The new Corona virus, should we worry?

CutDownTheNet

Well-known member
Joined
Oct 4, 2018
Messages
1,239
Reaction score
246
Points
63
The recently revised Minnesota models under any of the 8 scenarios estimated ~ 100 deaths per day at this point and 1700 deaths by the end of May. Something seems amiss. Perhaps the curve is just time-shifted, but looking at relatively flat hospital numbers they will probably need to scramble. Cell phone data indicates MN social distancing hit its nadir many weeks ago.





View attachment 8069
They could certainly do a better job on their slides. For instance (on the slide quoted) when they say 1700 mortality through end of May, would that be the mortality that we *would have had* if there had been no stay-at-home mitigations? On another slide they show 3 curves, unmitigated, stay-at-home through mid-May, etc. They're certainly not going to hit 1700 deaths through May, that's for sure. Similar problems as the IHME model had, namely not fitting to current data properly.

I may have to download their code from github and take a peek, but I fear the code will be a black hole.
 

CutDownTheNet

Well-known member
Joined
Oct 4, 2018
Messages
1,239
Reaction score
246
Points
63
COVID-19 may be white noise in many places where people die of many other reasons at young ages. India is ranked 105th in population over 60, with only 8.2% of their population over age 60.

Looks like their average age is 29. Life expectancy of males 68.4 years.
Apparently, India's would-be COVID-19 dead are either already dead from something else, or else taking hydroxychloroquine to defend against malaria.
 
Last edited:

Pompous Elitist

Well-known member
Joined
Aug 18, 2013
Messages
13,882
Reaction score
729
Points
113
Apparently, India's would-be COVID-19 dead are either already dead, or else taking hydroxochloroquine to defend against malaria.
Yes, it’s hard to kill someone that’s already dead except in Hollywood. More seriously, India is probably just getting going. Due to a number of confounding factors hard to say what will ultimately happen until the numbers come in and I certainly won’t be looking at any SEIR models for answers. “We don’t know“ is ok to say.
 

Pompous Elitist

Well-known member
Joined
Aug 18, 2013
Messages
13,882
Reaction score
729
Points
113
He did say, "it could get worse before it gets better." Brilliant insight. Or, "it could maybe go away." Not so brilliant. "We'll see what happens." We saw.
Yeah DJT has said some really dumb things but to be fair DJT was as accurate as all the experts, armchair or otherwise. Guess high, guess low, go with your gut. Throw a dart.
 

Pompous Elitist

Well-known member
Joined
Aug 18, 2013
Messages
13,882
Reaction score
729
Points
113
They could certainly do a better job on their slides. For instance (on the slide quoted) when they say 1700 mortality through end of May, would that be the mortality that we *would have had* if there had been no stay-at-home mitigations? On another slide they show 3 curves, unmitigated, stay-at-home through mid-May, etc. They're certainly not going to hit 1700 deaths through May, that's for sure. Similar problems as the IHME model had, namely not fitting to current data properly.

I may have to download their code from github and take a peek, but I fear the code will be a black hole.
No, it was scenario 4, extending Stay Home Orders by 4 weeks. You have to go back to the April 28 model revision slide show to understand what they’re talking about. It appears the Walz administration ultimately chose scenario 5, extend SHO to end of May. This will push the curve and ultimate peak cases and ICU demand back somewhat. What may be interesting is their own model predicts worse mortality numbers (scenario 5) vs protecting only the vulnerable (scenario 3). It only pushes the peak back. ? why they chose that option, unless they are concerned about supply chain issues right now which is plausible. PPE, medicines, etc. Will any hospitals have staff left by July?


1908BE7E-7DA8-43F3-92A6-6ADA348BE9FF.png
 
Last edited:

GopherWeatherGuy

Well-known member
Joined
Oct 24, 2013
Messages
7,033
Reaction score
581
Points
113
No, it was scenario 4, extending Stay Home Orders by 4 weeks. You have to go back to the April 28 model revision slide show to understand what they’re talking about. It appears the Walz administration ultimately chose scenario 5, extend SHO to end of May. This will push the curve and ultimate peak cases and ICU demand back somewhat. What may be interesting is their own model predicts worse mortality numbers (scenario 5) vs protecting only the vulnerable (scenario 3). It only pushes the peak back. ? why they chose that option, unless they are concerned about supply chain issues right now which is plausible. PPE, medicines, etc. Will any hospitals have staff left by July?


