Reading at night shortens life. Video games provoke mass murders. Soda makes people aggressive. What?! No! Every day, the media writes about sensational discoveries and shocking research results. But not all of them can be trusted: statistical errors, deliberate falsifications, and non-obvious flaws have not been cancelled.

The COVID-19 pandemic has forced humanity to take a crash course in statistics: now we are quite familiar with graphs, have heard a thing or two about the normal distribution, and are familiar with the survivorship bias. But we still have a lot to learn: how do mathematical models work? What is the difference between absolute and relative risks? What do ratings tell us? What is the Texas shooter fallacy?

Science journalist Tom Chivers and economics lecturer at Durham University David Chivers use examples of the sensational headlines of the COVID-19 era to show how not to be fooled by numbers.