My Bet on Covid-19 and Why I Might Lose

My good friend and co-author, Charles L. Hooper, and I made a bet on the number of lives lost in America due to the Covid-19 disease.

Here are the terms:

We’re betting the grand total of $40 at even odds.

I bet that by the end of the calendar year, the number of deaths that can clearly be attributed to the disease will be greater than 100,000.

I talked to Charley on the phone this morning to see why he’s so optimistic. Here’s what he said: these are my words for his thoughts.

First, read Ari Libsker, “Corona Is Slowing Down, Humanity Will Survive, Says Biophysicist Michael Levitt,” CTech, March 13, 2020. By the way, it’s one of those rare cases where the headline dramatically understates the message. Levitt isn’t saying simply that humanity will survive; no one challenges that. He’s saying that the deaths will be way way lower than many are predicting.

Then, noted Charley, pay a lot of attention to this paragraph:

Quarantine makes a difference, according to Levitt, but there are other factors at work. “We know China was under almost complete quarantine, people only left home to do crucial shopping and avoided contact with others. In Wuhan, which had the highest number of infection cases in the Hubei province, everyone had a chance of getting infected, but only 3% caught it,” he explained. “Even on the Diamond Princess (the virus-stricken cruise ship), the infection rate did not top 20%.” Based on these statistics, Levitt said, he concluded that many people are just naturally immune to the virus.

Charley homed in on the quote about the Diamond Princess. He pointed out that that infection rate of 20% happened when people were in very close quarters for many days. So why, he asks, would we expect the rate to be higher in the larger world that is probably a little more “socially distanced” than that, even before the current level of social distancing? And by March 6, the death toll had reached 7. So of at least 696 people who got Covid-19, almost exactly 1% died.

So, argues Charley, let’s say we get an infection rate in the United States of 10% of 330 million. That’s 33 million people. He thinks it will be substantially lower. Let’s say it’s 6%. He thinks it’s reasonable that under 0.5% of them will die. That gives him 99,000.

I will update if Charley contacts me and tells me that I got anything wrong.

Here’s another part of the article that I highlighted to him that gave me hope that way under 50% of people would get infected:

There are several reasons for this [his optimistic prediction], according to Levitt. “In exponential growth models, you assume that new people can be infected every day, because you keep meeting new people. But, if you consider your own social circle, you basically meet the same people every day. You can meet new people on public transportation, for example; but even on the bus, after some time most passengers will either be infected or immune.”

Charley has at least persuaded me that it’s close to even odds and that I have only a slight upper hand. He thinks, of course, that he has the upper hand. But differences in opinion are what make horse races.

By the way, I bet a friend of Charley’s on Facebook, Jean Lockhart, $40 at even odds that over 50,000 people in the United States will die of Covid-19 by the end of the calendar year. That’s one check that I would LOVE to write out.

READER COMMENTS

READ COMMENT POLICY

R R Schoettker

Mar 16 2020 at 2:56pm

Given that this bet is regarding death numbers from a pandemic, offering to write a check after the time period elapses might not be a very prudent pay off option if one of the bettors is unfortunate enough to be one of the deceased. If both die that ‘solves’ (sic) the problem. Otherwise having a third party hold the funds with note of beneficiaries might be a good idea.

David Henderson

Mar 16 2020 at 3:08pm

The odds that both of us die are really low. Moreover, we are both married. Also, if even one of us dies, that one and his spouse will have way more issues than whether his estate is $40 heavier or $40 lighter.

Mar 16 2020 at 3:45pm

Positive thinking, always a good plan even if it is occasionally crushed by reality. Good luck on beating the ‘reaper’. I hope (like you I’m sure) that Charlie and Jean win the bets!

J Storrs Hall

Mar 16 2020 at 4:12pm

This is one HELL of a hard bet to adjudicate. Most of the COVID-19 mortality is among the elderly and those with pre-existing conditions. You can’t claim that anyone who dies who happens to have been infected died of the disease. On the other hand, you certainly can’t claim the opposite. Your best bet, IMHO, is to take the background death rate in the US (about 2.4M/yr) and see by how much the death rate this year exceeds it. Given that, I would lean slightly (but not strongly) towards Hooper’s side of the bet.

Mar 16 2020 at 4:17pm

You’re right. Charley and I are very close friends, though, and of all the people in the world whom I trust to adjudicate fairly, he is at, or close to, the top.

