Scientists are often quick to remind the public: no model is perfect.
After Professor Neil Ferguson’s computer model suggested in March that 500,000 people in the UK would die of the coronavirus without lockdown measures, the UK quickly abandoned its herd immunity strategy and resorted to shutting down schools and businesses.
A June review in the journal Nature showed that some researchers were able to reproduce Ferguson’s findings, but software engineers said the code was messy and some public-health experts said the results were unreliable.
One finding in particular didn’t seem to make sense: The model showed that closing UK schools and universities during a lockdown actually led to more COVID-19 deaths than if schools had stayed open.
A new peer-reviewed report published in the BMJ on Wednesday suggests that takeaway could be accurate.
“Our first thought was that it was a mistake, but after a little work on the