Plot makes differences.

All three could have been tables. Same numbers, same honesty, nobody reading past the first row.

So each comes with its receipts.

Ranked bars · Regional economies

Rust Belt cities still run on hospitals and universities

Sorted horizontal bar chart of anchor institution GDP share across major U.S. metros, colored by region, with Cleveland highest at 14 percent and San Jose lowest at 4.3 percent.

Two sentences sit inside the panel, one by the teal, one by the navy. Between them you learn what the colors mean, without leaving the chart.

Trend lines · Traffic speeds

Chicago drives faster, but every city crawls by rush hour

Line chart of average weekday driving speed by hour for Chicago, Philadelphia, and Boston, with faint scatter points behind each line and a shaded evening rush hour band.

The faint dots are the raw weekly numbers, left in so you can see how noisy the measurement is. The bold lines stay the thing you follow.

Scatter · Pedestrian safety

Where children walk in danger, everyone does

Scatter plot of child versus total pedestrian fatality rates across 30 U.S. metros, with a fitted trend line, gray unlabeled metros, and colored labeled outliers including Bakersfield, Fresno, and Florida cities.

Thirty labels would be a mess. Most points stay gray and nameless, here to make the cloud. Only the metros with something to say get a name, and McAllen gets one for breaking it.

Why bother styling a chart?

Because the plain version is not neutral. It was styled too, by software, for nobody in particular.

A headline that says something. A sentence inside the panel. A source line at the bottom. None of it is hard. It is mostly deciding to treat a figure as writing.

The maps are on North Is a Choice. The rest is in Thoughts.

Built with
R, ggplot2, ggrepel
Figure 01
Federal Reserve anchor economy data
Figure 02
Hourly traffic speeds, weekdays
Figure 03
FARS 2017 to 2022, ACS population
Status
Practice, on public data
Back to Thoughts