Also, let’s say that by one measure 90,000 people have died and by an extreme measure on the other end, it’s “only” 130,000 people. In that case, I’ll send him the check. I’ll simply be relieved.

Mark Z

Mar 16 2020 at 8:07pm

Hmm, if about 4 million people die in a year from something other than COVID-19, but 10% of them contracted it, and the average person who contracts it is positive for about ~5% of the year (2 and a half weeks), then we might expect around 20,000 people to die of other causes while COVID-19 positive. And some of those of course are car accidents, homicides, etc. So it would probably have to be at least somewhat close for this to be a deciding factor I think.

Apr 3 2020 at 9:00am

One more point about how hard it is: I got that 2.4M number on a CDC website. Turns out there’s another CDC website that says 2.8M. I haven’t done the deep research on which is closer to right, but from a simplistic standpoint there’s already an uncertainty of 400,000 in the number!

Josh

Mar 16 2020 at 4:20pm

I’m also encouraged by the WHO report on China. They make it very clear that nearly all cases were caused by close contact, and very rarely from brief contact (eg passing someone who’s coughing on the street). So a dinner party or a biogen conference are going to spread the disease. But grocery shopping wont.

Though they also note a lot of couples didn’t even give it to each other (which again might point to natural immunity).

https://www.who.int/docs/default-source/coronaviruse/who-china-joint-mission-on-covid-19-final-report.pdf

The disease they describe seems to be highly susceptible to social distancing. It doesn’t seem like a disease that can’t be stopped.

It also seems like there may be far more people with the disease than we think already. Donovan Mitchell (NBA player who got it) reported today that he has zero symptoms. He only got tested because he’s an NBA player; how many “regular folks” also have the disease and don’t know. Of course the more people who have it without symptoms, the lower the actual death rate will turn out to be if everyone does get it.

Charley Hooper

Mar 16 2020 at 4:52pm

Thank you, David, for the chance to explain my reasoning.

Overall, I think it’s highly unlikely that David will win this bet.

Even in perfect conditions for the virus on the Diamond Princess cruise ship, only 17% even contracted the disease. In Wuhan, China, where the disease originated (as far as we know), the infection rate was only 3%.

The ultimate rate in the U.S. must certainly be lower than both of these cases, but assume 1-3%.

Covid-19 has had a fatality rate of about 0.5% in China. Let’s say it turns out to be a little higher in China, perhaps 0.6% or 0.7%. However, it should be lower in this country due to our better medical system. Let’s assume a range of 0.2% to 0.5%.

With 330 million people, I expect the 2020 death rate from Covid-19 to be between 6,600 (lower infection rate and lower death rate) and 49,500 (higher infection and death rates) deaths, or roughly equivalent to a regular flu season, in which about 40,000 Americans die each year.

Brian

Mar 17 2020 at 5:09pm

Charley,

I think you’re right on with this and should win your bet easily (sorry, David). In addition to the arguments you present, current growth rates imply something well below even your lower bounds. Based on the current rate of slowing, we should reach a maximum around April 16 with less than 300,000 (confirmed) infected. Based on current death rates, which are probably too high, that would give <5000 deaths. I will be surprised if we even reach 10,000 deaths.

Luther Martin

Mar 22 2020 at 11:38pm

I’d go with the less than 100k deaths in the US. Based on the data that’s available as I write this, China has had about 2.3 deaths/M population from COVID-19. Right now, it looks like the US is at about 1.3 deaths/M. This number has to increase a lot to get to the catastrophic possibilities that we hear about. Italy is now around 91 deaths/M, which is much worse than any other country. But even at that death rate, we’d only get about 30k deaths in the US.

I worry more about the economic damage that our response is causing. The implicit value we’re attaching to each life saved is potentially huge. Unfortunately, the public-health goals and the economic goals don’t seem to always align.

Cindy

Mar 21 2020 at 9:55am

Why do you think the U.S. has a better medical system? I would say for this argument , and the % of infected and response by the countries, the reverse is true.

China testing was more thorough and quarantine mandatory.

I like your optimism but, would increase the % for the States in comparison to China.

