1. I agree with this, and think that this pandemic has revealed the worst of data journalism and technocracy. America is a poor, segregated, and obese country with dysfunctional public and private institutions. The death toll in the US will likely exceed 80k. https://twitter.com/Chris_arnade/status/1248976490387189761

2. Models and data are only tools to help us think through what is happening now and what *might* happen in the future. Case numbers and models are *easily* manipulated and inherently wrong.

3. Misuse of data is a political problem. The most obvious example is the Chinese government's use of fictional numbers for strategic reasons. Then there's a measurement issue in countries that aren't lying. At *best* the numbers we have show direction, at worst they mislead.

4. These numbers don't just undercount Chinese deaths bc of PRC propaganda, they undercount the poor and unseen. In the U.S. that's jails, nursing homes, and dense areas with poor people. Abroad it's much worse. Ecuador has 315 confirmed deaths. And yet. https://twitter.com/mrochabrun/status/1247984914009595908

5. The FT's @jburnmurdoch been the single worst data journalist, using fake data to generate click-bait. I pointed out a week ago Ecuador numbers were obviously fake. Today I looked at his chart, Ecuador has disappeared from his list of countries. https://twitter.com/matthewstoller/status/1246479531365535744

6. Even the choice to compare countries is absurd. Yeah, the U.S. has more cases than Italy. It is also five times the size of Italy. If you compare Western Europe to the U.S., Western Europe has many more cases and deaths. That matters just to understanding what is happening.

7. Then there's modeling. Trying to model a disease we don't know very much about and its impact on many different countries with different social systems, levels of testing, and political openness is impossible. We have no idea what it means to shut off supply chains.

8. Every model, from the ones that over-estimated the initial death wave to those that under-estimated it, is premised on a host of assumptions that won't hold. We will reopen our economy. Hungry poor people in America or elsewhere will not adhere to a quarantine. Period.

9. The nice pretty curves with colored spaces showing health care capacity and geometric lines are just not how this will play out. There will be new spikes after the initial ones. It will not hit rural areas as bad as it hits dense areas. The poor and unhealthy will suffer most.

10. Data and models are useful. But as @superwuster once pointed out, our elites are sort of 'math freaks' obsessed with measurement, even though the future of extremely complex social systems can't be measured but must be understood through observation and anecdote.

11. This assumes we don't find a treatment/vaccine tomorrow. Which again, is an assumption. Newspapers should stop making assertions about what the data shows, beyond what it might imply for directional cases. Or get confirmation from other sources of data. Above all, stop lying.