View attachment 8077
Their model has been flawed, lacking of mathematical sense, and useless since the first release. Nothing new there.
 

MplsGopher

Well-known member
Joined
Nov 4, 2017
Messages
12,324
Reaction score
1,477
Points
113
^^^ has no clue how the model was made or what methods it was based on, has no access to the parameters or data it used, and wouldn't have been able to do any better himself, yet blathers out accusations like a coward. :rolleyes:
 

GopherWeatherGuy

Well-known member
Joined
Oct 24, 2013
Messages
7,033
Reaction score
581
Points
113
^^^ has no clue how the model was made or what methods it was based on, has no access to the parameters or data it used, and wouldn't have been able to do any better himself, yet blathers out accusations like a coward. :rolleyes:
My stance hasn't changed from the first release and I explained why then, nearly two months ago. MN keeps doubling, and tripling down on this model, and it makes no mathematical sense. I have a modeling background and absolutely could have done better with just an excel spreadsheet.

None of the MN predictions have been in the same stratosphere of reality to date. I don't need to know methods or parameters to know that it's a useless pile of garbage, and has been since day one.
 

MplsGopher

Well-known member
Joined
Nov 4, 2017
Messages
12,324
Reaction score
1,477
Points
113
My stance hasn't changed from the first release and I explained why then, nearly two months ago. MN keeps doubling, and tripling down on this model, and it makes no mathematical sense. I have a modeling background and absolutely could have done better with just an excel spreadsheet.

None of the MN predictions have been in the same stratosphere of reality to date. I don't need to know methods or parameters to know that it's a useless pile of garbage, and has been since day one.
Bald faced lie. You would've done it then.

Your model would've failed just like every other model has failed: because this is brand new! There was no prior information.

Also highly doubt you have the slightest clue about what math the model is based on, or how it works. Just because you know how to input weather data into a decades-long established weather model system and turn the crank, doesn't mean you know jack about math.


Done arguing with you about this today. Not going to let you drag me down like goldteam. You get the last word, if you want it.
 

GopherWeatherGuy

Well-known member
Joined
Oct 24, 2013
Messages
7,033
Reaction score
581
Points
113
Bald faced lie. You would've done it then.

Your model would've failed just like every other model has failed: because this is brand new! There was no prior information.

Also highly doubt you have the slightest clue about what math the model is based on, or how it works. Just because you know how to input weather data into a decades-long established weather model system and turn the crank, doesn't mean you know jack about math.


Done arguing with you about this today. Not going to let you drag me down like goldteam. You get the last word, if you want it.
As a meteorologist who dealt with numerical models every day of my career, and also helped to build one, I really have no clue. Many meteorologists also double major in math, because this degree requires a significant amount of math. Models are updated constantly, and new ones come out every few years.

I like to discuss data and reality. Of course you don't want to hear anything that doesn't fit your massive death and destruction narrative. What's obvious every day is that you are the most clueless poster on this board. There isn't even a close second.
 

Blizzard

Well-known member
Joined
Nov 20, 2008
Messages
3,445
Reaction score
48
Points
48
As a meteorologist who dealt with numerical models every day of my career, and also helped to build one, I really have no clue. Many meteorologists also double major in math, because this degree requires a significant amount of math. Models are updated constantly, and new ones come out every few years.

I like to discuss data and reality. Of course you don't want to hear anything that doesn't fit your massive death and destruction narrative. What's obvious every day is that you are the most clueless poster on this board. There isn't even a close second.
What he said.
 

GophersInIowa

Well-known member
Joined
Nov 21, 2008
Messages
25,167
Reaction score
770
Points
113
Their model has been flawed, lacking of mathematical sense, and useless since the first release. Nothing new there.
In your opinion, what's off about the model? Do you think things like the R0 and mortality rate that they're guessing on are way off?
 