Mark Bahner

Mar 16 2020 at 5:13pm

Hi David,

I don’t know about your interest in further betting, but I’ll give you even odds on up to $40, with me predicting less than 40,000 in the U.S. by the end of this year, as recorded on the Johns Hopkins website:

https://coronavirus.jhu.edu/map.html

Best wishes,
Mark

P.S. It may seem from previous comments like I’ve been all over the map on COVID-19. The situation is that the high death rate in the state of Washington shocked me…but the extraordinary response in the U.S. has also surprised me.

Mar 16 2020 at 5:49pm

You’re on. $40.

john hare

Mar 17 2020 at 11:51am

I’m thinking direct cause deaths may be a four digit number. Indirect though could have you winning if there were a way to keep track. Indirect as in auto fatalities that would have been avoided by flying. Not sure enough of my thoughts to do a bet.

Mar 17 2020 at 12:02pm

John,

Good points. But if we have to go as far as you suggest to get it to 100,000, I will be as happy as Charley that he won the bet.

Yaakov

Mar 18 2020 at 8:10pm

People are driving instead of flying? I thought everybody is just sitting at home. I would expect fewer accident casualties.

Mar 20 2020 at 8:22am

If the betting window is still open, I have another $40…this time I’m betting less than 4,000 COVID-19 deaths in the U.S. by December 31, 2020, as recorded by the Johns Hopkins coronavirus website:

I’m probably still taking advantage of you, but I need the money. 🙂

Mark

Rebes

Mar 16 2020 at 5:15pm

David,

Your and Charley’s different views are informative, but I wish you had found a more thoughtful way to express them than through a bet. A bet implies that you hope for the outcome you put your money on. That’s why people don’t bet on the opposing team, even when they expect their team will lose. I have no doubt that you don’t want more than 100,000 people to die to win $40, and since you don’t, cancel your bet and find out whether your fears materialize without money changing hands.

Mar 16 2020 at 5:52pm

You wrote:

A bet implies that you hope for the outcome you put your money on.

No it doesn’t. It’s often a way of hedging against bad outcomes. I bet that Trump would win when a colleague put up $100 to $20. I wasn’t happy. The $100 slightly reduce my unhappiness.

I have no doubt that you don’t want more than 100,000 people to die to win $40,

Exactly. So there goes your about implications.

Oh, and by the way, I sometimes do bet on the opposing team.

Mar 16 2020 at 6:11pm

You will think of my advice when you find yourself mentioned in an article with the headline “Economist bets $40 that more than 100,000 people will die from Coronavirus.”

It feeds into the stereotype that economists are heartless.

Walter Boggs

Mar 17 2020 at 11:45pm

I see headlines every day that are meant to imply something they do not in fact imply. I don’t think we’ll ever fix them all.

Henry

Mar 16 2020 at 7:45pm

Neither bettor has much confidence in their analysis. They bet only $40.

Mar 16 2020 at 10:16pm

It’s not the money, it’s the principle of the thing. (Also, I won’t have more than $40 by the end of the year, given the current path of the stock market. ;-))

Paolo Cecchi

Mar 17 2020 at 7:24am

Before accepting the bet you should define the concept of “clearly attributed to the disease”.

Consider the number of deaths here in Italy (13 March) https://www.epicentro.iss.it/coronavirus/bollettino/Report-COVID-2019_13_marzo.pdf

Deaths and pre-existung pathology

1,1% no pre-existing pathology

26,1% 1 pre-existing pathology

25,7% 2 pre-existing pathologies

47% 3 or more pre-existing pathologies

****

The pre-existing pathologies considered are:

37,3% ischaemic heart diseases

26,5% atrial fibrillation

8,2% stroke

76,5% arterial hypertension

37,3% diabetes mellitus

4,5% dementia

9,7% Chronic Obstructive Bronchopneumopathy

19,4% Cancer in the last 5 yrs

2,6% chronic liver disease

17,5% chronic renal insufficiency

In my opinion the distinction between deaths with coronavirus and deaths of coronavirus is specious: I assume that all these deaths are “clearly attributed to the disease”.

But what is your opinion?

Tyler Wells

Mar 17 2020 at 10:44am

Obviously cheering for Charley on this one. I think that the case against Charley is how the virus spread so far and so fast with only a 3% infection rate? I wonder whether the disease was already spreading well outside China before people began to recognize it as something different? Outside of China and Italy, the death toll is very tiny, so much that you would be sure that there are problems with measurement.

