COVID-19 Exponential Growth forecasting It's a big building with patients but that's not important right now Life nonlinear dynamics predictive analytics Problem Solving real world Risk Statistics

argumentum ad ignorantiam

We have it totally under control. It’s one person coming in from China, and we have it under control. It’s going to be just fine.

President Trump — Jan. 21, 2020 [Interview]

President Trump’s was confused. At least President Trump’s confusion is consistent over time. Very similar to his praise of West Virgina Governor Jim Justice the other day: “Big Jim, the governor, he must be doing a good job of that. That’s what’s reported. …West Virginia is the only one that has no cases. So obviously that’s being treated differently than a New York or California.” Trump continues to show he fails to understand absence of evidence is not, in fact, evidence of absence. Hours later, West Virgina would announce it’s first case. As of March 20th, they would have 7 cases.

SARS-CoV-2 is like honey badger. It just don’t care. It just doesn’t give a shit about human created artificial borders and boundaries. Trump and Co. travel bans notwithstanding.

For many of the smart people I know watching the COVID crisis unfold here in the US, I see a common trend (I am partially guilty here too). We have been waiting for the data we feel will help us better assess the state of the situation. We have been taking comfort in hiding behind the lack of data. Likely influenced by Lord Bertrand Russell’s admonishment:

The demand for certainty is one which is natural to man, but is nevertheless an intellectual vice. … So long as men are not trained to withhold judgment in the absence of evidence,  they will be led astray by cocksure prophets, and it is likely that their leaders will be either ignorant fanatics or dishonest charlatans. To endure uncertainty is difficult, but so are most of the other virtues.

Bertrand Russell – Unpopular Essays

We all seem to wonder about the cost-benefit of the more extreme measures that will have an undeniable, obvious, deleterious impact on the economy. One of my *very* smart friends recently asked, “Is there a middle ground where we accept seasonal flu-like deaths, don’t overrun the hospitals, and avoid wrecking the economy?” It is a rational question in the context of the very data driven world of Silicon Valley in which my peer group resides. However, are these the right questions to be asking in the face of COVID-19?

Probably not. The normal questions our minds think to ask are designed to fit the data we do happen to have in order to make a coherent story for ourselves and others. Since Aristotle, philosophers have made the point that we are deep-thinking animals. We humans can learn by reasoning. It is certainly easier to narrate backward in order to give ourselves an illusion of understanding, but we do have the capacity to reason forward.

In times of crisis, we must remember to rank beliefs not according to their plausibility (availability bias), but rather by the harm they may cause. Knowing that we cannot predict the future with great accuracy does not mean that we should not benefit from smart choices in the present. Managing regret must rise to the forefront of our thinking as we consider public policy around COVID-19. This must guide our forward thinking capacity and activity.

 In terms of public policy in the face of a novel pathogen, it is highly desirable to follow a much more sound policy of taking risks we can measure, as opposed to trying to measure risks that we are taking. What do I mean by that? Simply put, it makes sense to stay at home and largely isolate ourselves knowing that we are risking economic recession, rather than demanding that we increase testing to better measure the risk of continuing behavior as usual.

We look at the data from China and we see that the lockdown seems to have had an impact and turned things around there in about 5 weeks. Keep in mind that when China locks things down, it is a very different experience than what we can expect in more open societies like Italy and the United States. It is far more draconian than many of us would be willing to tolerate.

Even if the United States cuts its rate of transmission in half — a tall order — some 650,000 people might become infected in the next two months. This was the conclusion of researchers who used a database of known cases and Census Bureau transportation data to model how the outbreak could evolve. Models are just models, but it is consistent with what I was saying a couple of posts ago, that left unchecked, we’ll be at 1.2 million fellow Americans infected by May 1.

Now that New York is following California in terms of asking everyone to do their part by staying the fuck home, we should expect impact on the number of cases in time. If we all do our part to maximize social distancing, we can bring down the R to 1 or less. That might do the trick over the long run, but what I am really concerned about, as someone who has always considered NYC a second home, is what is going on in NY today.

What is concerning to me is that the Governor of NY seemed to struggle with the reality longer than he should have. Again, thinking out loud, perhaps Governor Cuomo fell into that trap of needing more data (here waiting on increased testing to scope the size of the problem). Whatever the case, it now looks like our friends and family in NY are in for a worse than Italy scenario.

Consider these two graphs of cases per 100,000 in population. The first one is from Friday’s data, the second is from Saturday’s data. By setting Day 1 in each of the individual regions (Italy, NY, CA, USA) based on when they had 100 cases respectively looked rather bleak for NYC. So, I added an additional set of data for Italy, setting their Day 1 where the cases were >= 900 to match NY’s Day 1 reality (they went from 60 to ~900 due to an increase of testing). Still not so good for NY in comparison.

… and a day later …

Now, a few items in NY’s favor:

  • NY’s overall population is younger than Italy (median age 38.2 years for NY vs. 47.3 years for Italy).
  •  NY has lower rates of lung cancer ( 59.3/ 100.000 New York vs. 67.8 / 100,000 Italy ) which might be a good proxy for compromised respiratory issues.
  • 34.7 ICU beds per 100,000 in the US vs 12.5 per 100,000 in Italy.

Generally, about 20 percent of coronavirus patients will require hospitalization. A quarter of those will be put on mechanical ventilator machines to help them breathe, and clear out their lungs. As of yesterday, more than 1,200 people had been hospitalized with the virus. About 170 patients were in intensive care units in city hospitals, according to The New York Times.

All we can do now is watch wait and see. We are placing a great deal of faith in social distancing. How well we practice it seems to be one of the key inputs into how much suck there is going to be. Please, please, please, take it seriously.

Absence of evidence is not evidence of absence. The risk from this pathogen may not be fully measured with testing. With that in mind, we cannot hide behind our lack of knowledge to adopt a position of fatalism. We can all do our part and increase our vigilance in our social distancing practice. Especially if we live in New York City.

It seems highly likely New York is in for the worst of this. If you have friends and family there, like I do, call them. Let them know you care. Offer to do what you can to help them stay in their homes. Do they need money because they are out of work due to the crisis? What other things do they need that you might be able to send to them?


[4:01 PM Eastern] UPDATE: I decided to also look at deaths / 100,ooo residents and NY is not looking so hot on that front either. I stated all entities on the day they reported > 10 deaths. Graphic below:

[ 23 Mar 2020 Updated Graphs]

One problem NY state seems to have is that their population density is creating a situation where they did in one day, what it took two days to do in Italy. At the current pace, you can expect by tomorrow mid-day sometime that NY will cross into the >= 100/100,000 confirmed cases territory. If we cannot tamp down the exponential growth, you could arrive at a situation of 1000s/100,1000 before the end of April. Not great news.

This image has an empty alt attribute; its file name is image-8-1024x566.png
This image has an empty alt attribute; its file name is image-9-1024x559.png

Leave a Reply