IHME Model: Stephen King has been fired

The IHME COVID-19 model, notorious for “projecting” disease stats that look more like an apocalypse model than reality, has been adjusted once again. This time, though, instead of “fixing” it to make hindcasts even worse, they actually revised it such that numbers are now into the merely exaggerated.

I first noticed the model’s issues when examining March 31 “projections.” I chose that date because it had become the present, and thus a useful test of the model’s accuracy.

It failed. It’s death projections weren’t too bad, but the hospital beds needed numbers were insane.

Then:

3/31 projected beds: 95,581

3/31 actual beds: 38,743

For giggles, earlier this month (April) they revised the model and the 3/31 prediction (a hindcast, mind you) became…

107,638

With this latest, post-Stephen King, revision, 3/31 now hindcasts 41,070 beds needed vs. the actual 38,743 that I estimated. That’s much more reasonable.

The total projected deaths looks more like what I expect, too: 60,415.

Don’t get me wrong. For those that do get COVID-19 and go sour, it’s bad. Very bad. But that’s also true of the seasonal flu (which has already killed more people than COVID-19 is projected to kill). And the people who are particularly vulnerable to SARS-CoV-2 are the same people particularly vulnerable to the flu.

COVID-19 looks less like what most people see as a pandemic, and more like an extended “flu” season (yes, I know influenza and corona viruses aren’t the same thing).

If you found this post useful, please consider dropping something in my tip jar. I could really use the money, what with ISP bills, and general life expenses. And the rabbits need feed. Click here to donate via PayPal.
(More Tip Jar Options)

IHME Model Revisited

So how’s that IHME COVID-19 model doing today? When last we looked, it appeared to be run by Stephen King

For April 5, it projected:

  • Beds: 179,267 (126,649-225,921)
  • Deaths per Day: 1,529 (1,228-1,790)
  • Total Deaths: 9,893 (9,023-10,646)

Reality says:

  • Beds: 37,589*
  • Deaths per Day: 1,030
  • Total Deaths: 9,536

* I’ve estimated that using Georgia’s 19.3% of cases needing hospitalization. I applied that figure to US total cases (333,173), then subtracted the 17,177 recovered cases and 9,536 dead, neither of which need hospitalization for COVID-19 anymore.

DPD comes in almost 200 fewer than the low end of IHME’s range, but nearly 500 below the mean, which they’ve inflated by about 50%. Total deaths predicted is still high, but the range includes reality.

But beds are pure science fiction (there’s nothing scientific about that). They projected 4.77 times as many beds as appear to be in use. On this planet/spacetime continuum, at any rate.

Amusingly, they tweaked the model since my last reality-check. They had “predicted” that 95,581 beds would be needed on March 31. Now it claims we needed 107,638 (90,119-122,430) beds that day. With hindcasting getting worse, I have zero confidence in forecasting.

Holy shit, guys. If you’re going to change your model, change it to get closer to reality, not farther.

If you found this post useful, please consider dropping something in my tip jar. I could really use the money, what with ISP bills, and general life expenses. And the rabbits need feed. Click here to donate via PayPal.
(More Tip Jar Options)

Dear Up-State New York,

Your scumbag Governor wants you dead.

Cuomo plans to use National Guard to seize ventilators from upstate facilities
Gov. Andrew Cuomo said that he would use the National Guard to take ventilators and protective equipment from hospitals and other health care facilities in upstate New York where they’re not in use to deploy the lifesaving equipment in the New York City area.
[…]
“I apologize for the hardship to those institutions — ultimately there is no hardship, if you don’t get the ventilator back, I give you my personal word I will pay you for the ventilator — but I’m not going to let people die because we didn’t redeploy these ventilators.”

Let’s be clear about what this fucking scumbag is saying.

  1. Up-state Lives Don’t Matter. Paraphrasing: NYC people are dying. We’re taking gear that <our projections say you need. Fuck you; you aren’t real people.
  2. NYC residents are under hardship. Paraphrasing: If you up-staters die, that isn’t a hardship. I’ll just throw some money at the hospital — not the families of the deceased — that I should have used to buy ventilators years ago.
  3. An emergency on my part is an emergency on your part. Paraphrasing: My “prior planning” was to let up-staters die. You’re worthless; a waste of resources. Besides, you probably voted Republican.