Mar 17 2020 at 1:35pm

Outside of China and Italy, the death toll is very tiny, so much that you would be sure that there are problems with measurement.

I don’t agree. The death toll (case fatality rate) in Hubei province in China was vastly higher than in the rest of China, because Hubei was the probably origination point, and so isolation procedures and treatment protocols weren’t established.

Italy has the following problems, relative to the U.S., with regards to the COVID-19 pandemic:

A) The population is older…median age in Italy of 47, versus the U.S. at 38.

B) Italy is a more densely populated country: 206 people per square kilometer, versus the U.S. at 40 people per square kilometer.

C) Italian dwellings and buildings have less floor area per inhabitant than U.S. dwellings and buildings.

D) Italian social customs involve much more physical contact than U.S. customs:

How to do Italian cheek kissing (not advisable during viral pandemics)

Mar 17 2020 at 3:19pm

You disagree that there are likely problems with measurement? With the symptoms being so similar to influenza (in most cases), and with the very small numbers infected, I would think that it is likely that there were many misdiagnosis (COVID 19 cases that were misdiagnosed as influenza), especially at first.

Interesting link, all are asymptomatic and, if it weren’t for the match against the Italian team, would not have been tested at this point. Perhaps some will subsequently show symptoms.

https://www.beinsports.com/us/laliga/video/valencia-cf-say-35-percent-of-tests-positive-/1435199

Mar 17 2020 at 11:30pm

Hi Tyler,

You write:

As I commented to Thaomas above, I don’t think anyone in the U.S. who dies with symptoms similar to COVID-19 from this point forward will die without testing. So I’m very confident Johns Hopkins’ numbers for deaths in the U.S. from COVID-19 will not suffer from failures to test people who die:

Mar 18 2020 at 10:18am

For sure, I would agree that virtually all deaths in the United States from now to whenever the public gets bored with this issue will be correctly identified as COVID 19 or otherwise.

I would also speculate that we will see spikes in the case counts as testing is rolled out. This will create the illusion (and hysteria) of a rapidly ballooning contagion when really it is just better measurement of existing contagion. My understanding is that there is no test for prior contagion (you may have had it in the past and, therefore, would be resistant).

Thaomas

Mar 17 2020 at 10:54am

Given the state of testing, will we ever know?

Mar 17 2020 at 1:13pm

I don’t see the current state of testing as having much effect on knowing the *death rates*. Testing is improving very fast…I’d expect the U.S. to have testing capability of over 100,000 tests per day, certainly by April.

And anyone in the U.S. going forward who actually *dies* from COVID-19 is likely to at some point have a medical professional look at them and say, “this could be COVID-19” and test for COVID-19. So I think the vast majority (95+ percent) of people going forward who *die* from COVID-19 in the U.S. will be tested for COVID-19.

Hazel Meade

Mar 17 2020 at 6:09pm

Based on these statistics, Levitt said, he concluded that many people are just naturally immune to the virus.

Maybe they are immune because they already had it.

I had something very similar to the described symptoms of COVID-19 last year. Lower respiratory tract viral infection that developed into bronchitis. No head cold. Urgent care gave me a prescription for an inhaler and just said “there’s something going around. It’s viral, and there’s no treatment so you just have to get over it. Most people get over it in 2-3 weeks”. I keep saying that I suspect that it was the same virus, or maybe an ancestral strain, but people seem to think that’s crazy. I got it from my kid who got it from preschool. So I’m thinking maybe the kids are immune because they already had it, along with a bunch of other people, and somehow it just didn’t hit the radar until now? Given what we know about the FDA and CDC regulations regarding testing, how would we be able to detect a new virus?

I know the genetic data indicates that it originated recently in Wuhan, but maybe it’s a recent mutation of an existing strain that was already circulating among humans. Maybe it mutated into a more aggressive or more infectious form.

Mar 18 2020 at 4:15pm

We don’t need to be previously exposed to a virus to avoid getting sick from it. Our immune systems fight off invaders all the time. Sometimes they are known invaders, which our immune system remembers from before, and sometimes they are new.

The symptoms of the illness we experience are usually the result of our immune system fighting the virus, bacteria, or fungus. Even when we don’t have symptoms, our immune system might be busy at work dealing with some sort of invader.

