Data visualisation expert on what’s wrong with the UK coronavirus charts

He makes some decent points on their specific choices of graphs but it feels a bit unfair to say the presenters should slow down or that more time should have been spent "to dot the Is and cross the Ts". I'm pretty sure they are under insane external pressure to meet a deadline and to fit everything into a media friendly slot. Delays in communication feed will not be met with nods of approval over their commitment to accuracy - it'll feed into the narrative that no-one knows what they're doing. I've no time at all for the British government but I have to chuckle at some commentary. "It's easy bro, just take your time."

More generally I feel like this is an instance where our reporting/visualisation tools have totally outpaced the data pipeline. We've got interactive dashboards with integrated maps and widgets showing us global figures in real time and they're great. And yes, I was one of those people who morbidly kept the Johns Hopkins dashboard open at the start of all this. But the data that's feeding these is coming from a various Excel spreadsheets updated (partly updated by hand) once a day (but maybe not weekend). These in turn are based on other spreadsheets/CSV files (probably sent by email). The flashy "real time" element gives us a warped perception of what we're looking at. Yes, last weeks figures may change tomorrow - not because of a cover up (or even individual incompetence) - it's because they're adding in the cases that didn't come through last week from one lab/region because someone was on holiday or the file wasn't ready by the deadline.

There was a delay the other day updating the UK daily figures and the data.gov.uk website had a note saying "Last updated : invalid date". I think it was there for an hour or so but even that brief period was enough to invite mocking commentary.

I found the green/brown map particularly confusing at the time - glad it wasn't just me!

You also think that between HMG and the BBC someone would have done a run through to check if the slides displayed properly on a normal TV!

Mass testing with an unknown false positive rate doesn't help things either.