Pages

Wednesday, March 18, 2020

Flu Fiction

Essential to review this episode of Mythbusters these days, where Adam and Jamie show how easy it is to spread a virus, and how best to mitigate it.  Works best with Firefox.  May require you to fill a captcha.

https://ww1.1movies.is/series/mythbusters-season-8/26250/33746/B6dUOl1g-watch-online-free.html

Saturday, March 14, 2020

Get With the Program on COVID-19






So I asked myself about five times this evening, “How many times do I have to deal with this bullshit today?”

Here’s a common meme that’s going around that’s trying to assuage everyone’s fears and claim that this is an overreaction.





Let’s break this down.

As of this writing ( March 14, 19:20 pacific time), there are 156,169 cases of COVID-19 worldwide. I’m using the numbers from Johns Hopkins University.

80,995 of those are in China.

Which means that since this idiotic meme was published the infection rate outside of China has gone, well, viral. So the number of people in China vs the numbers outside of China are meaningless about your chances of catching it.

Yes, there are mild cases. No, if you’re a reasonably healthy person, you’re not going to die from COVID-19. That’s not the point! There are a huge number of at-risk people who can easily die from this. Elderly. Immunocompromised. Asthmatics. People with COPD. Probably others that don’t come to mind at the moment.

Taking a single sample from the “worst” day in China is disingenuous, because China has managed to control the spread by using the power of its totalitarian communist dictatorship to seal off provinces, impose strict quarantine, and established curfews. Since they did this, the infection rate has leveled out and it’s stopped spreading. For a country of a billion people, to have a mere 80,995 reported infections from something this infectious is pretty damn impressive. Of course, it begs the question, can we really trust the numbers coming out of China? I don’t know if we can trust the actual numbers, but I’m reasonably sure we can trust the derivatives scaled over time. They might have ten times more people than they admit, but I’m reasonably confident they’ve stopped the spread of the virus and held down the number of new infections.

Now, we need to look at some realities.  

Worldwide, there has been currently 156,169 cases of COVID-19 reported worldwide. Of those, 79,785 have been reported resolved. Of the resolved cases, 5,830 died. Of the remaining 76,384, 65,594 are outside of China. That means for all the patients for whom the disease has run its course, 7.31% of them have died worldwide. Yes, these are mostly old people or people in high-risk categories.  

The question is, do we disregard the health of these people because we personally aren’t going to get very sick?

The Chinese experience informs us that without medical care, the mortality rate is about 12%. With medical care, this comes down to about 1%.  

A very conservative estimate based on the rate of infection in the United States, assuming we take no preventive measures to increase social distance is that we’ll hit 1 million reported cases sometime in late April. A less conservative estimate is that we’ll get there in early April. Again, if no action is taken, we’ll be on track to hit 10 million by mid-May.  

Actual mileage may vary. Personal habits, sanitation, hygiene all play into this. Americans are a standoffish lot, we’re reasonably clean and we generally eschew public transportation. All of these lower our chances of contracting the virus, yet in the last 24 hours the number of cases reported in the USA have nearly doubled.

This is the power of exponential compounding. It starts out slow, and then ramps up fast once it hits critical mass.

And this is why these idiotic memes comparing absolute numbers of flu cases and flu deaths are meaningless and counterproductive. The seasonal flu is endemic, and the rate of infection isn’t exponential. COVID-19 is epidemic, and its rate of infection is exponential. So the next time you see one of these that tells you how many people contracted the flu and how few died and try to show how paltry the COVID-19 body count is in comparison, remember that we’re in the early stages of this, it’s highly contagious, and an order of magnitude more deadly. It kills the same demographic as the flu, but it does so with ten times the efficiency. The flu is planned for, it doesn't come in a huge wave, we have seasonal models that predict how many people contract it and how many people will need medical care, and we're ready for that.




The United States has 924,107 fully staffed hospital beds. I don’t know, I’m guessing about 50% of those are in use on any given day. Suppose we get to the 10 million reported cases we’re expected to get in May if we do nothing. According to the meme above, 19% of the cases are moderate to critical, and will require medical intervention to some degree. That’s 1.9 million patients, more than twice the number than we have facilities available to treat. Based on the numbers from the epidemic worldwide, this means a death toll from 400,000 to 700,000. And that’s just May. 