Mar 18 2020 at 8:23pm

COVID-19 has relatives – other coronaviruses – that are already common in humans and just cause ordinary colds. I think the evidence is pretty good that the current virus is fairly recently arrived in humans. Similar viruses have been found in bats (with 96% similarity) and it there’s mostly some intermediate animal that got it from bats and passed it to humans (pangolins being the most likely, but so far closer relatives of COVID19 haven’t been found in pangolins) and many of the early infected people in Wuhan happened to work at a live animal market.

If you’ve had a viral infection recently that is related to COVID-19 (such as another coronavirus) you’re probably more likely to be immune to it. It’s partly a matter of luck: if you’re immune system happens to target part of a protein in one virus that is identical to the corresponding protein in another virus, then when you catch the second virus, your immune system already knows what to look for.

David Mathis

Mar 18 2020 at 11:56am

I agree with the hypothesis Hazel puts forward. A couple who are very close friends had a respiratory illness in late December or early January in which they described the symptoms as being eerily similar to those of COVID-19. The kicker is he works in the aviation service industry and works closely with people who travel all over the world all the time…..I think it has been in the US for far longer than advertised.

Todd D Mora

Mar 18 2020 at 2:30pm

Great blog post! One concern, who is still writing checks? Can’t you Venmo or Cash the money to your friends? 🙂

Bruce StJohn

Mar 19 2020 at 8:15am

This is a great thread! The comments are very informative, too.

My $40 would be on fewer than 100,000 deaths. But please enlighten me. 100,000 deaths out of a population of 320 million is not a HUGE percentage, is it? That would be 5 – 10 times the number of influenza deaths, but still not huge.

Also, from a risk assessment standpoint, should we shut down the economy to reduce that number by maybe only 50% this year and then the other 50%+ in the fall?

There is a very nice piece of art online showing a sharp peak and a drawn out peak of what looks like a normal distribution. Well, there are other probability distributions. I am very concerned that we might encounter a bimodal situation in which this roars back big time in the fall.

Masks appear to be working in South Korea. I would propose that we get people into masks here. The CDCs story that they only work for health care workers and sick people has been shown to be questionable.

robc

Mar 19 2020 at 11:18am

If the hydrocloroquine treatment really does work (Big If), David going to be paying off a number of bets.

Mar 19 2020 at 11:55am

Let’s hope.

Mar 19 2020 at 12:37pm

I hope so to, it would mean the “cower in fear” phase is over and we can move on to “dont come to work if sick, see the doc, and get your 65 year old generic medicine” phase. Which isn’t a phase I saw coming, but probably should have.

Mar 19 2020 at 12:46pm

This does lead to an interesting question:

How come none of the models put in a probability factor for “the cure is on the shelf, we just got to find it”?

It is looking like p(shelf) is very large, if not quite 1 yet.

Alan Reynolds

Mar 19 2020 at 1:41pm

Let’s take Dr. Fauci’s arguably high estimate that the death rate for COVID-19 is 1%, and that for seasonal flu is 0.1% on average. Last season’s flu, says the CDC, hAD “35.5 million people getting sick with influenza… and 34,200 deaths.” To reach 34,200 deaths from COVID-19 we would have see 1/10th as many cases as we did for the flu, or at least 3.4 million.

This requires having no epidemic curve at all, but an exponential straight line with cases doubling at the same rate forever. In reality, the only epidemics to affect millions in a single country are influenzas. No deadly corona virus (including SARS and MERS) ever infected more than 81,000 in a single country before stopping, and that was in a country with more than four times the U.S. population. They infect thousands, not millions.

I’d be glad to bet that U.S. deaths from COVID-19 do not exceed deaths from seasonal flu this year (which looks a bit worse than last year). This is largely because we are flattening the curve like crazy (and no, it’s never too late), but also because curves are not exponential lines.

Abe

Mar 20 2020 at 3:54am

Lets say its the same death rate as the flu.

Lets just say its another type of flu.

So basically the death rate for flu just went up 100%!! fro 40k to 80k.

Its also adding another thing a person can die from. It is adding to the total percentage of things people can die from. So as its own it might not be higher than other things people can die from, but its definetly another serious addition to that!

It also moves pretty quickly and the ability to help those in need is lower.

For people with complications its probably a very high % of death rate. Flu does not have that.

Yogi

Mar 22 2020 at 7:16am

You are wrong.

