[arblify]

I think you're right, we do need a plan, but we need trained epidemiologists and other public health figures to define that plan. Many features of disease spread are counterintuitive and we need people with training to help us, the public, understand the problem and reasonable solutions. When people like Michael Osterholm tell me that it is possible one or two million people will die in the United States, I believe him and I'm willing to go along with the plan. That said, I would like to see more clear definition of where the risk lies, how we can take small chances that have a low chance of going catastrophically wrong, and how we can responsibly try to revert to some kind of normality.

[CPLX]

I mean, it's not like this crisis started an hour ago. What's their plan? Like can someone write it down and post in on a government website?

We're all commenting on what appears to be an official government press release that is announcing an entire month of some of the most serious restrictions on the public ever put in place, that does not appear to even try to take any kind of quantitative approach to explaining why that policy was put in to place, and how its effectiveness is being assessed.

Broad, destructive public policy not tied to success metrics is fucking insane and I think the strong negative reactions to it are completely warranted.

[arblify]

Once you commit to some success metric, you're going to be stuck with it even it proves to be a bad metric. Let's say we all agree the "2 weeks with decreasing number of deaths" is a good target. Then we go back to work, and we discover that going back triggers a rapid spread and kills a bunch of people and starts to overwhelm hospitals. Now we need a new metric.

This is a novel scientific problem. Caution and study are warranted. People's lives are at stake.

[CPLX]

People's lives are always at stake. That's how public policy works. Like 100% of the time. Try painting lines on the highway or administering a school lunch program without putting people's lives at stake.

The only thing novel here is completely abandoning the concept of public policy goals. There's nothing scientific about setting your public policy without any metrics at all. This is kind of the opposite of a scientific approach.

[arblify]

I'm not suggesting having no metrics, I'm talking about hard targets. Acknowledging that there are things we don't know about the problem and that we're not overcommitting to a course of action based on some specific target is reasonable.

At the same time, to your point, I would like to see clear explanations of "these are the aspects of the disease we are trying to understand (virality, mortality, etc)," these are the constraints on our healthcare system, these are the economic effects, here are the tradeoffs we're trying to make. All of that stuff is good, but it's a complex problem and assuming that we know enough at this point to set a clear numerical goal seems wrong to me. Describing general parameters for our data gathering and decision-making is good, though, and I agree that I'd like to see more of it.

[CPLX]

I think that the government agency in charge of the shutdown should be able to answer each of three basic questions with clarity and some kind of numerical response:

1) What are you hoping this policy will accomplish?

2) What sequence of events, if any, would cause you to accelerate your timetable for easing restrictions?

3) What sequence of events, if any, would cause you to delay your timetable for easing restrictions?

I mean those should be the raw basic cost of even having this conversation. I am used to seeing magical thinking and emotional political arguments in many places but I am surprised to see such hostility to a basic quantitative approach on HN of all places.

[arblify]

Your questions are all reasonable. I'm not personally opposed to quantitative decision-making, but I'm also painfully aware of the limitations of quantitative methods, especially when applied under pressure. I would argue that a blind faith in mathematics is just as wrongheaded as the magical thinking you're describing.

To give a clear example of why I'm skeptical, look at the use of quantitative methods to conduct governance in the banking industry. It's not that we shouldn't have used numerical methods, it's just that they ended up being woefully insufficient because of how they were applied. There's no reason we couldn't make the same mistake here in a premature bid for some kind of certainty.