Look, as contagious as this is, you’re probably going to get it. It’s eventually going to take its place along with the common seasonal flu. It will eventually burn itself out. The thing we need to work towards is to contain the spread to a manageable rate, to spread the infections out over time so our medical resources aren’t overwhelmed.  





Eyewitness accounts in Italy tell us that this has overwhelmed their system. Doctors are conducting triage, determining which patients can go home without medical care, which patients will survive with medical treatment, and which are going to die if they get treatment or not. It sounds heartless but it's normal when medical services get overwhelmed. Look, Italy is like the worst possible place to be letting a bug like this loose. They're the oldest demographic in Europe. They greet each other by kissing each other's cheeks. They have a personal space of about four inches. They all smoke. It's a dirty country. Literally everything you can think of to do wrong when there's a highly contagious respiratory disease going around is literally built into the Italian culture. So a LOT of people are going to get sick, and the medical services are overwhelmed. But it's a numbers game. Given an unrestricted contagion, it can happen anywhere. If we don't start isolating people who have been exposed right now, this calculus could be coming soon to a hospital near you.

The Science

Seasonal flu is an “all human virus”. The DNA/RNA chains that make up the virus are recognized by the human immune system. This means that your body has some immunity to it before it comes around each year... you get immunity two ways...through exposure to a virus, or by getting a flu shot.

Novel viruses, come from animals.... the WHO tracks novel viruses in animals, (sometimes for years watching for mutations). Usually these viruses only transfer from animal to animal (pigs in the case of H1N1) (birds in the case of the Spanish flu). But once, one of these animal viruses mutates, and starts to transfer from animals to humans... then it’s a problem, Why? Because we have no natural or acquired immunity.. the RNA sequencing of the genes inside the virus isn’t human, and the human immune system doesn’t recognize it so, we can’t fight it off.

Now.... sometimes, the mutation only allows transfer from animal to human, for years it’s only transmission is from an infected animal to a human before it finally mutates so that it can now transfer human to human... once that happens..we have a new contagion phase. And depending on the fashion of this new mutation, thats what decides how contagious, or how deadly it’s gonna be..

H1N1 was deadly....but it did not mutate in a way that was as deadly as the Spanish flu. It’s RNA was slower to mutate and it attacked its host differently, too.

Fast forward.

Now, here comes this Coronavirus... it existed in animals only, for nobody knows how long...but one day, at an animal market, in Wuhan China, in December 2019, it mutated and made the jump from animal to people. At first, only animals could give it to a person... But here is the scary part.... in just TWO WEEKS it mutated again and gained the ability to jump from human to human. Scientists call this quick ability, “slippery”

This Coronavirus, not being in any form a “human” virus (whereas we would all have some natural or acquired immunity). Took off like a rocket. And this was because, Humans have no known immunity...doctors have no known medicines for it.

And it just so happens that this particular mutated animal virus, changed itself in such a way the way that it causes great damage to human lungs..

That’s why Coronavirus is different from seasonal flu, or H1N1 or any other type of influenza.... this one is slippery AF. And it’s a lung eater...And, it’s already mutated AGAIN, so that we now have two strains to deal with, strain s, and strain L....which makes it twice as hard to develop a vaccine.

We really have no tools in our shed, with this. History has shown that fast and immediate closings of public places has helped in the past pandemics. Philadelphia and Baltimore were reluctant to close events in 1918 and they were the hardest hit in the US during the Spanish Flu.

The Misinformation

But why didn't we see this kind of panic with SARS?  It had more people infected and killed more than COVID-19 has (so far).

SARS was easy. It presented with very specific symptoms that were easily identifiable.  If you had these symptoms, you reported to medical services, got treated and quarantined. It didn't have a significant incubation period during which you were contagious. There was no reason to disrupt life for most of the country.  COVID-19, on the other hand, has a significant incubation period during which it's contagious, and the carrier doesn't even know they're sick. It presents with a wide variety of symptoms, as reported by the Italian medical experience, so most people think they have a cold or the regular flu and don't seek treatment, or necessarily even isolate themselves.  This makes the spread extremely difficult to identify and contain, unlike SARS.

Another common trope is that we don't know how many cases aren't reported, so the death rates aren't really as bad as they seem.  Bullshit.  It doesn't matter how many aren't diagnosed.  The ones who are diagnosed are showing an alarming trend. Reported cases in the US have more than doubled in the last 24 hours. At the current rate, in the US alone we're headed for a million sometime in April, as many as 10 million by mid-May if we don't quarantine to the level China has. Those are reported cases.  Yeah, there might be 20 million more that are unreported and of no concern, but the death rate of all the resolved cases that are reported is running around 7% at the moment. There's no reason to think those absolute numbers are suddenly going to change or that the number of critical cases that are currently going to be reported is going to change if we start counting total infections vs reported infections.  The unreported number of infections is a red herring.

