Published on 2026-07-25
'Thousands of cases': is that actually a lot?
A raw number like 'thousands of cases' means nothing until you know thousands out of how many. Turn it into a rate and the picture usually changes.
'Thousands of cases reported' is a sentence built to sound alarming no matter what it's describing. It works exactly as well for a rare disease in a small town as it does for a common one across an entire continent, because the word 'thousands' never changes, even when the population behind it is a thousand times bigger. That's the whole trick, and it's a much simpler one than most statistical sleight of hand: a big number, with the total it's drawn from left out of the sentence.
The number that needs a partner
Science journalists Tom Chivers and David Chivers, writing for BBC Science Focus, use a deliberately unglamorous example to make the point. Imagine reading that 10,000 people get ingrown toenails every year. It sounds like a lot. But 10,000 out of how many? If that figure covers the whole of Britain, a country of around 70 million people, it works out to roughly one person in 7,000 each year. The raw count and the real rate tell almost opposite stories, and only one of them fits comfortably in a headline.
That's the whole move: any count, on its own, is just a number floating with nothing to compare it to. Attach it to the population it came from and divide, and it turns into a rate, something you can actually judge against other rates. Strip the denominator back out, and you're left holding a number that sounds precise while telling you almost nothing.
When a report count gets mistaken for a case count
There's a sharper version of this problem in public health, one that fact checking organisation FactCheck.org, a project of the Annenberg Public Policy Center, has documented repeatedly. The US Vaccine Adverse Event Reporting System, VAERS, logs reports of health problems that happen after a vaccination, whether or not the vaccine actually caused them. A raw count of reports for a given vaccine tells you how many forms were filed. It does not, on its own, tell you how many doses were given out in total, and without that second number, the report count cannot be turned into a rate at all.
FactCheck.org's reporting shows how easily that missing denominator gets exploited. Some claims have pointed to specific vaccine 'lots' with unusually high report counts as proof that those batches were dangerous, without mentioning that lots vary enormously in size: a lot of 100,000 doses will naturally rack up more reports than a lot of 3,000 doses, even if both are equally safe. The reports pile up faster simply because more doses went out the door. Public health researchers have described VAERS itself as a 'hypothesis generating system', a first alarm bell meant to flag patterns worth investigating, not a finished measurement, precisely because it counts reports rather than confirmed, doses adjusted outcomes.
A count tells you how many forms got filed. A rate tells you how often something actually happens.
Same word, wildly different weight
Picture two towns. One has 5,000 residents and reports 'thousands of complaints' about a new policy in its first month. The other is an entire region of 20 million people and reports the exact same words: 'thousands of complaints'. In the small town, that phrase might describe most of the population up in arms. In the large region, the identical wording could describe a rounding error, a handful of vocal residents in an otherwise indifferent population of millions. Nothing about the sentence changes. Everything about what it means does.
The same trick works just as well moving in the other direction, shrinking a number down until it looks harmless. 'Only a few hundred incidents' sounds reassuring right up until you learn it's a few hundred incidents in a workplace of sixty people, which would mean the average employee was involved several times over. Small numbers hide exactly as much as big ones when the total they're measured against stays out of view. The direction of the illusion changes. The missing ingredient never does.
None of this means big raw numbers are always misleading, or that a count of 'thousands' is automatically nothing to worry about. Sometimes the underlying rate really is high, and the raw number and the rate tell the same alarming story either way. The point isn't to distrust the number. It's to notice when the sentence hands you a count and quietly withholds the total, and to ask, before reacting, the plain question that turns a count into something you can actually judge: out of how many, exactly?