One thing the last crisis taught me was a deep distrust of dominant narratives and numbers expressed or compared without understanding the uncertainty around them. Popular threads around #COVID19 feature ample amounts of both today. Tools I use to get around these:

1. Write out the opposite of the argument advanced. Eg: The UK death rate, higher than The Irish rate, is not a function of the policy failures of an inept govt but the increased population densities of U.K. cities. This argument might be true. It might not. But it is plausible.

2. Swop the main characters in the argument to discover your priors. “The Irish govt dilly-dallied when it should have locked down earlier” is also plausible. Again, it might be right it might be wrong. But ask yourself: would we all be RT’ing threads bashing us, not the Brits?

3. Learn how the data are constructed. Where do these datasets come from? You learn a *lot* about where the uncertainty comes from if you do this. Eg: @jburnmurdoch is very good in describing all the potential flaws in his now famous data visualisation for the @FT.

4. Just be skeptical of any narrative shared and reshared by tens of thousands. Chances are high it is either joining the flow of an existing narrative, confirming your existing biases, or not comparing like with like.

Finally the above points say nothing about whether these threads are right or wrong. The authors are arguing in good faith and are good people. It just says be careful what arguments you accept and share and why.