Skip to content
Notes on probability, panic & better guesses
← tools

File 03The Carcinogen, to Scale

The Carcinogen, to Scale

Same hazard group as plutonium. Nowhere near the same risk.

Almost everything “causes cancer” these days — bacon, a daily drink, sunbeds, smoking. Pick one and watch the scary relative percentage get struck out and rewritten as what it actually adds: extra cases per 1,000 people. The grid floods red for the exposure that earns it and shows a thin stripe for the rest. Then translate any headline you read yourself. Every figure sourced.

Interactivethe carcinogen, to scalepick an exposure

A headline gives you a relative risk — “+18%!”. What decides whether that matters is the absolute risk underneath it. Pick an exposure and watch the scary percentage get struck out and rewritten as what it actually adds, per 1,000 people.

eating the most vs. the least — 50 g a day is about three rashers of bacon

17% higher risk of bowel cancer about 10 more in 1,000

Anyway

56

in 1,000

With it

66

in 1,000

+10

the +17% you read about

Out of 1,000 people, about 56 get bowel (colorectal) cancer anyway. This exposure is linked with roughly 10 more — the rest of the thousand is unchanged.

Group 1 · carcinogenic to humans

The evidence that it can cause cancer is conclusive. This says nothing about how much.

  • get it anyway
  • the added cases
  • unaffected
The honest footnote

Cancer Research UK's own illustration: of 1,000 UK people about 61 develop bowel cancer at some point; among the lightest processed-meat eaters about 56, among the heaviest about 66 — ten more. The famous “+18% for every 50 g a day” is IARC's separate relative estimate. Either way it's a small addition to a modest baseline, and it rises with how much you eat.

Now do it to any headline you read

A study says something raises a risk by a scary-sounding percent. Set it here, with the baseline it sits on, and read off what it actually adds.

56 in 1,000 66 in 1,000 — that scary percent is 10 more people in a thousand.

A big percentage of a small number is still a small number. A small percentage of a huge number can matter. The relative figure alone never tells you which you are looking at.

Proportion, not prescription — this sizes a risk, it doesn't tell you what to eat, drink, or fear. Not medical advice.