Corona model for PA just significantly improved

[psu00]

Don’t know if an error or new update (still says last update 4/1).

-Now says no PA ICU bed shortage (was 87)
-Now says PA surplus of approx 12,000 beds and surplus of 700 ICU beds.
-Now says 267 ventilators needed (was 904)
-Now says peak deaths on 4/15- 32 expected to die (was 79 on 4/19).
-Now total PA deaths predicted at 782 to be over (for the most part) on May 1 (was 2,023 deaths predicted).


The model still says last updated 4/1 but these changes just happened within the past 2 hours or so. Is this real or some computer error?

https://covid19.healthdata.org/projections

 

[blion72]

Don’t know if an error or new update (still says last update 4/1).

-Now says no PA ICU bed shortage (was 87)
-Now says PA surplus of approx 12,000 beds and surplus of 700 ICU beds.
-Now says 267 ventilators needed (was 904)
-Now says peak deaths on 4/15- 32 expected to die (was 79 on 4/19).
-Now total PA deaths predicted at 782 to be over (for the most part) on May 1 (was 2,023 deaths predicted).


The model still says last updated 4/1 but these changes just happened within the past 2 hours or so. Is this real or some computer error?

https://covid19.healthdata.org/projections

Gov Cuomo has said he cannot get a straight answer from statisticians. hard to make the models work when all you know is the # deaths. infection rate, incubation time, time with disease, etc are all just guesses.

 

[PearlSUJam]

Don’t know if an error or new update (still says last update 4/1).

-Now says no PA ICU bed shortage (was 87)
-Now says PA surplus of approx 12,000 beds and surplus of 700 ICU beds.
-Now says 267 ventilators needed (was 904)
-Now says peak deaths on 4/15- 32 expected to die (was 79 on 4/19).
-Now total PA deaths predicted at 782 to be over (for the most part) on May 1 (was 2,023 deaths predicted).


The model still says last updated 4/1 but these changes just happened within the past 2 hours or so. Is this real or some computer error?

https://covid19.healthdata.org/projections

That's assuming people are following the suggestions put forth. See Luzerne County, PA.

 

[HazletonLion]

700 0f 741 cases in Luzerne county are in Hazleton, Wolf playing Sgt. Schultz with the NY & NJ boarders. Thanks for killing Easter Pa.

 

[psu00]

Gov Cuomo has said he cannot get a straight answer from statisticians. hard to make the models work when all you know is the # deaths. infection rate, incubation time, time with disease, etc are all just guesses.

I agree- it’s all pretty much best guess estimates. That said, it’s nice to see a significant decrease rather than a huge increase. Hopefully it’s trending in the right direction and holds up over time.

EDIT-
-There was apparently a significant reduction in deaths the past 24 hrs in the US as well. One day isn’t a trend, but still good to see.

 

[wbcincy]

Don’t know if an error or new update (still says last update 4/1).

-Now says no PA ICU bed shortage (was 87)
-Now says PA surplus of approx 12,000 beds and surplus of 700 ICU beds.
-Now says 267 ventilators needed (was 904)
-Now says peak deaths on 4/15- 32 expected to die (was 79 on 4/19).
-Now total PA deaths predicted at 782 to be over (for the most part) on May 1 (was 2,023 deaths predicted).


The model still says last updated 4/1 but these changes just happened within the past 2 hours or so. Is this real or some computer error?

https://covid19.healthdata.org/projections

Those charts seem to show this being essentially over almost everywhere by May 1, so I'm going to have take it with a grain of salt.

 

[Steve G]

Those charts seem to show this being essentially over almost everywhere by May 1, so I'm going to have take it with a grain of salt.

want to be careful with the word/ idea of "over". After peak and if cases drop due to social distancing that will be a good thing. But, 95% of us will still be vulnerable to infection and the virus will still be freely circulating in the population. So we will need to continue to engage in social distancing (maybe in some what modified form) to prevent second wave of infection. Until there is an efficacious vaccine, or a whole lot of us get it and survive this will unfortunately not be over.

