Coronavirus daily deaths trend

Update: July 8, 2020
I stopped updating this Google Sheet some time ago, as other sources meaningfully improved their data representations. The two that have most impressed me are the New York Times (for all my reservations about the Times’ reporting, they have a truly outstanding data visualization team) and Standard Code‘s COVID Mapping Project (they were, I believe, among the very first to begin providing useful and accessible state-level data). For international trends, Our World in Data is my preferred source.

We have enormous room for improvement in the richness of data being reporting, and I think our individual and collective decision-making abilities are meaningfully impaired by a lack of consistency in how information is being recorded and reported. Who has been hospitalized with Covid as opposed to for Covid? Who has died from Covid as opposed to with Covid? How many cases are asymptomatic vs severe? Etc.

The deeper I got into my exploration of the available data, the more I began to feel like the drunk man searching for his keys under the street light.

I couldn’t find any trend information comparing nation-to-nation on Covid-19 cumulative deaths per capita, so last weekend, I built a simple spreadsheet with Coronavirus deaths based on data from Worldometers. A few days later, I added trend lines for Japan with data from Covid19Japan.

The charts show deaths in 10 nations, with each nation’s trend line beginning on the date of the first death attributed to the virus. I wanted to see trends about deaths, thinking that while they probably weren’t being attributed in a consistent way within countries or from country to country, they were a better indicator than cases, given the inconsistencies in testing protocols.

Two notes:

  1. Given the vastness of the US and China, I think it would be better to look at their data by state and province, but as of this writing, I don’t have that daily death data easily accessible.
  2. It’s pretty clear that nations (and probably regions) have different standards for attributing deaths to Covid-19, so that’s worth remembering as you compare nation to nation.

The easiest way to view the charts is on a desktop via the spreadsheet itself. If you’re on mobile, you’ll want the Google Sheets app. I’ve embedded some of the charts below. If you have recommendations or comments, please comment below or share them on this Twitter thread.

Deaths per Million of Population (log)

Cumulative Deaths (log)


Why I’m being especially careful about Covid-19 (and you should, too!)

I haven’t read a succinct explanation for why one should take the Coronavirus seriously, so I thought I’d take a shot at writing one.

I have heard this a lot: “Over ten thousand people have already died from the flu this year. Coronavirus is nothing compared to the flu!”

I sincerely hope that Covid-19 turns out to be no big deal. That would be great! However, there is a meaningful chance that is not how things will shake out.


Some stats:

Covid-19 is 2x as contagious as the flu…

…and significantly more deadly. Those who catch it are something like 20x more likely to die from it.

Here’s how I interpret these two things:

The average person who gets Covid-19, gives it to somewhere between 2 and 3.1 people (let’s call it 2.5). And those 2.5 will give it to 6.25. And those 6.25 will give it to 15.6. And so on.

As I understand it, it’s more contagious because it has a longer incubation period. Something like two weeks vs a few days. And people are asymptomatic for longer, so they aren’t doing as much as they could/should to avoid spreading it to others—infected people are just naively lallygagging around smearing a deadly virus on their friends!

More contagious + more deadly = lots and lots and lots more risk.

The regular flu isn’t all that deadly. It knocks you out, but it is really unlikely to kill you (even if you are elderly). Covid-19 is different. At current mortality rates, if somewhere between 30 and 50 people get it, one of them is going to die. 1 in 400 for those under 40, 1 in 30 for those in their 60s, and 1 in 7 for those over 80. If every grandparent gets Covid-19 within the next few months, in 2020 we could lose 10% of all the world’s grandparents to the virus!

That sounds terrible, and especially so if there is nothing we can do about it. However, there is something we can do!

We are some indefinite period of time away from a vaccine (maybe a year, maybe more, maybe less), and while everyone is eventually going to be exposed to this, the more we can slow the spread, the better a chance we will give to those who are unexposed that they can be immunized via a vaccine or protected via herd immunity. Also, I imagine a year of experience will improve our treatments for those infected.

If, however, it spreads super fast—say, everyone on the planet gets it within three months—that’s 8 billion souls getting exposed, and 2% to 3% dying. At 2.5%, that’s (gulp) 200 million people dead.

“Only” about 500,000 humans die each year from the flu.

However, if we are super diligent about slowing the spread (canceling conferences, washing hands, elbow bumps, etc), we can dramatically decrease the mortality rate and the rate of infection. I assume that in a best case scenario, we can reduce the Covid-19 rate of transmission and mortality rate down to something like that of the flu. Maybe we could even eradicate it (I don’t know if this is a possibility, but I guess anything is possible).

Of course, I’m unlikely to die from this; I am a young (46 year old) whipper snapper. But I take it as my personal responsibility as a gentleman to do my best to postpone my inevitable exposure for as long as reasonably possible, thereby limiting my likelihood of transmitting it to others.

If I get it today, then someone who is somehow connected to me is going to die. But if I am not exposed to it for another year, then that’s probably (hopefully!) not going to be the case.

So wash your hands, people!