How exactly are there "only" supposed to be 60,000 deaths?

[D_Master81]

I know many probably were like me and saw that University of Washington projection that Trump has run with that now projects “only” 60,000 deaths by August as highly optimistic. But I’m looking at the numbers and it would literally take almost an act of God for that projection to even be close to accurate. I haven’t been keeping close track but it seems we are having 2,000+ deaths a day for some time now so even a 50% drop is a thousand a day, 75% drop is 500+. At the current rate it’s over 10,000 a week, a 75% drop would still be 2500 a week. Even if that occurs it’s 10,000 a month for 3 months which would be 70,000 by August assuming a nearly immediate drop of that. I look for another week or so of peak until we see a reduction but still nearly 50% of peak for a few weeks if it’s anything like Italy and Spain. So I expect another week of 10,000 deaths followed by 2 weeks of 5000+. So right there is 20,000 that would be almost 60,000 by mid May. Tell me again how it’s going to be 60,000 by August? What right wing think tank sponsored this? I wouldn’t care but I’ve heard trump and others on the right bring this up to show how much lower estimates are now because of his actions and I’m like only 20,000 are gonna die in 3 months when it’s been almost 40,000 in 6 weeks basically.

[Bernardo de La Paz]

1. Only by flattening the curve, sustainably. But it could easily be four times that if "opened up". nt

[Amishman]

59. Unfortunately staying closed isn't sustainable either

We have no good options.

The only path is a tightrope between economic suicide and viral homicide.

This is going to hurt, I don't think we are even in the worst of it yet.

The official death counts are likely low. There is a very interesting example in a neighboring county. Berks county coroner's office reports 78 deaths, but the official totals reported by the state is only 40. https://www.readingeagle.com/coronavirus/berks-county-going-in-wrong-direction-with-biggest-leap-yet-in-daily-coronavirus-case-count/article_96e02906-817d-11ea-9087-53662f772f7e.html

[captain queeg]

2. Trump will keep moving the target no matter how it shakes out he'll declare victory:

Only X number have died. That’s better than a million! I give myself a 10. Of course if a million die that’ll need to be reshuffled or more likely covered up somehow.

[Ferrets are Cool]

5. That's one reason I hated to see that "million" number out there

The bastid can always look back, point at that number and declare HIS victory. If it were to exceed that, it would be Obama's fault.

[Bernardo de La Paz]

12. By August it should be clearer whether we have a nice German death curve or an Italian one

Neither is "great". In either case, funerals that touch extended families and friends are hard to hide, obits hard not to notice, etc.

When red hatters have lost a nephew or aunt and a bowling buddy and a guy on the same shop floor, even they will notice. And remember.

[Crunchy Frog]

20. Probably different curves in different regions of the country.

Also may be ebbs and flows in many regions, especially ones that try to reopen prematurely.

This country is far more expansive than countries like Spain and Italy, so the course of the pandemic will likely be more variable and complex.

It will also likely take a few years to get accurate death estimates.

[SoonerPride]

3. It is even crazier than that!

https://covid19.healthdata.org/united-states-of-america

It shows that August 4 the total will be 60,308 dead.

But it is bizarre because it shows the country reaching 59,184 by May 10th and then for the rest of May, all of June, and all of July only have 800 more people die?

In what universe are they talking about?

The level of deaths is essentially FLAT for 3 months.

On what planet does that make sense?

There will be 100,000 dead by August 4th
And 200,000 by Election Day.

If not more.

[qazplm135]

11. I don't know if it will be that much

but I would expect it to be well over 50K by May 1. Even if the peak was today, I would expect another 15K to die in the next two weeks (which is only a little over a 1K a day). I would then expect at least another month of hundreds per day bare minimum. Probably two months. I would expect to see something around 75K-100K by mid summer. After that, there's just a whole lot of uncertainty. Will there be a second wave? Will there be an effective treatment ready? How soon will social distancing fall apart and where?

Models aren't designed to be perfect, just to give us a ballpark idea.

200K is pretty unlikely at this point not thanks to Trump, but to many of the Governors who have stepped up.

[Celerity]

21. 60,000 seems really low, I'll be surprised if the US does not pass that within the next 2 to 3 weeks

I also think the death counts are being massively manipulated on a number of fronts.

https://www.worldometers.info/coronavirus/country/us/

[uponit7771]

35. +1

[LisaL]

4. Clearly there is going be a lot more.

We are already at over 36,000.

