University of Washington IHME projections may be somewhat off. At least for Michigan

[Kaleva]

According to the model, peak deaths was supposed to have happened on the 7th and should now be on the decline but that's not the case. The website projected deaths for the state on the 10th is 111 but 205 actually died.

The site also predicted resource use in the state to have peaked on the 7th of April but new cases continue to climb at a great rate.

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

[Squinch]

1. Shocked. Shocked, I tell you.

[gibraltar72]

2. It is not lost on we in Michigan

That like Donnie Death the virus also doesn't read.

[Buckeye_Democrat]

3. That model makes some overly-optimistic assumptions...

... based on data from Wuhan, and that our stay-at-home orders will have similar effects.

As if anyone in the USA is coming close to matching China’s draconian measures at mitigating the spread of the virus!

The Washington Post had an article about those assumptions a few days ago, showing that the models created for Washington DC by local scientists have much worse projections.

The optimistic IMHE model is the only one used at the national level, unfortunately!

[Victor_c3]

6. Here in NW Ohio, I still see masses of people congregating in certain areas, like grocery stores

Everyone is touching fruit, touching freezer doors, and crowding together in the isles. The spread might slow a bit, but it’s still happening- I’m sure of it.

[Buckeye_Democrat]

7. No doubt in my mind. nt

[thesquanderer]

4. You can bet that the administration favors the most optimistic models. n/t

[Squinch]

8. I'm betting we will find out down the road that they PAID FOR that model somehow.

[Victor_c3]

5. As long as the rate of new infections is increasing, the rate of deaths will increase

I read somewhere that people who die typically die around 20-23 days after being infected. With that in mind, we won’t see a decrease until about 20 days after the rate of new infections levels off and starts to decline.

I don’t see that happening for a long time.

[snpsmom]

9. Yep

this model is pie-in-the-sky b.s.

[RockCreek]

10. What about this modelling site?

http://covidactnow.org/
Data by state for hospitalizations predicted and peak, county data in beta test stage.
Updated and readjusted every 24 h per the site.

[Buckeye_Democrat]

12. Thank you for that link!

[HarlanPepper]

13. That site is a mess

[Kaleva]

14. I think that'll be more accurate then the UW model

And the model you link to breaks it down by county which makes sense. The apex for Upper Michigan will come later then for the Detroit area in south east lower Michigan.

[HarlanPepper]

11. Oddly, it's been over estimating deaths in my state

Models are just that, so I wouldn’t put much stock in them unless you know what kind of data is being used.

[Kaleva]

16. What state is that?

[GulfCoast66]

15. It's impossible to model the entire US as a whole.

Lots of our country is nothing like China on which early models were based. The rural areas there are more densely packed than even our suburbs. Not to mention all the multigenerational households.

The only place I encounter others is at the grocery and I’ve found a few with little traffic. When I go in I feel like a surgeon. N95 mask and a baggy of alcohol soaked paper towel to wipe down everything.

Our dense cities are going to get hit hard. Social Isolation will really help in the suburbs and true rural areas.