 

[BobPSU92]

We’re screwed.

 

[AWS1022]

I agree- it’s all pretty much best guess estimates. That said, it’s nice to see a significant decrease rather than a huge increase. Hopefully it’s trending in the right direction and holds up over time.

EDIT-
-There was apparently a significant reduction in deaths the past 24 hrs in the US as well. One day isn’t a trend, but still good to see.

The governor here in Kentucky gets on tv every night and goes over the latest number of cases and deaths...he said the last 24 hours was way down because a lot of testing places aren’t open on Sunday. Kentucky dropped from 100 cases the day before to 36 cases, but he said if all testing facilities were open it would likely be around 100 for yesterday as well.

 

[MJG-90]

There is such a lag time from test to results. PA shut down less than three weeks ago. Those that were catching the virus prior to staying at home were just getting their results last week and maybe just now. That’s part of the reason for the huge jumps we were seeing (along with just more people being tested). Hopefully that has run its course.

But then I went to the grocery store for an order pick up and watched a bunch of people go into the store. Not one wore a mask or gloves. So who knows how much longer this could last.

 

[mn78psu83]

Don’t know if an error or new update (still says last update 4/1).

-Now says no PA ICU bed shortage (was 87)
-Now says PA surplus of approx 12,000 beds and surplus of 700 ICU beds.
-Now says 267 ventilators needed (was 904)
-Now says peak deaths on 4/15- 32 expected to die (was 79 on 4/19).
-Now total PA deaths predicted at 782 to be over (for the most part) on May 1 (was 2,023 deaths predicted).


The model still says last updated 4/1 but these changes just happened within the past 2 hours or so. Is this real or some computer error?

https://covid19.healthdata.org/projections

The U of W model was updated yesterday. They must not have updated that piece of info at the time you checked on it. The projected number of deaths in the US by August 4 is also down dramatically - by around 12,000. The model assumes "full social distancing through May 2020."

https://covid19.healthdata.org/projections

COVID-19 projections assuming full social distancing through May 2020
Last updated April 5, 2020.

81,766COVID-19 deaths
projected by August 4, 2020

 

[interrobang]

A model can't be a good model when with only 5 days of additional data is changes that drastically.

 

[mn78psu83]

A model can't be a good model when with only 5 days of additional data is changes that drastically.

Yeah, I'm trying to be optimistic. Hope I'm wrong, but when this is all said and done they might have to rename this model the "Pollyanna Model." Of course, there are some that will be renamed the "Gloom and Doom" models.

 

[cjrugger]

A model can't be a good model when with only 5 days of additional data is changes that drastically.

The model didn’t have good data to begin with

 

[cjrugger]

Gov Cuomo has said he cannot get a straight answer from statisticians. hard to make the models work when all you know is the # deaths. infection rate, incubation time, time with disease, etc are all just guesses.

Funny, a bunch of people said that a couple weeks ago and were all labeled science hating virus deniers

 

[psuro]

There is such a lag time from test to results. PA shut down less than three weeks ago. Those that were catching the virus prior to staying at home were just getting their results last week and maybe just now. That’s part of the reason for the huge jumps we were seeing (along with just more people being tested). Hopefully that has run its course.

But then I went to the grocery store for an order pick up and watched a bunch of people go into the store. Not one wore a mask or gloves. So who knows how much longer this could last.

I wore one of my masks that I have when I ride my bike.....had no problem with people staying 6 feet away. Most were 15-20 feet away. How thoughtful of them.

 

[Nitt1300]

I work one of my masks that I have when I ride my bike.....had no problem with people staying 6 feet away. Most were 15-20 feet away. How thoughtful of them.

I'm wearing a mask and a full face helmet- people can't seem to get far enough away from me.

 

[psuro]

I'm wearing a mask and a full face helmet- people can't seem to get far enough away from me.

