tfw you definitely know how averages and time work https://twitter.com/andrew_lilico/status/1245613091208826885">https://twitter.com/andrew_li...

this thread deals with a related though slightly different problem & I think is useful https://twitter.com/DFisman/status/1245689823949729792">https://twitter.com/DFisman/s...

wrt Lilico& #39;s tweet & other similar ones going round, you can& #39;t plug in 23 days ago as the average time since infection for current deaths for a number of reasons

* with an epidemic you& #39;ve got (at the current stage we& #39;re in) an increasing number of daily cases. Even if you assume you& #39;ve got the same distribution for time-to-death for infections on each day then a larger-than-naively-expected number are going to be more recent cases.