[Marie Marie]

7. And I'm sure that number is under the actual number.

[brush]

16. Yes, a lot more. There were 4500 yesterday, the highest daily total so far. At that rate it's 31k...

a week. We could be over 60,000 by the end of the month.

[onenote]

24. The 4500 included 3778 deaths that occurred between March 11 and April 14.

It was not a one day total.
https://www.worldometers.info/coronavirus/us-data/

[ProfessorGAC]

27. Still...

...with a first death on Feb 29, the smoothed out average is 743/day.
That includes that first couple weeks where the daily average was around 15.
So, taking off those 14 days, we're at pretty close to 7k/week.
We're likely peaking now, and probably so for another week.
Hard to envision <9,000/week for the rest of April.
With back-reporting, as you suggested, I can see bumping against that 60k value by May 1.
I believe distancing, sanitizing, etc. HAS to be helping, so a slowing is likely. But in the next 12 days? Unlikely.

[brush]

50. Sorry, it was only 2000 for the day. Only.

[onenote]

52. How do you get that number?

The number reported: 4591. The number that were previously unreported deaths occurring as far back as March 11: 3778.
That's 813 deaths not potentially attributed to an earlier period. So how do you get to 2000?

[brush]

53. Google. It's been near that daily for a while now.

[malaise]

54. almost 39,000 now

will be 41,000 by tomorrow night
At this rate I'll say easily 100,000 by the end of the month

[RestoreAmerica2020]

6. A reporter needs to ask malevolent in WH, JUST 60,000? What do you mean by JUST?

Last edited Fri Apr 17, 2020, 11:10 PM - Edit history (1)

..Other qestions: Can you explain "just" to families of the 40,000 PEOPLE that have died under your watch? And, if it were Ivanka, her children would it be "JUST" or would it matter then if it were your children, your wife that died of covid19?

Damn, he should be forced to physically sign approx 40 K [are we at 40k persons?have to check worldmwter] condolences send out to individual families.

Yes, Jake Tapper I'm looking at you--God forbid you challenge the guy [tapper recebtly said he's not there to coreect him, just report (his lies?) Tapper and other reporters just siiting on their nalgas [butts] are cobardes; imo that is.]

[Squinch]

8. That model is ridiculous. Has been since the start.

[former9thward]

9. Now the U of WA is a right wing think tank?

Heard everything now...

Their projection is 60,000 by May 19. After that it will basically level off for the rest of the summer with almost no additional deaths. Their model also has a certain amount of statistical uncertainty in it like every other model ever. By May 19 there could be as high as 138,727 deaths or a little as 34,052 deaths. For the rest of the summer it could be as high as 140,000 deaths or a low as just over 34,000. But the most likely number is the 60,000 number

https://covid19.healthdata.org/united-states-of-america

If you think the U of Washington is a right wing CT group maybe you should post a model showing what you think.

Their modeling has certainly more accurate than the absurd 2.2 million dead that was put out just a short time ago by many. Or some who posted that tens of millions would die and there would be almost no one left to vote in the November election.

[D_Master81]

13. 2.2 million was if nothing was done

And considering the numbers we’re seeing after extreme measures have been taken I doubt it’d be far off. I guess we’ll find out how it does. I find it hard to believe that less than a thousand people will die over multiple months.

[former9thward]

31. That is a false assumption.

Even if government did not do anything at any level people would do social distancing if hundreds of thousands and millions were dying. So "if nothing was done" was always a false assumption put out by people trying to scare people and get attention.

[greyl]

15. "By May 19 there could be as ... little as 34,052 deaths."

Two current aggregating websites sourcing data from CDC, WHO, CSBS, 1P3A, JHU show deaths right now at 32,823 or 37,054.

Also shows new cases have been about 30,000 per day for the past two weeks, with 32,557 new cases reported yesterday, April 16.

https://infection2020.com/

https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6

And that's with testing rates being criminally low.

[former9thward]

29. Models always have a low and high estimate.

including any that you may present.

[greyl]

55. Your reply is totally irrelevant. I posted data showing how innacurate your favorite model is.

Since the low estimate has already been surpassed, one should stop referring to that model as being accurate.


+6171 deaths reported in the past 24hrs.

[former9thward]

57. 6171 deaths where?

