Squaring the Corona-chan circle

Corona Chan, le coronavirus vue par les Otaku - le site du ...

Looking over the highly alarmist reports from the whorenalists of the lying (((media))) these days is enough to give one fits of rage-induced vomiting. And in some cases it is doing precisely that. I spent at least an hour on the phone recently with a friend of mine who has vast experience with mass-casualty disaster responses and deep expertise in precisely these sorts of highly complex scenarios, and he was practically spitting blood at how crazy the world has gone over something that simply doesn't merit the reaction.

But there does appear to be one major logical problem here, which we have to look at calmly and dispassionately.

By any rational measure, the epidemiological models used to formulate the various governmental responses from around the world have proven to be ridiculously wrong.

I'm not talking in terms of "off by a few percentage points", either. That's to be expected. It's normal to have a degree of error and uncertainty surrounding any mathematical model.

No, I'm talking about being OVER EIGHT THOUSAND PERCENT wrong.

Consider: Prof. Neil Ferguson of Imperial College London, one of the world's top scientific research universities, originally came up with a forecast that COVID-19 would kill 500,000 Britons if allowed to rampage through the UK unimpeded, and 2.2 million dead in the USA if America did the same.

Even His Most Illustrious, Noble, August, Benevolent, and Legendary Celestial Majesty, the God-Emperor of Mankind, Donaldus Triumphus Magnus Astra, the First of His Name, was taken in by these estimates. Even he has been using the "2.2 million" figure in press briefings to tell people that if he hadn't taken action, that is how many would have died.

Now, I happen to have a background in mathematics. I'm quite a bad mathematician, mind you - I last attempted stochastic differential equations nearly 14 years ago, and if you asked me to solve one today, I would simply throw my hands up in the air and tell you I can't do it. But I know enough reasonably advanced mathematics to understand how exponential curves and bell-curve distributions work - and I know plenty of financial mathematics at a highly advanced level.

Furthermore, I've seen financial models go horribly, dreadfully, spectacularly wrong - firsthand. I worked in banking and finance for nearly 10 years before my career took a sudden and, it appears, terminal trajectory off a cliff. I know how sensitive these models are to their inputs. If you put in garbage assumptions, you WILL get absolutely garbage results.

So when people start throwing around numbers like "TWO POINT TWO MEEEEELLION!!! WILL DIEEEEE IF WE DO NOTHING!!!!!!!", in a country with the population, geographical distribution, and general curmudgeonliness of the United States of America, well, that's when my bullshit detector jumps straight up and starts yammering.

And that was what set me off on a bit of a side-quest to go figure out whether these models match reality.

OK, so the keyword in the statement above about Prof. Ferguson's model is "unimpeded", which is to say, no protective measures of any kind taken to drastically slow the spread of the disease, in terms of lockdowns and shelter-in-place orders and quarantines and so on.

Both the USA and the UK have enacted such policies - but with dramatically different levels of strictness.

And yet the overall death tolls remain quite low.

In the US, as of this exact moment of writing, there have been 5,133 deaths from 216,362 confirmed cases. In the UK, there have been 2,352 deaths from 29,474 confirmed cases. The actual infection rates are vastly higher, since we know that roughly 80-85% of those infected show such mild symptoms that they don't even feel the need to get tested.

So in reality, you have to multiply the total infection rate by about 5 to get a rough estimate of the real number of infected people. And then, all of a sudden, those death rates - already not particularly scary once put into context - look quite reasonable and tolerable.

There is more evidence emerging every single day that we have wildly overreacted to the lethality of this virus. Take this one from Dave Cullen, who did a voice-over of a German epidemiologist questioning the numbers around the COVID-19 outbreak:



The key part of that video is where the good doctor points out that a French study found that the mortality rate from this latest SARS-related coronavirus, is in fact not statistically significantly different from that of other coronaviruses - including ones that cause the common cold.

Time and time again, the absolute worst and most terrifying scenarios that the world's foremost virologists, epidemiologists, immunologists, and biological warfare specialists, have all laid out for us, have proven to be totally WRONG.

And that's becoming quite noticeable. Some of the less stupid and more responsible media outlets in the world - such as those funded by the Russians, interestingly enough - are cottoning on to this fact and have tried to point this out to the rest of us:

Many people are waking up to the fact that the Covid-19 “pandemic” is not turning out as billed. When we finally emerge from it, the big question will be how many people have died from the virus. Here’s the most likely outcome.

You can bet that the institutions of international government, and the “experts” advising them, will try to massage and cherry-pick statistics to present the version of events that most closely matches their worst-case scenarios. The fact is, according to their early predictions, we are already long overdue millions of Covid-19 deaths that have failed to materialise.