GopherJake

Well-known member
Joined
Mar 21, 2009
Messages
16,641
Reaction score
256
Points
83
As a meteorologist who dealt with numerical models every day of my career, and also helped to build one, I really have no clue. Many meteorologists also double major in math, because this degree requires a significant amount of math. Models are updated constantly, and new ones come out every few years.

I like to discuss data and reality. Of course you don't want to hear anything that doesn't fit your massive death and destruction narrative. What's obvious every day is that you are the most clueless poster on this board. There isn't even a close second.
I honestly don't doubt your math acumen. But you better than anyone should know that there is sometimes a fine line between a driving rainstorm and a sunny day. Give me a radar map and I can probably predict tomorrow a reasonable % of the time without historical data, sophisticated modeling and Excel. Add 60,000 variables over which you have no control and things get a little more complex. Show your work like CutDown does. Give us the picture in 3 months. Get that right and we'll all advocate for you taking the helm in MN.
 

MplsGopher

Well-known member
Joined
Nov 4, 2017
Messages
12,324
Reaction score
1,477
Points
113
^^^
It's much, much easier to come on GopherHole OTB and lazily lash out at the Walz and Osterholm, with no evidence and putting in zero effort.
 

MplsGopher

Well-known member
Joined
Nov 4, 2017
Messages
12,324
Reaction score
1,477
Points
113
In your opinion, what's off about the model? Do you think things like the R0 and mortality rate that they're guessing on are way off?
Don't ask him that. He has no clue, off the top of his head.

He'd have to actually look under the hood of the models (if that information is publicly available), see what methods and parameter values were used, etc. That would require actual effort.
 

short ornery norwegian

Well-known member
Joined
Feb 9, 2011
Messages
8,917
Reaction score
646
Points
113
Minnesota Covid-19 Update - Tuesday, May 19. Data reported by 4pm the previous day.

Positive Cases 17,029 +665.

Deaths 748 +17.

Deaths at long-term care and assisted living facilities 608 +13.

Patients Hospitalized 545 +57.

Patients In ICU 229 (no change).

Total tests processed 161,835 +5,229.

Top Counties:
Hennepin 5,650 - 476 deaths
Stearns 1,782 - 10 deaths
Ramsey 1,720 - 74 deaths
Nobles 1,394 - 2 deaths

(note - that is by far the biggest one-day jump I remember for hospitalizations. something to watch to see how many wind up in ICU.)
 

GopherWeatherGuy

Well-known member
Joined
Oct 24, 2013
Messages
7,033
Reaction score
581
Points
113
In your opinion, what's off about the model? Do you think things like the R0 and mortality rate that they're guessing on are way off?
Those could be a factor, but just the basic math behind social distancing and what they've modeled doesn't make any sense. They keep modeling a delaying of the peak, and show very little flattening of the curve. Their different scenarios even show that social distancing isn't necessarily preventing deaths, it's just delaying them.

When they released their early April update, they even said within a 95% probability that we'll be at peak ICU demand beginning in mid May. At that time there was ~70 in the ICU, and they said we were going to need ~3000 ICU beds at peak. That was complete crazy talk, and it still is, especially when considering they extended the SAH order at that time.

This video is a great explanation of the math behind social distancing and how it flattens the curve. It's like the MN modelers have completely ignored basics behind this.

 

howeda7

Well-known member
Joined
Nov 22, 2008
Messages
41,900
Reaction score
1,710
Points
113
Sweden has the highest deaths/million in the world last week. Are we still emulating them?
 

GopherWeatherGuy

Well-known member
Joined
Oct 24, 2013
Messages
7,033
Reaction score
581
Points
113
I honestly don't doubt your math acumen. But you better than anyone should know that there is sometimes a fine line between a driving rainstorm and a sunny day. Give me a radar map and I can probably predict tomorrow a reasonable % of the time without historical data, sophisticated modeling and Excel. Add 60,000 variables over which you have no control and things get a little more complex. Show your work like CutDown does. Give us the picture in 3 months. Get that right and we'll all advocate for you taking the helm in MN.
There is definitely a fine line between a thunderstorm and a sunny day. However, today's radar tells you nothing about tomorrow's weather.