The only way of slowing this down and managing it is to increase social distance. Communist China did it with totalitarian efficiency. In spite of what the Democrats and Liberals would like, we can’t get away with that here in the West. We have to convince people of the seriousness of the problem, so they’ll self-isolate and get the same effect China did. So stop the bullshit memes saying this is an overreaction. The numbers say it’s not, for those of you who can do a basic mathematical extrapolation. Stop saying it’s no worse than the flu. The numbers tell us it kills ten times as many people as the flu does. Stop saying it’s a media hoax. It’s not.  

Like a tidal wave in the open ocean, there’s a monster headed our way. I at first thought it was an unnecessary panic myself, until I started extrapolating the numbers and reading through the propaganda. It’s real. So for the sake of the elderly, the immunocompromised, the asthmatics and those with COPD, stop your bullshit, stop saying there’s no problem. Isolate, watch the numbers, take precautions when you’re out, and work with society to keep the deaths to a minimum. You may be safe, but to say that’s all we need to worry about is nothing but a big “fuck you” to millions of Americans who statistically are likely to die if we don’t keep a handle on this.







Sunday, March 1, 2020

Electoral Apportionment

A bone of contention these days is the Electoral College.  The Democrats want to get rid of it because, well, Democracy.  Republicans want to keep it, because it's one of the cornerstones that keeps the country together.  The common mantra is that if we went to a direct majority wins election that New York and California would decide the president.  That's a little bit disingenuous, but not far wrong.  Texas and Florida would also get a say.

Currently in 48 of the 50 states, electors are chosen by a winner-take-all system.  Which ever candidate wins the states gets all the electors for that state.  Each state gets a number of electors equal to the number of congressional seats they have in the House, plus two representing the Senate seats for each state.

Sidebar: Electors for each party for each congressional district are chosen at the party state conventions.  That elector gets to cast his vote only if his party wins the overall state election, even if the candidate lost the district that elector represents.

 That's all the US Constitution says about the electoral college. How the electors are chosen by the states is entirely up to the State legislatures.  There's no Constitutional mandate that states use the winner-take-all system.

The purpose of districting to determine a nationwide election is to actually increase the power of your vote. I won't go into the math about how it accomplishes this here, that was done very will in an old article by Will Hively in Discover magazine, called Math Against Tyranny. The problem with this, though is it only works if the districts are roughly equal size -- like congressional districts.  When your uber-districts are states, they no longer are equal size, and in fact dampen the effect of their constituent districts.

There seem to be a couple of different ways of dividing the electoral votes.  The winner-take-all system is one.  The problem with this is that it concentrates power in the urban areas of the largest states.  Win those urban areas, and you win the state.  By this method, California, New York, Texas and Florida together account for 28% of the voting power in the country, and much of that power is concentrated in some very small densely populated congressional districts.

Then there's Individual apportionment.  Under this scheme, Each Elector is selected based on the winner within his own district, without regard to how the candidate did statewide.  The two electoral seats for each state would then be apportioned to the statewide winner of the election. Under this scheme, you could easily have a minority of electors from a state that do not represent the candidate who won the state.

I suggested this recently on a public forum, and was shouted down by conservatives, who acted more like liberals do when you pee in their cornflakes.  I was called all kinds of vile epithets, accused of being a liberal and had a whole lot of people who apparently didn't read what I wrote and thought I was suggesting abolishing the electoral college.

So I ran the numbers.  I examined the 2016 election and calculated how the Electoral College would have voted if the Electors were selected by their individual districts, instead of winner-take-all.  The results were interesting.  Trump would have still easily carried the election with 291 vs 247, but not by as wide a margin.  The overwhelming significance of New York and California would have been dampened.  Florida's significance disappeared entirely, as it split pretty evenly between the two candidates.  Texas' significance also diminish, counteracting the significance of New York and California, as many of the border district voted for Clinton.

Of most interest was Minnesota, whose 10 electoral votes for Clinton came up 5 and 5 for the two candidates. The pro-Trump districts were weakly for Trump, and the pro-Clinton districts were strongly for Clinton, giving Clinton the state, even though more districts voted for Trump.