But even when Covid-19 deaths are recorded, we have seen how it could be that people are dying with coronavirus rather than dying of it. This concept is easy enough to understand, and it encourages one to take a closer look at the breakdown of deaths across an entire society. The more you follow this rabbit hole down, the more interesting the numbers become.
[...]
To check this out, I looked at the British government’s own statistics on total deaths registered weekly across the UK. It shows that in the week ending on the 8th of March 2019, 10,898 people died in total in the UK. This year, in the week ending the 6th of March 2020, the equivalent figure was almost identical: 10,895. Make of that what you will. Statistics are currently available up to March 20, and while there is a lag between the spread of the virus and the resulting deaths, so far only about 1 percent of all mortalities bear any relation to coronavirus, and there is no visible spike. If nothing else, it helps to view the extent of the crisis in proportion - thousands of people die each week, and from the long-term view what we are seeing is not a plague, but a blip.


That is indeed what we are seeing right now.

In fact, let me show you by example exactly how easy it is to make HUGE mistakes from limited data sets without putting things into context.

The headline numbers available from Bing's COVID portal suggest a mortality rate of about 5%, which is the mortality assumption used by most of the early predictive models.

But if you take what we know already and put it all into a table, it is very easy to see why those models are so badly wrong.

Here's a table that I put together myself using publicly available data and a 5% mortality rate assumption:

RegionInfectionsProjectedActual% ErrorLikely True Rate
Global937,56746,87847,2560.81%1.01%
USA216,35210,8185,133-52.55%0.47%
Italy110,5745,52913,155137.94%2.38%
Spain104,1185,2069,38780.31%1.80%
China81,8594,0933,318-18.93%0.81%
Germany77,9813,899931-76.12%0.24%
France56,9892,8494,03241.50%1.42%
Iran47,5932,3803,03627.58%1.28%
United Kingdom29,4741,4742,35259.60%1.60%

I'll explain this table in simple terms:
  • The Projected column simply takes the total number of infections in any given region and computes 5% of that number;
  • The Actual column is the total number of recorded deaths ascribed to COVID-19;
  • The % Error column divides the Actual by the Projected and figures out what percentage of the Actual deaths were, well, Projected - and as you can see, the projections are WILDLY off in a number of cases;
  • That error is captured in the % Error column, of course;
  • The last column, Likely True Rate, accounts for the fact that the total count of confirmed infected people is only about 20% of the actual number, and so adjusts the Infections by multiplying them by 5 and divides that result into the total number of deaths;
So if you assume a 5% overall mortality rate, using very simplistic and naive base values, you can get close to a globally accurate number. But it completely ignores local realities that significant affect actual casualties on the ground.

That last column is the important one. And it shows an average mortality rate of about 1.22%, once you adjust for the fact that most people who have this virus, don't even know it.

And even that number is STILL wrong, because it is massively distorted by China, Iran, and Italy.

We simply cannot trust China's numbers. They are complete bullshit. We have good reason to believe that true deaths in Wuhan, alone, are more like 36,000-45,000. The Chinese numbers are anywhere from 15 to 40 times too low, both in terms of overall infections and actual deaths.

We also cannot trust Iran's numbers. Independent sources indicate that Iran has something approaching 15,000 deaths, and God only knows how many infected, and is now busy spreading the problem to Syria - which is also busy repressing their numbers.

As for Italy, we know that the infection is heavily concentrated in the Lombardia region with its large population of Chinese illegals and migrant workers, and its lousy air quality.

All of this is before we get to the fact that the death counts are badly distorted. Everyone keeps counting people who died WITH COVID-19 infections, as though they died FROM COVID-19 infections. It's not even remotely correct to do this, but that's the basis for public policy right now.

I just showed you how badly wrong people can be from mathematical modeling without resorting to a single exponential curve or fancy statistical distribution. Everything I did involved readily available public data and the kind of maths that I learned back in the 6th grade. And what I did just showed you how easy it is to make huge mistakes.

All of this indicates that the true death rate from this virus is substantially under 1%. And that is increasingly the view taken by a number of experienced epidemiologists:



There is yet more reason to doubt what the government "experts" are telling us. The very man whose model initially predicted 500,000 deaths in the UK alone from COVID-19, also advised the British government during the outbreaks of both BSE (mad cow disease) and foot-and-mouth (FMD) disease. And he wasn't just wrong in both cases - in the case of BSE, his worst-case scenario was 28,409% wrong, and in the case of FMD, his advice was directly responsible for costing the British economy nearly $15 billion with, as far as I can tell, zero deaths.

All of this naturally gives rise to three major questions:
  1. Why in God's name are we trusting "experts" whose worst case scenarios have been proven repeatedly wrong and wildly ridiculous, and whose models remain untested, unverified, and unexamined?
  2. Why are we not bothering with proper cost-benefit analyses to figure out when the cost of shutting down the economy exceeds the benefit to patients of doing so, especially given that the disease disproportionately affects the oldest and least productive people in the country?
  3. If the mortality rate is actually between 0.1% and 1%, which is about the same as a really bad flu... why is this virus such a terrible threat?
The first question is relatively easily answered:

Humans are herd creatures, and a herd is easily spooked. When you want a herd of cattle or sheep to do what you want, you pose a threat to it in a particular direction that seems so terrifying that it has no other choice but to go where you wish to direct it. And there is no doubt or question that several governments have used this outbreak as an excuse to grant themselves powers that give them unparalleled degrees of control over their populations.