The infection will not necessarily spread through the population gradually over the course of the year. It could spread very quickly. It is possible for the whole population to be infected in a couple of months.

On the grand princess, the number of serious cases is about 3x the number of deaths… If we run with your numbers for the US population, that is 300K-500K people who will need hospitalization in a span of a few weeks. Do you think the system will really be able to save over 2/3rd of them ?

Also, I think a fatality rate of 0.5% and a 6% infecton rate are both serious underestimates.

Take the example of Bergamo in Italy.

The town of Bergamo had 164 fatalities in the first two weeks of March. This is 0.14% of the population of the town… in just two weeks. In a normal 2 week interval, about 0.05% of the population dies. The extra fatalities have not all been attributed to the virus but they almost certainly are. To put this in perspective, it is the equivalent of 300K extra people dying in the US in a 2 week interval.

And the deaths in Bergamo didn’t start in March. Funeral parlors had informed the mayor that something is going on in January… The Government disregarded it.

And we are not done yet. In the third week of march, the local newspaper has increased the number of obituary pages by a factor of 6-10.

So, yeah, if it spreads uncontrolled, it is going to kill a heckuva lot more than 100K people. Hopefully, by the end of next month, the disease will run its course in Bergamo and we can take stock and figure out how fatal it really is.

Chris

Mar 23 2020 at 6:22pm

There’s one thing that might impact your numbers here: The average age of people living in a given town or region in Italy. I don’t know what it is in this particular case, but people in the Mediterranean region overall live longer. If all other factors are generally equal, an average age of 75 would have a much different percentage result than an average age of 60. We already know this from the data about the risks versus age groups. If everyone on the planet was 5 years old, maybe practically no one would die from coronavirus.

Ott

Mar 22 2020 at 3:42pm

If current trends continue over 140,000 people, (perhaps up to 250,000) will be infected in the next 10 days in the USA alone, never mind the 6-8wk modelling of this disease.

Peg your fatality rate at whatever percent you want 1%-5%. Globally the shit will continue to hit the fan and the USA will likely be front and centre to it all. Unless you believe Iran’s numbers… and Egypt’s; however I digress.
Good luck, stay safe. Peace out.

Mar 22 2020 at 8:03pm

The death rate exponential growth rate removes bias due to different testing rates and policies. I calculated it for several different countries before quarantines and found they all fell between 18 and 30% daily exponential growth rate. Pretty tight. Whatever happened on the Diamond Princess is of little interest compared to Italy almost 2 weeks into lock-down and seeing a 10% daily growth rate. I sure wish some white knight came down from the sky and fixed this, but I don’t see the US, with 30% saying they aren’t going to change their habits, getting below 15%, let alone 10%. But even if we did go magically go from 30+% daily rate down to 10% over the course of 2 weeks, we’re still going to see millions of cases and well over 100,000 deaths.

S. Korea has better Health Care than the US, and did a huge amount of testing (i.e., detected lots of mild cases), and is still seeing just over 1% mortality. US has less spare capacity than most developed countries. US could see 5% mortality. I would gladly take any bets on >100K death at any amount, so that I can transfer that money from whoever whats to take me up on that to a fund for the families of health workers who die of this.

And it will reach such a percentage in the population within a couple months that it’ll produce enough herd immunity that we’ll be done with the bulk of this by the end of the summer.

If 30% are immune, it just shifts the peak by a week or so.

Jenny

Mar 25 2020 at 4:12pm

The latest data on the Diamond Princess shows 10 deaths and 115 still sick with 15 in ICU so it’s too early to know the cfr. Some people spend 6 weeks or more in hospital.

https://www.worldometers.info/coronavirus/#countries

Terri

Mar 26 2020 at 10:57am

FYI The WHO’s most recent situation report has different numbers. They’re still reporting 7 deaths. I see a lot of links to Worldometers, but I don’t know where they’re getting their data, if not from The WHO. I thought perhaps some new patients had died, so I searched for any articles about new deaths on the Diamond Princess, but found nothing. Also seems odd that after all this time three passengers would die since the WHO last reported yesterday. I did find this, however:

https://www.nbcnews.com/health/health-news/10-dead-over-800-test-positive-coronavirus-outbreaks-princess-cruises-n1167716

This article combines the totals from the Diamond Princess and the Grand Princess.