So, this is you? You are the Stig?

That's nice.

 

[rohrmd9]

You want people to stay really far away from you? Don't wear a mask. That scares people way more than any ominous looking mask does these days.

 

[interrobang]

You want people to stay really far away from you? Don't wear a mask. That scares people way more than any ominous looking mask does these days.

I'm awaiting the first reports of mask vs. no mask violence in stores.

 

[BobPSU92]

You want people to stay really far away from you? Don't wear a mask. That scares people way more than any ominous looking mask does these days.

And don’t shower.

 

[Howie'81]

Gov Cuomo has said he cannot get a straight answer from statisticians. hard to make the models work when all you know is the # deaths. infection rate, incubation time, time with disease, etc are all just guesses.

Models such as these are very complex. Often times if a single variable is incorrect, you don't get a reliable result. That's the problem with models.

 

[Howie'81]

Funny, a bunch of people said that a couple weeks ago and were all labeled science hating virus deniers

You said it brother. All I said then was that we don't know the denominator, i.e. number of people infected, to calculate a proper mortality rate. You'd swear I said water isn't wet. Since then, I've heard several scientists echo my exact point.

 

[Steve G]

You said it brother. All I said then was that we don't know the denominator, i.e. number of people infected, to calculate a proper mortality rate. You'd swear I said water isn't wet. Since then, I've heard several scientists echo my exact point.

directly due to failure to have wide spread testing. As RT-PCR testing capacity increases and antigen and antibody tests come on line will have better data on which to base models

 

[BobPSU92]

Since we’re talking models, isn’t it time for another Midnighter poll?

 

[mn78psu83]

U of W model was updated this morning.

COVID-19 projections assuming full social distancing through May 2020
Last updated April 8, 2020 (Pacific Time).

Numbers dropped significantly again.

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

Total deaths
60,415 COVID-19 deaths
projected by August 4, 2020

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

Projected PA deaths also with a dramatic drop.

Total deaths
969 COVID-19 deaths
projected by August 4, 2020

 

[bison13]

U of W model was this morning.

COVID-19 projections assuming full social distancing through May 2020
Last updated April 8, 2020 (Pacific Time).

Numbers dropped significantly again.

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

Total deaths
60,415 COVID-19 deaths
projected by August 4, 2020

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

Projected PA deaths also with a dramatic drop.

Total deaths
969 COVID-19 deaths
projected by August 4, 2020

I'd like to see us data without nyc. Two different timeframes I'm sure.

 

[Pardlion]

U of W model was this morning.

COVID-19 projections assuming full social distancing through May 2020
Last updated April 8, 2020 (Pacific Time).

Numbers dropped significantly again.

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

Total deaths
60,415 COVID-19 deaths
projected by August 4, 2020

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

Projected PA deaths also with a dramatic drop.

Total deaths
969 COVID-19 deaths
projected by August 4, 2020

These models are garbage. I predict 700 deaths, but reserve the right to keep changing my guess as real data comes in. I'll be right eventually.

The real shame here is that we hit the self-destruct button on our economy in response to this crap.

 

[bison13]

These models are garbage. I predict 700 deaths, but reserve the right to keep changing my guess as real data comes in. I'll be right eventually.

The real shame here is that we hit the self-destruct button on our economy in response to this crap.

That's why a us model just based off totals is crazy. Over half of cases and deaths are from one 50 mile radius.

 

[mn78psu83]

These models are garbage. I predict 700 deaths, but reserve the right to keep changing my guess as real data comes in. I'll be right eventually.

Yeah, I have followed several models and hear where you're coming from on this.

