A flu season simulation

How it works

The simulation follows 10,000 US adults through one typical flu season, showing how many get sick with flu, how many need hospital care, and how many die. It then re-rolls the dice for the same people as if all had been vaccinated.

This is a Monte Carlo simulation: for each adult, we roll dice weighted by likelihood.

A typical season. The chances come from CDC’s estimates of flu illness, hospitalization, and death by age for 13 seasons from 2010–11 to 2024–25, averaged. The two COVID-19 seasons are excluded. The rates are adjusted to people who aren’t vaccinated and weighted to today’s mix of US adults by age, so older adults, who account for most flu hospitalizations and deaths, count in proportion.

The vaccine prevents about 40% of flu illness serious enough to see a doctor in a typical season, the median of CDC’s season-by-season estimates since 2010–11. The simulation applies the same 40% to hospital care and death, as CDC’s own analysis of vaccine benefits does.

Caveats. Real seasons vary widely, from mild to severe, and so does how closely the vaccine matches the flu strains circulating that year. The simulation shows an average season, not a forecast for this one or a personal risk estimate.