As for a prediction, I said yesterday we are either peaking now or will very shortly in MN. There will be no late summer peak as the MN models now suggest. That's based on a combination of what's already occurred in MN and elsewhere, demographics, population density, weather, and increased testing.
 

GopherJake

Well-known member
Joined
Mar 21, 2009
Messages
16,641
Reaction score
256
Points
83
There is definitely a fine line between a thunderstorm and a sunny day. However, today's radar tells you nothing about tomorrow's weather.
Nothing? :rolleyes:

As for a prediction, I said yesterday we are either peaking now or will very shortly in MN. There will be no late summer peak as the MN models now suggest. That's based on a combination of what's already occurred in MN and elsewhere, demographics, population density, weather, and increased testing.
Looking for something more specific. But I'll mark you down for no August peak. Kinda broad though.
 

balds

Well-known member
Joined
Nov 22, 2008
Messages
2,402
Reaction score
119
Points
63
He did say last week, not overall. I have no idea of what he is saying is true though.
Highest in EUROPE over the past week, based on a 7 day moving average. Deaths still going down in Sweden. Over the course of the pandemic Sweden is doing better than UK ,France, Belgium, Spain and Italy, but worse than neighboring Denmark and Norway.

"Sweden’s open strategy seems to have softened the blow on the economy, with growth shrinking much less than in Denmark and Norway in the first quarter"

 

CutDownTheNet

Well-known member
Joined
Oct 4, 2018
Messages
1,239
Reaction score
246
Points
63
Sweden has the highest deaths/million in the world last week. Are we still emulating them?
Maybe on the weekly totals.

But actually on the grand totals, no, San Marino is the country that has the largest deaths/million. As a state, New York beats San Marino by a bit, and San Marino beats New Jersey by a bit. The following chart shows all the countries plus all the US states that have a deaths/million of 80 or greater. I had it color coded (state vs country) in Excel, but unfortunately that did not carry over with the paste here. But you know which ones are states.

The US is trailing Sweden by 90 deaths/million. This has been the case for a while now. The US has been behind Sweden by a number ranging from about 75 to 100. I'd have to look to see whether or not the US is closing the gap on Sweden.

Country/StateDeaths per Million Population
New York
1,469​
San Marino
1,209​
New Jersey
1,176​
Connecticut
967​
Massachusetts
850​
Belgium
786​
Andorra
660​
Spain
594​
District Of Columbia
567​
Louisiana
555​
Italy
532​
UK
521​
Michigan
502​
Rhode Island
502​
France
429​
Pennsylvania
374​
Sweden
371​
Sint Maarten
350​
Maryland
344​
Netherlands
334​
Illinois
334​
Ireland
317​
Delaware
312​
Isle of Man
282​
USA (all states)
281​
Indiana
271​
Channel Islands
253​
Switzerland
219​
Colorado
213​
Montserrat
200​
Mississippi
186​
Luxembourg
174​
Ecuador
161​
Canada
157​
Georgia
157​
Ohio
147​
Bermuda
144​
Washington
134​
Minnesota
134​
New Mexico
129​
New Hampshire
126​
Portugal
122​
Virginia
122​
Nevada
119​
Iowa
118​
Alabama
103​
Monaco
102​
Missouri
100​
Germany
98​
Arizona
97​
Florida
96​
Denmark
95​
Vermont
87​
Peru
85​
Iran
85​
California
84​
Brazil
82​
Wisconsin
80​
 
Last edited:

MplsGopher

Well-known member
Joined
Nov 4, 2017
Messages
12,324
Reaction score
1,477
Points
113
Country/provincial borders are arbitrary boundaries that the virus does not obey.

The true way to do it would be the grid the habitable area of the world into square km/miles (whatever) and look at how many deaths/1000 population that live in each square area.

Where is that map? Would be interesting.
 
Top Bottom