You know… Back when NYUNSAFE passed, I recall a few misgivings from cops who didn’t look forward to being sent to confiscate arms in up-state NY. I hope the NGs give that some thought.

And should one or two up-staters die for lack of ventilators because more important NY City folk had them, I doubt that any Las Vegas oddsmaker will make book on Cuomo’s life.

If you found this post useful, please consider dropping something in my tip jar. I could really use the money, what with ISP bills, and general life expenses. And the rabbits need feed. Click here to donate via PayPal.
(More Tip Jar Options)

Some States Aren’t Shutting Down! The Sky Is Falling!

I see Jason Silverstein is a New Yorker who hasn’t grasped relative population density.

42 states now have stay-at-home orders for coronavirus. These are the 8 that don’t.
It may seem like the entire United States is shutting down and staying indoors due to the coronavirus pandemic. But some parts of the country are still holding off.

With the exception of South Carolina, which is a comparatively extreme outlier, those states are low population density region which are inherently “social distanced,” and — as you noted — have very few COVID-19 cases.

Let me put this in perspective with examples closer to home for both of us.

New York City: 27,751 people per square mile.

Camden County, GA: 86.5 people per square mile.

1 sq mi = 27,878,399.996383 sq ft.

Camden County averages 322,294 sq ft per person even without “social distancing.” That’s an average “social distance” circle with a diameter of 102,589 ft. In NYC, you have to work to get 6 ft of spacing. In Camden County, we have to work to get that close (I managed it twice 12 days ago).

Now, Wyoming on your list has a population density of 5.97 people per square mile: 4,669,749 sq ft per person. That’s a 1,486,429 diameter circle per person, on average. When I was out there, I went weeks without seeing another soul without trying.

The biggest city in Wyoming is Cheyenne; population density 1,990.14/sq mi. An average 4,459 ft diameter circle per person.

NYC? You’ve got 1005 sq ft per person. Around here, we call that a “tiny house.”

As of 4/4/2020, 8:25AM EDT:

South Carolina, #24
1,700

Arkansas, #35
738

Iowa, #37
699

Nebraska, #45
279

North Dakota, #48
173

South Dakota, #47
187

Utah, #28
1,246

Wyoming, #49
166

Given the population densities in those states, anyone not in a COVID-19-positive person’s immediate family has to work to get exposed to a known case. By comparison, NYC residents would have had to try hard to avoid exposure, and by all reports, few made the effort. Silverstein shouldn’t judge other states by the poor lifestyle and location choices of New Yorkers.

If you found this post useful, please consider dropping something in my tip jar. I could really use the money, what with ISP bills, and general life expenses. And the rabbits need feed. Click here to donate via PayPal.
(More Tip Jar Options)

HOAs Behaving B/a/d/l/y/ Criminally

I’ve seen some bad HOA reports, but this one takes the cake house.

“As a result of your actions, we will be invoking section II, Article 4 of the Bylaws, which allows the Board to foreclose on any property that is in violation.
[…]
If you are an owner, you or your tenants must vacate the property by April 30, 2020. You will still be responsible for any liens on your property, including mortgages.

Attorneys, please weigh in. I think I see a few issues here.

1. First off, the eviction notice appears to violate Tennessee eviction law, which I understand requires allow the resident an opportunity to “cure” the alleged offense, not to mention certain time frames (and specific reasons which don’t seem to include working from home in an emergency).

2. The HOA will “foreclose”? The HOA is the mortgage holder? That seems unlikely, as buyers normally get their own mortgages from whomever will give them the best rate.

3. Now, if the HOA believes that working from home during a pandemic somehow damages the Association so badly that they are entitled to these homes in recompense, wouldn’t they need to file a lawsuit and win, first?

4. If the HOA is the mortgage holder, and can theoretically foreclose, even that must follow state and federal law. This appears not to.

5. If the HOA is simply deciding to take the home, leaving the owner of record on the hook for the mortgage — presumably so it can resell the home to another sucker — that would seem to my unlawyerly mind to be theft (the unauthorized taking) and fraud (selling someone else’s property to another party). Isn’t that massively criminal?

