. @EdConwaySky is having a good COVID-19 war & this is worth reading. It’s become obvious COVID-19 models are crap, to put it bluntly. And more importantly, it’s not clear they add anything that can’t be achieved by, for eg, what @BristOliver does. That is, an accurate observation https://twitter.com/edconwaysky/status/1324819551335403520

of the data and it’s trend over the short term, and a simple statement about where that could lead. Then it’s up to everyone to make a decision based on that analysis. In contrast, the models are presenting a fake accuracy for something too complex and uncertain over long time

frames. Epidemics are already driven by multiple variables but with this pandemic we’ve started to interact and respond to epidemic growth so this creates a situation that is clearly impossible to model. We should keep the COVID-19 analysis more simple and data driven, start

including other concerns (eg, economic, focused on public health effects if you like), and then be clear that fundamentally we’re faced with ethical decisions similar to what we routinely do in other areas (eg, healthcare rationing)