Now everyone has heard of the Pueyo Model, which modeled the development of the coronavirus pandemic in the world. In fact, these are Excel sheets in which Pueyo downloaded data on coronavirus statistics and screwed up a graph, idealizing the ranges.

The "model" turned out to be a calculation of the arithmetic mean (!) growth rate based on the number of cases detected in 17 days in 10 countries, and then direct extrapolation into the infinite future.

But it simply calculates the growth rate percentages incorrectly! Because, in the existing version of the formula, the calculations in the cells are set up based on the fact that newly detected cases are entered into the cases table. But the table is filled with data according to a different logic (!) - the total number of detected cases of infection is an increasing total! As a result, it has a growth rate of at least 100% every day. Naturally, with such a table, it is impossible to do otherwise - because tomorrow the number of detected cases cannot be less than today! As a result, for example, the transition from 30 cases to 31 cases, he has a growth rate of 103%?

So this "expert" calculates the amount of interest in different countries. Then he stupidly calculates the arithmetic mean and gets a graph with this growth rate. He doesn't take into account whether the data is from the closed ship Diamond Princes, a small country or a huge one, or the methodology for identifying infected people. The main thing is to put numbers in cells with absolutely no understanding of how the data was obtained and how the real processes developed.

Ask why, having a growth rate of 120%, he multiplies by 1.2 when trying to predict the next value from 2480? Any mathematician would answer the question of what the value of N would be if it grew (the growth rate!) by 120%, and it would be N + (N/100) * 120, in which case 2480 + (2480/100) * 120 = 5456

But the growth rate is the coefficient or value at which the next value increases from the previous one, and for financiers and effective managers, the growth rate means the state of the next step relative to the previous one, expressed as a percentage. And what all normal people call the growth rate, financiers and effective managers call the growth rate. And the growth rate is equal to the growth rate minus 100. Accordingly, in the case of our epidemic, with an average growth rate of 121%, the average growth rate is 21%.

The growth rate is not the growth rate at all.

It is clear why financiers do this. We all remember wealth growth charts with a growth rate of 100,500%. Invest money with us and it will grow exponentially, like the coronavirus! Lenya Golubkov from MMM also drew such charts. And for business owners to write reports, it is better to write a growth rate of 107% instead of 7%. Even if the business is losing money, the growth rate will be great. You lose more than half of your fortune, but your growth rate is 40%!!! Managers will keep quiet about a growth rate of -60%. Well, it's just an education for investors and managers: if you see a growth rate of less than 100%, then everything is going downhill. And no one will tell you much about the growth rate. This is for internal consumption. It was done naturally to pick the brains of managers and clients. But has it become used to assess global threats?

But even in this, the guy was pissing off. The way he uses his term "growth rate" is properly called "chain growth rate". The category, in my opinion, was introduced solely to brainwash people, since the linguistic meaning is completely different from the content of the formula for calculating it.

So, back to the coronavirus. You've probably realized by now that an epidemic growth rate of 100% means that the epidemic has stopped and is not moving anywhere. And a growth rate of 80% means that it is declining by 20%. Don't try to understand it, just accept it. And remember the difference between the growth rate and the growth rate. There are so many effective managers now that, apparently, the understanding of the term cannot be changed. Just when you are told that the epidemic is developing at a daily growth rate of 80%, know that it is decreasing by 20% every day.

What was the point of all this? You can already tell from the approach who was doing it. A financier who is used to counting money and selling something to others. Not to research processes and build models of process functioning.

He calculated the growth rate at the moment when everyone was showing this growth. Each day naturally had more expressions than the previous ones. It simply had a growth rate of more than 100 a priori simply because of the data structure! By chance, it turned out to be such that it obviously did not detect terrible errors. In fact, his calculations show that the number of infections increases daily by 21% from the previous day. This gives a visually logical graph, without checking the calculations it looks like something that has the right to live... But it has nothing to do with the real characteristics of the process!

To make sure that this method of calculation does not make sense, here is an example.

Let's say you decide to count the number of cats around you and decide to devote 17 days to this task. On the first day you found 1 cat, and on the second day you found 30 cats. The growth rate of the number of cats around you was 3000% on that day, and then you ran out of cats around you and didn't find any for the remaining 15 days. A normal person who hasn't found a single new cat in 2 weeks will conclude that there are really only 31 cats. But not an effective epidemic predictor! Because do you know what your average growth rate of cat detection is according to the logic of Pueyo, Leni Golubkov, and most effective managers?

Let's calculate (0% + 3000% + 100% + 100 + ... and plus 13 times 100%) / 17 total 4500/17 = 264%!!! Do you understand? With just one cat found on the second day, you got an arithmetic mean growth rate of 264% on the 17th day, although there are no more cats at all! ... And if you extrapolate 31 found cats into the future with a daily growth rate of 264%, it suddenly turns out that the "model" predicted 31 * 2.64 = 81.8 cats around you on day 18, although there were only 31 cats, and there are still only 31 cats. And in 10 days, according to the "model", there will be 509,801 cats around you!!!

Passing off the number of detected cases as the number of actually infected, especially knowing that the disease in ~ 70% cases is so mild, does not fall into any statistics... well, only a fool can do that. And ignoring this fact leads to a terrible overestimation of the lethality of the infection!

As a result of these "predictions," we get an a priori predetermined constant growth of the epidemic with a surprisingly high mortality rate.

And this was sold to the whole world as a prophecy of an exponential increase in mortality, as well as the inevitable end of the world, unless everyone hides in holes until they completely sink into the Stone Age.

But the worst part is that none of the scientific institutes, expensive consultants with world renown names even tried to tell the world that this "model" was a ridiculous mistake. It was enough to open this file on Google.

PS Let's not talk about the medical side of things and virology here. In this case, we're only talking about the "model" that has everyone shitting their pants. The "model" is manipulative. We will talk about the rest later.