6. Assuming #5, isn’t this a an organization (the HOA company/corporation) conducting an organized criminal action? Looks like a RICO case to me.

OK, attorneys; tell me where and how I’m wrong. Or right.

If you found this post useful, please consider dropping something in my tip jar. I could really use the money, what with ISP bills, and general life expenses. And the rabbits need feed. Click here to donate via PayPal.
(More Tip Jar Options)

I think we can safely say that model is flawed

The University of Washington’s Institute for Health Metrics and Evaluation has a model that projects potentially useful data like US COVID-19 deaths, and hospital beds and ICU units needed.

So how good is it?

For today, it had predicted:

  • Beds: 95,581
  • ICU: 19,638
  • Daily Deaths: 719 (638 – 781)
  • Total Deaths: 3,710 (3,629 – 3,772)

Meanwhile in the real world:

  • Daily Deaths: 439
  • Total Deaths: 3,580

As you can see, reality was not only considerably lower than the prediction, reality didn’t make into the low end of the predicted range. (ETA: I see the numbers have been updated. Daily deaths is now 648 which does bring it into the low end of predicted. Total deaths increased to 3,789.)

Beds and ICU units are tougher to figure. If anyone has a link to US data for those, drop it in comments below. But I can use Georgia as a proxy.

Georgia has 818 COVID-19 patients hospitalized, and 3,817 total cases. The United States has 180,789 total cases, making Georgia 2.1% of the total. If Georgia’s beds are also 2.1% of total beds, then…

38,743. A far cry from 95,581; a bit more than a third of the prediction.

Some studies have shown that roughly 80% percent of COVID-19 infectees are asymptomatic or quite mild cases. That would leave 20% who might need hospital beds: 36,158. Which is fairly close to my proxy estimate of hospitalizations, which works out to 21.4%.

Gloom and doom doesn’t help. Sure, you can use worst case projections to plan… for the worst, but when a model’s best case is nowhere near lower reality, it’s time to scrap the model, and start over.

If you found this post useful, please consider dropping something in my tip jar. I could really use the money, what with ISP bills, and general life expenses. And the rabbits need feed. Click here to donate via PayPal.
(More Tip Jar Options)

More COVID-19 Perspective

The United States is at week 9.5 of COVID-19, since the first reported US case. As of 3/28/2020, there are 116,057 cases.

At the same point in the current flu season there had been approximately 130,000 hospitalizations. The vast majority of flu cases do not require hospitalization. So serious flu cases exceed total (known) COVID-19.

That suggests a couple of things. Either COVID-19 is far less infectious than the flu, or it’s far less serious.

The fact is, unless we do widespread COVID-19 testing of the general population we have no idea how widespread it. The vast majority of known cases are mild to moderate; many people never show symptoms at all. In the case of the cruise ship Diamond Princess, we have a possible model.

  • Population: 3,711
  • Tests: 3,068
  • COVID-19 positive: 634 (17.1%)
  • No symptoms: 328 of the 634 (51.7%)
  • Deaths: 7If that is a good model for the general population, and our 116,057 known cases represent 48.3% of actual infections, then there may really be 240,284 cases out there. So the mortality rate, rather than the seeming 1.67%, is only 0.81%. The rate the Diamond Princess saw was only 1.1%.

    That suggests COVID-19 is less infectious than the flu. OTOH…

    If 17.1& of the population will be infected, then we will reach 56,259,000 cases. Compared to flu’s estimated 38,000,000 – 54,000,000 to date, it may be about as infectious, but much less severe in most cases. Remember: at the same 9.5 week point in each, more flu victims required hospitalizations than we have total COVID-19 victims.

    From March 15 through March 21 — one week — the flu killed 1,000 people.

    From January 21 to March 28 — 9.5 weeks — COVID-19 nailed 1,988; an average of 209.3 per week.

    If you found this post useful, please consider dropping something in my tip jar. I could really use the money, what with ISP bills, and general life expenses. And the rabbits need feed. Click here to donate via PayPal.
    (More Tip Jar Options)