Skip to content
← Insights
Nov 14, 20263 min read

Five striking numbers, five different measurement problems

Shark incidents, meat production, seabed coverage, manufactured mass and sand use need their units and dates before they can inform a decision.

A pastel measuring scale surrounded by a globe, a small sample jar and a research notebook.

A large number can change the scale of a problem without telling us what decision to make. The unit, observation year, and method determine what the number can support.

Five examples make that visible. They concern sharks, food production, the seabed, manufactured material, and sand. Their figures cannot be placed on one common risk scale. Each reveals a different measurement problem worth keeping beside the headline.

Shark deaths and shark bites describe different populations

The International Shark Attack File's 2025 summary records 65 confirmed unprovoked bites worldwide, with nine unprovoked fatalities. Its classifications distinguish those incidents from provoked bites and other encounters. The summary also explains that investigations can change classifications.

A 2024 fishing-mortality study estimated at least 80 million sharks killed by fishing in 2019. That is an estimate about sharks, using a different method and observation year from the human-incident record.

The contrast can direct attention towards fishing impacts. Dividing the two figures would not produce a person's chance of being bitten. A beach decision needs local conditions and current advisories; an annual worldwide incident total lacks the relevant exposure denominator.

Food production needs an explicit quantity

FAO's agricultural release covering 2024 reports 374 million tonnes of world meat production, across meat types. It identifies chicken, pig, and cattle meat as the largest items. This is a mass total, not a count of animals.

Converting chicken slaughter totals into a daily average can be useful, provided the annual data and unit are known. Estimating the birds needed for a proposed meal requires an edible-yield assumption. A raw carcass weight cannot simply become a plated portion.

A hypothetical universal daily chicken portion also assumes a dietary pattern people do not share. The arithmetic can illustrate demand under that assumption. It cannot explain the observed global total without information about actual diets, yields, and waste.

A mapped seabed has not all been seen directly

Seabed 2030's April 2026 update reports 28.7% of the ocean floor mapped in its global coverage measure. That replaces older figures around 26-27% when describing this release.

Mapping coverage is a statement about bathymetric data. It does not mean the same proportion has been examined visually at close range. A map can support a question about depth and shape while leaving questions about small organisms or local conditions unanswered.

The reverse misunderstanding is possible too. Unmapped to a modern coverage standard does not mean there is no information whatsoever. Coarser global data and direct modern surveys answer questions at different resolutions. A project needs the resolution relevant to its intended use.

Manufactured mass is a stock with a defined boundary

The 2020 study comparing human-made mass with biomass estimated a crossover around 2020, with an uncertainty of plus or minus six years, on a dry-weight basis. The manufactured total includes material categories beyond concrete.

Dry weight excludes water from the biomass comparison. Changing that definition changes the comparison. The headline should keep the boundary visible instead of implying that one concrete measurement outweighed every living organism under any possible definition.

The estimate describes material accumulated in use under the study's accounting. It is not an annual production figure. It should not be placed beside a yearly extraction total without explaining that difference.

Sand use is an annual flow

UNEP's May 2026 report announcement gives global sand use at about 50 billion tonnes per year. Its discussion distinguishes material used for development from sand remaining in natural systems.

That global scale cannot identify the impact of a particular extraction site. The location, material properties, and extraction conditions affect the relevant decision. A construction specification also cannot assume every sand deposit is interchangeable.

The number is an annual flow, unlike the accumulated manufactured mass in the previous example. Both figures can inform material-use questions. They measure different parts of the system.

For any striking statistic, retain enough context to reconstruct the claim: what was counted, when it was observed, and how the estimate was made. That small record makes a number useful after its surprise has worn off, especially when the next release changes it.

ScienceDataEnvironment