The point is either Worldometers is wrong or The WHO is wrong (or lagging), which either way means an oft-referred to “legitimate” source is off by 30%.

Mar 25 2020 at 4:18pm

My guess USA has over 5,000 deaths by Easter. With swine flu 2009, it took 8 months to reach 5,000 worldwide deaths (WHO official Oct 2009). Remember for flu deaths each year, they are just estimates. They don’t actually count them. So future estimated deaths from this will be far higher than what is officially counted also. In Wuhan, funeral directors said most people died at home and were never counted.

John

Mar 27 2020 at 5:47pm

Professor Levitt also predicted a week or two ago that the total number of deaths from Covid 19 in Israel would not exceed 10, and more likely not 5 if restrictions were in place. As of the end of the day Friday, March 26, in Israel, with restrictions in place, the reported number of deaths is already 12.

genxmike

Apr 3 2020 at 3:18pm

40 deaths so far as of today. I think we can all agree that Levitt was really, really wrong in taking his model from the initial outbreak and attempting to apply it universally across the world. Israel is in deep trouble right now and I hope not much of their policy was based on his predictions or things will get far worse.

The US will be lucky to keep things under 100k at the moment and that would require far greater constraint than is currently being shown. Even places that shut down earlier, like NJ, are still seeing monster increases. Less affected states are just a week or two behind is all.

William Moore

Mar 28 2020 at 8:38pm

Did my eyes deceive me, or did somebody REALLY wager fewer than 4,000 deaths in the US by December back there?

And not on March 16th (though even that would be foolhardy), but just over a WEEK ago?

I mean, a fortnight is a long time in politics, but a week is an ETERNITY when your cases and fatalities are doubling every two and a half days…

Since the US has hit already 2,000 deaths several days before March is out, that is one bet I wish I’d taken. And 5,000 by Easter is also wildly under par.

I’m not sure if it’s hubris (“No, not here, not in our neck of the woods”), or just a simple inability to grasp how this thing is going to swamp health services (and that includes health services that are in far, far better shape than in the United States) and ICUs, and how it is going to have a terrible attrition rate on medical professionals, too.

The most bizarre thing is that to make this bet on March 20th, one would have had to ignore that Spain (into fullblown lockdown on March 14th) was already going ballistic, and France, too. Exceptionalism much?

Jerry Gaines

Mar 29 2020 at 8:52am

It’s now early on March 29, and the number of deaths in the US has doubled in the last 60 hours from around 1,000 to more than 2,000.

Everybody still comfortable with their bets that the US death total will be under 100,000 by December 31?

Mar 31 2020 at 7:43pm

There’s one bet out there, Jerry, that is already teetering on the edge, as my numbers off worldometers.info just hit 3,850, (and Johns Hopkins are hot on their heels at 3,810) and it is still (at least in the US) only March 31st. That was the individual who was “taking advantage” of David Henderson, but hopefully David will now being taking control of Mark’s $40, possibly within an hour or two. What was it P.T. Barnum said?

Apr 1 2020 at 12:27am

Sadly I have to agree with you. We’ll hit 4,000 no more than 12 hours from now (0420 GMT on Wednesday April 1).

Just based on some really simple extrapolations (using Worldometers as a starting point), I think we’ll hit 10,000 deaths before the end of the weekend. Those kinds of extrapolations are useless if you’re looking out a month or two, but for the next, say, 100 hours, they’re probably pretty accurate. And honestly whatever’s going to happen over the next 2-4 weeks is already set in stone.

Looking back through this thread, another commenter said on March 20 that the US is “flattening the curve like crazy”. Hoo boy.

Apr 1 2020 at 12:43am

And yep, we’ve hit 4,000. Currently at 4,054 as of 0440 GMT.

Apr 14 2020 at 4:52pm

Well, here we are, a fortnight on, Easter just about done and dusted, and the numbers are getting very ugly. Won’t be THAT long at the current rate before that $40 for the “less than 40,000” is very much in play. Probably before the end of the month. By that stage, the U.S. will be into some serious numbers in terms of the key parameter of deaths/million inhabitants. Then again, the UK isn’t looking too cheerful, either. What is it with countries that came late to the table and yet signally failed to take any benefit, learn any lessons, from the window of opportunity offered by seeing the plight of Italy, Spain, and France, a couple of weeks up the road? Hoo boy, indeed.

Comments are closed.