unequal bins make the bar heights misleading, because a wider bin naturally collects more values and looks taller for the wrong reason.

Reading the shape

Once the bins are sensible, the outline of the bars tells you most of what you need to know:

• Symmetric (normal-like). A single hump in the middle that tails off evenly on both sides — values cluster around a central point and extremes are rare in either direction.
• Skewed. A hump pushed to one side with a long tail stretching the other way. A right (positive) skew has its tail toward the high values; a left (negative) skew trails toward the low values.
• Bimodal. Two distinct humps, a strong hint that you are looking at two different groups mixed into one chart and might want to separate them.
• Uniform. Bars of roughly equal height across the range, meaning values are spread evenly with no preferred region.

Also read the spread — a narrow, tall cluster means the values are consistent, while a low, wide spread means they vary a lot — and watch for isolated bars far out on the axis, which mark unusual values worth investigating.

When not to use a histogram

Make a histogram

chart.biz does not yet have a dedicated histogram tool. The closest option is the bar chart maker — note that it draws categorical bars with gaps, so if you use it for binned data you would label each bin yourself and remember the bars represent ranges, not categories. For the wider picture, see the complete guide to chart types.