State Number of districts Voted Trump Voted Clinton at large number of electoral votes winner take all independently apportioned
Trump Clinton Trump Clinton

 Alabama
7 6 1 2 9 9 0 8 1

 Alaska
1 1 0 2 3 3 0 3 0

 Arizona
9 5 4 2 11 11 0 7 4

 Arkansas
4 4 0 2 6 6 0 6 0

 California
53 7 46 2 55 0 55 7 48

 Colorado
7 3 4 2 9 0 9 3 6

 Connecticut
5 0 5 2 7 0 7 0 7

 Delaware
1 0 1 2 3 0 3 0 3
 Florida
27 14 13 2 29 29 0 16 13

 Georgia
14 10 4 2 16 16 0 12 4

 Hawaii
2 0 2 2 4 0 4 0 4

 Idaho
2 2 0 2 4 4 0 4 0

 Illinois
18 7 11 2 20 0 20 7 13

 Indiana
9 7 2 2 11 11 0 9 2

 Iowa
4 4 0 2 6 6 0 6 0

 Kansas
4 3 1 2 6 6 0 6 0

 Kentucky
6 5 1 2 8 8 0 7 1

 Louisiana
6 5 1 2 8 8 0 7 1

 Maine
2 1 1 2 4 1 3 1 3

 Maryland
8 1 7 2 10 0 10 1 9

 Massachusetts
9 0 9 2 11 0 11 0 11

 Michigan
14 10 4 2 16 16 0 12 4

 Minnesota
8 5 3 2 10 0 10 5 5

 Mississippi
4 3 1 2 6 6 0 5 1

 Missouri
8 6 2 2 10 10 0 8 2

 Montana
1 1 0 2 3 3 0 3 0

 Nebraska
3 3 0 2 5 5 0 5 0

 Nevada
4 2 2 2 6 0 6 2 4

 New Hampshire
2 1 1 2 4 0 4 1 3

 New Jersey
12 5 7 2 14 0 14 5 9

 New Mexico
3 1 2 2 5 0 5 1 4

 New York
27 8 19 2 29 0 29 8 21

 North Carolina
13 10 3 2 15 15 0 12 3

 North Dakota
1 1 0 2 3 3 0 3 0

 Ohio
16 12 4 2 18 18 0 14 4

 Oklahoma
5 5 0 2 7 7 0 7 0

 Oregon
5 1 4 2 7 0 7 1 6

 Pennsylvania
18 12 6 2 20 20 0 14 6

 Rhode Island
2 0 2 2 4 0 4 0 4

 South Carolina
7 6 1 2 9 9 0 8 1

 South Dakota
1 1 0 2 3 3 0 3 0

 Tennessee
9 7 2 2 11 11 0 9 2

 Texas
36 22 14 2 38 38 0 24 14

 Utah
4 4 0 2 6 6 0 6 0

 Vermont
1 0 1 2 3 0 3 0 3

 Virginia
11 6 5 2 13 0 13 6 7

 Washington
10 3 7 2 12 0 12 3 9

 West Virginia
3 3 0 2 5 5 0 5 0

 Wisconsin
8 6 2 2 10 10 0 8 2

 Wyoming
1 1 0 2 3 3 0 3 0
DC 1 0 1 2 3 0 3 0 3
Total electoral votes 436 230 206 102 538 306 232 291 247

All in all, the net results don't change much. There are a number of red districts in the blue states that have no real say in the winner-take-all system.  These are offset by the minority of blue districts in the red states, but there's a lot more red states with one or two blue districts.

The net result is a more even distribution of voting power, with the 100 at-large electors who vote according to the whole state creating a bit of a smoothing function.  The dominance of the four big states is reduced, as their electoral votes will usually be split pretty decisively.  The net result would be that candidates would have to court more widely and consider the rural vote more.  Because of the way the electors were defined in the Constitution, I doubt the framers had in mind that Electors would be assigned on a winner-take-all basis from each state.

Click here for a map of the 2016 election by congressional district.

Unfortunately, there's no Constitutional prohibition from states casting their electors according to the national popular vote.  It's an exceedingly idiotic thing to do, and essentially the states who pass this have disenfranchised themselves.  I expect numerous court challenges based on lack of representation within the states who have passed this.  I also expect that if Trump quite likely wins the popular vote this 2020 election, you'll see these states falling over themselves trying to roll back the National Popular vote movement.