Please link to that. The model has been entirely accurate. Your comments show you don't understand how models work at all.

[greenjar_01]

18. What?

No stock tips?

[Squinch]

26. Here are experts explaining why the IHME estimates are crap:

https://www.democraticunderground.com/?com=view_post&forum=1002&pid=13300791

From the article:

According to a critique by researchers at the London School of Hygiene & Tropical Medicine and Imperial College London, published this week in Annals of Internal Medicine, the IHME projections are based “on a statistical model with no epidemiologic basis.”

[former9thward]

30. I don't click DU blue links.

But if you have a another estimate, with specific days, like the U of WA does, why don't you tell us? Then we can compare....

[Squinch]

32. Lol! Seriously?

Why is that? You don't trust the site?

[former9thward]

36. Post your estimate with specific days.

Why won't you do that if you think the U of Washington is so terrible? Let's compare....

[Squinch]

37. Wow. You're really invested, aren't you? But the discussion here is the validity

of IHME, which is a crap model that is being idiotically used to formulate public policy, not what my estimates are on specific days.

I imagine your ongoing position will be that, if I don't give you MY estimates for specific days, that is proof that the IHME model is totally valid and you totally won the argument that is happening in your head. Which is ... odd.

But have at it.

[Bernardo de La Paz]

43. Why on earth not? Seems highly illogical and self-defeating. . . . nt

[former9thward]

48. I asked the poster to post a model he thinks is correct.

Which he did not. The poster just threw stones at the U of Washington model without offering anything of his own. Maybe an alternative model was not offered because both Cuomo and the governor of NJ said today the curve was going down -- exactly at the time the U of WA said it would.

[Bernardo de La Paz]

49. I asked "why not?" after you wrote "I don't click DU blue links." I'm stonkered why you might even

I'm stonkered, mystified, and at a loss as to why you might even consider not clicking on a DU link. To say the least, I'm curious.

[obamanut2012]

40. +1

[Bernardo de La Paz]

10. There are already 37,000 deaths. It won't be 25% of the current rate for 3 months

The rate of new deaths seems to have declined to less than 7% per day, a good sign. If it is 6% for a week, then 5% for a week, the 4, 3, 2, 1 ... six weeks.

1.06 ^ 7 = 1.50. Times 37,000 = 54,600
1.05 ^ 7 = 1.47. Times 54,600 = 80,700
1.04 ^ 7 = 1.31. Times 80,700 = 106,000
1.03 ^ 7 = 1.23. Times 106,000 = 130,000
1.02 ^ 7 = 1.15. Times 130,000 = 149,000
1.01 ^ 7 = 1.07. Times 149,000 = 160,000

If we do better: Four weeks ago there were 100 deaths. If we posit we are at the peak and can have a four week decline ahead of us, ...

1.055 ^ 7 = 1.45. Times 37,000 = 53,800
1.040 ^ 7 = 1.31. Times 53,800 = 70,500
1.025 ^ 7 = 1.19. Times 70,500 = 83,800
1.010 ^ 7 = 1.07. Times 83,800 = 89,700

(I have retained extra digits in my calculations, so yours probably won't match digit for digit but should be very similar).

So my best case is at least 90,000 deaths. I think it is more likely to be the first scenario, perhaps 160,000 deaths. But it could easily be more depending on how red hatters roll the dice in the next 10 to 14 days. https://www.democraticunderground.com/?com=view_post&forum=1002&pid=13306154 We could be soon into a second wave, a red hat wave and go beyond 160,000.

[Demsrule86]

17. Very impressive calculations...and I think very likely to happen...the best if we continue to

flatten the curve and worse if the idiotic red states just let Covid spread...dumb dumb DeSantis opened the beaches today.

[Blue_true]

22. Your numbers look right if we don't open up too fast.

The thing that may keep infection and death numbers down overall is if the big problems shift to less populated areas. That would be hell for rural people, but fewer people in the country will die

[Squinch]

25. This. Thank you. Watching the trends, this is just about the same as my calculations.

And yes, the push to re-open will revise that upward. Only question is how much.

I've been called a ghoul here, and I've been told here that I am cheering for more deaths, when I quoted similar numbers a while back. Also for saying IHME is crap and for questioning their motives for putting out such crap.

Time tells.