The second question is also relatively easily answered:

The people who are most threatened by this disease are, indeed, the least economically productive - those over the age of about 60 or so, and especially those with existing medical conditions. But those people also happen to be the most valuable in terms of political power and influence; in most democracies, they vote more than younger cohorts do.

There are some important caveats to that analysis. Data coming out of Russia tell us that almost half of the confirmed cases coming out of that country are under the age of 45, and there is a small but growing body of evidence coming out of Europe and the UK that the virus is, in fact, dangerous to younger folks with no existing health issues, though with far less lethality than to older and weaker people.

But overall, it has been clear for at least two weeks now that the models used to justify these economy-destroying lockdown measures are hugely flawed.

So what about the third question - why is this, in fact, a public health emergency?

That, again, comes down to some basic mathematics.

I don't have hard data to work with here beyond a few limited points. So much of what follows is merely logical deduction.

First, most influenza epidemics have a mortality rate of a small fraction of 1%, and of the total number of infected, very, very few people with influenza actually require hospitalisation. The data that we have on hand indicates that about 80% of COVID-19 infected show no to very mild symptoms; 10% show symptoms severe enough to require a doctor's attention; and the remainder require hospitalisation. Of that remainder, 1-2% from the original population of infected will die.

Let's take a city like NYFC - about 8 million people. We have 47,000 confirmed infected cases there. That means that about 313,000 people are actually infected, roughly speaking. And of those 47,000, some 2,820 will require hospitalisation.

Assuming a 2% mortality rate among confirmed infections, which seems to be about the rate in bigger cities, close to a thousand of those 47,000 people will die. And in fact, over 1,300 people have died in NYC so far.

Now look at the logistics of a large hospital.

Most large hospitals don't have more than 20 or so ICU units and ventilators to go with them. That is because ICUs are really bloody expensive to maintain and run.

It doesn't take much computing to figure out that, even in a large city, 3,000 or so really sick patients are enough to overwhelm the medical facilities available at any given time.

That is the real nature of the threat: thousands of people getting sick simultaneously, who still represent a tiny percentage of the overall population of infected, overwhelming the resources of a system in a very big hurry.

And that is why you see so many scenes of horror and desperation coming out of the hospitals.

I've seen the videos myself of COVID-19 patients gasping for air and wailing in agony and terror. They are horrible to watch. It doesn't take all that many such cases to cause the most important, expensive, and complex critical-care systems in a region to completely collapse. COVID-19 is considerably more dangerous than the flu for this one reason - its ability to swamp existing support systems quickly once a certain number of patients get really sick.

That is precisely what we are seeing all over the world. Once the total number of critically ill people crosses a specific threshold, it's Katy-Bar-the-Door.

Therefore, the aim of public policy has to be to keep this scenario from happening. Unfortunately, it is becoming extremely clear from virtually every example that we have seen so far that authorities around the world are using chainsaws when scalpels are needed.

What would a more surgical, more precise, more effective, and less traumatic response look like? That's a topic for another post, as this one has gone on quite long enough. But I'll give you a hint: it looks quite a lot like what Taiwan and Singapore have done.

As for this crisis: my advice to all of you remains the same as ever.

Stay calm. Do not panic. Obey lawful orders to self-isolate and stay away from crowds. Look after the older folks with you, and take extra precautions to keep clean and healthy around them. Do not do anything stupid or take silly chances - like that idiot in my parents' housing compound did, when he arrived back from the UK, dodged screening tests, and violated sensible quarantine measures.

If you have the infection itself - DO NOT GIVE UP HOPE. Odds are very good that you will pull through just fine. Effective treatments using hydroxychloroquine with azithromycin and large doses of Vitamin C are already being used in NYFC and other "hot zones". You're most likely to be OK, so have a bit of faith and optimism.

Above all, have faith that the Lord is with us and looking out for us. Americans are blessed indeed to have the God-Emperor in charge right now, and I do believe that he is going to show us all exactly why he is a great leader, repeatedly, before this is done.

By June, this should all be winding down and we should all be looking to resume something resembling normal life - with some caveats. Have some faith and patience, and we WILL get there.

Comments

  1. Here are the numbers for Israel, 8 million people:
    ~7000 patients confirmed
    37 dead
    357 recovered

    Mortality rate of ~0.5%.
    The following is an age distribution of patients (March 29):
    https://www.clalit.co.il/he/your_health/family/PublishingImages/Pages/corona_updates/age3103.png
    Most of the cases are under 30. This makes sense as Israel is a young nation.

    ReplyDelete

Post a comment

NO ANONYMOUS COMMENTS. ALL ANONYMOUS COMMENTS WILL BE DELETED.
Contact the Didact: mantlesapproach@gmail.com

Popular Posts