 

[NittPicker]

Funny, a bunch of people said that a couple weeks ago and were all labeled science hating virus deniers

I have no expertise in virology but it's fun to watch other people who also mostly have no such expertise arguing as if they do. Virus deniers originally said there's nothing to worry about. "It's in China, not here. Look at the projections!!" Then when it got to the US and a few places were hit hard, they said, "It's here but it's not that bad. Don't pay attention to the projections!!" Now that some models are predicting fewer fatalities it's back to, "Look at the projections!!" It works both ways so flip it around for the chicken littles. That's how it is in 2020. Confirmation bias rules the day.

I wish I was in the MRE business. No doubt there was some money to be made there.

 

[mn78psu83]

I'd like to see us data without nyc. Two different timeframes I'm sure.

To your point about NY and this model:

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

Total deaths
13,307 COVID-19 deaths
projected by August 4, 2020

That's 22% of all projected deaths in the US through Aug 4, 2020.

 

[CentexLion]

I have no expertise in virology but it's fun to watch other people who also mostly have no such expertise arguing as if they do. Virus deniers originally said there's nothing to worry about. "It's in China, not here. Look at the projections!!" Then when it got to the US and a few places were hit hard, they said, "It's here but it's not that bad. Don't pay attention to the projections!!" Now that some models are predicting fewer fatalities it's back to, "Look at the projections!!" It works both ways so flip it around for the chicken littles. That's how it is in 2020. Confirmation bias rules the day.

I wish I was in the MRE business. No doubt there was some money to be made there.

I don’t follow any of the models, but did anyone even come close to predicting this correct? It seems like alll models were a few magnitudes off. Glad it is so low, but really reminds me that all of our models seem to focus on worse and not median case

My model has been following about 25 friends in China. They are more crowded and less hygienic than we normally are (except for you nyc and New Orleans). Still, nobody I know over there even knows someone with the virus. And they traveled all over the country for their new year holiday. They had a few lock downs but mostly in hubei province. Now, they are all back to normal. Why wouldn’t we follow their path in rates? Maybe they get hit hard now again. We’ll see. I really like this social distancing thing and being one of only 10 at work while the other 400 stay home

 

[royboy]

..and the total USA deaths predicted by August 4 is now 60,415 (fairly recently it was 93,000+ and then 81,000+)

with the predicted daily peak of 2,212 just 4 days away

 

[cjrugger]

I have no expertise in virology but it's fun to watch other people who also mostly have no such expertise arguing as if they do. Virus deniers originally said there's nothing to worry about. "It's in China, not here. Look at the projections!!" Then when it got to the US and a few places were hit hard, they said, "It's here but it's not that bad. Don't pay attention to the projections!!" Now that some models are predicting fewer fatalities it's back to, "Look at the projections!!" It works both ways so flip it around for the chicken littles. That's how it is in 2020. Confirmation bias rules the day.

I wish I was in the MRE business. No doubt there was some money to be made there.

Not at all what I’m talking about. I’m talking about the people who said the data was garbage because we had no idea when the virus arrived or how many people had it at any point of time. But those people, who were correct, were lumped in with the virus deniers

 

[PSU2UNC]

A model can't be a good model when with only 5 days of additional data is changes that drastically.

Models are only as good as the data you construct them with. The underlying data here is, to be kind, terrible. The models are probably pretty good, but the data is changing very rapidly, which is why the projections are changing as well.

 

[LTFox14]

The numbers keep improving because of the social distancing. Had we not aggressively pursued that the numbers would be different.

 

[cntblvitsbutter]

The numbers keep improving because of the social distancing. Had we not aggressively pursued that the numbers would be different.

Not sure why that's so hard to understand. The whole point of social distancing and stay at home orders was to "flatten the curve" as much as possible which seems to be exactly what's happening.

 

[BobPSU92]

Not sure why that's so hard to understand. The whole point of social distancing and stay at home orders was to "flatten the curve" as much as possible which seems to be exactly what's happening.

Except for New Yorkers and New Jerseysans going to the Poconos and spitting in people’s faces.

 

[mn78psu83]

The numbers keep improving because of the social distancing. Had we not aggressively pursued that the numbers would be different.

Of course.