[ProfessorGAC]

28. My Numbers Are Similar

The model says 60k by May 19. I'm calculating that as 2 weeks late for 60k.
So, closer to 70k by May 19.
A guestimated drop projects an average of declining weeks of 1,400-1,600/week.
So, my guesstimate is mid-80s.

[Marthe48]

41. What do you call your calculations?

I had some algebra, some geometry. Is this a statistical formulation? Thanks

[Bernardo de La Paz]

42. It's exponential growth with a few suppositions stated

The ^ symbol should be read as "raised to the power of". I'm using 7 because I'm am supposing a week at each growth rate, as a simplification. The base is the rate of growth. 1.0 would be no growth. Less than 1.0 would be a decline in numbers. Total deaths don't decline, they can only accumulate. The rate can decline to zero growth.

Imagine a 7 % daily growth rate (about what it is now), with the current death total what it is now (about 37,000) as your 1.0 start.

Day 1: 1.07 (7% more total deaths at end of day 1)
Day 2: 1.07 x 1.07 (because the 7% contributes to more growth)
Day 3: (1.07 x 1.07) x 1.07
Day 4: ((1.07 x 1.07) x 1.07) x 1.07
and so on.

Because you are multiplying each day by 1.07 it gets raised to the power of the number of days. Repeated addition creates multiplication. Repeated multiplication creates exponentiation.

With more math, we can model a continuous smooth rate of decline instead of approximating with a stepped decline week by week.

Some background: https://www.democraticunderground.com/?com=view_post&forum=1002&pid=13289484

[Marthe48]

60. Thank you for explaining

I like reading your in-depth posts.

[Bernardo de La Paz]

44. Typo in first calculations (week 2). Rather than edit, I will repost now

1.06 ^ 7 = 1.50. Times 37,000 = 54,600
1.05 ^ 7 = 1.41. Times 54,600 = 76,800
1.04 ^ 7 = 1.31. Times 80,700 = 100,600
1.03 ^ 7 = 1.23. Times 106,000 = 124,000
1.02 ^ 7 = 1.15. Times 130,000 = 142,000
1.01 ^ 7 = 1.07. Times 149,000 = 152,000

152,000 quite possible and could be worse with a red hatter wave.

[Steelrolled]

14. They are taking real world numbers and applying them

to a fairly simple mathematical model, based on previous experience, and an answer pops out.

They will constantly tweak their model and feed new data into it, and get new answers. At the end, they will have a 100 reasons why they did or did not get it right. It is what economists have been being doing for ever.

[Blue_true]

19. It looks like we will exceed 60,000 within the next three weeks.

We are only 21,000 under that number and are losing more than 2,500 people per day. The level of incompetence and lack of compassion from Trump is stunning, he is more concerned with instigating insurrection than he is about how many Americans are dying from the virus.

[spanone]

33. that's when trump stops counting.

[uponit7771]

34. People will stop dying in August, that's the IMHE calculation

[shockey80]

38. I was about to post the same question.

It's very hard to believe.

[shockey80]

39. With the Republicans going on a death march.

Last edited Sat Apr 18, 2020, 10:07 AM - Edit history (1)

It makes it really hard to believe those number.

[dawg]

45. Back when that model projected 68,000 deaths, the website said the model assumed ...

current levels of social distancing maintained until the end of May.

Now that the projection has been lowered to 60,000, it says it assumes current levels of social distancing are maintained "until infections minimized and containment implemented".

In other words, this projection is based on an alternate universe to the one we are living in. In *this* timeline, social distancing is already eroding even as deaths continue to peak.

[Tom Rinaldo]

46. It won't help if Trump and his allies attempt shell games to obscure the truth for political reasons

The elections are over six months away. There aren't any baskets large enough to hide the truth under for that long.

[elias7]

47. Deaths are being undercounted in large cities

Docs and nurses not allowed to speak out. People admitted to ICU’s with clinical diagnosis of COVID and do not really require testing, since it’s presumed. People die of pneumonia, heart failure, respiratory failure, multi-system organ failure, all from COVID, but their not being counted as COVID deaths.

We’ve turned into China, Russia, N Korea in our closed, hide the truth ways...

[OhioChick]

51. I wholeheartedly agree with you n/t

[Autumn]

58. Fucking with the numbers to keep the count low.

[Initech]

56. Chump will view any number of deaths under the original 2 million projection as a victory.

And then he will blame the democrats for over-inflating the statistics. Count on it.