From St. Petersburg to Pandemic: stupid comments to @dottorpax & @nntaleb paper ( https://www.nature.com/articles/s41567-020-0921-x)

Disclaimer: i sincerely love Pasquale, he is the person I would like to be; great mathematician, great cook, great education. An extremely respectable person.

And..
1\\

I presumptuously believe that he will appreciate my donkey thoughts more more than those of the many rimjobbers(aka asskissers).
St.Petersburg paradox: an infinite expected value implies a strictly limited(upper bounded) finite utility function.
this is the rule(Bernoulli)

2\\

If we think like a virus, the optimal goal to achieve is to live with the guest (homicide is equal to suicide)...and this is the main reason why today we are 7.7 billion people and at the time of the Spanish flu we were 1.8 billion people.
So, when you write(attachment)...
3\\

I believe you run into a logical fallacy. Do not expect survivors(no herd immunity and no consciousness of transmission ) also, makes this eventuality practically impracticable(sorry for the pun but is math).

the second(or third) mistake i found in a work...
4\\

i sincerely appreciated, is to consider epidemiological models "not fat tailed".
On the contrary...i think they are "too much fat tailed".
They are overly simplistic in their catastrophic assumptions.
This is R0 distribution left free "to reproduce himself".
5\\

(Pi is the percentage of population required (or already assumed) to achieve herd immunity)

It seems a Lévy alpha-stable distribution, no finite mean, non finite variance with this characteristic exponent.

There is nothing "normal", look at R0 rate of growth...
6\\

and this is the reason because some of us judge the forecasts made using this "stable Paretian distribution", as Mandelbrot said, rather unrealistic e dangerous often.
Think at technical commission report for reopening (Stefano Merler) @alexvespi

7\\

"100000 persons in Intensive Care Units in june..."
How? We should take&carry them by force i suppose.... :=)

Conclusion, my time is over..(i must go to work)..

theoretical models are useful if you DON'T BIND THEM to decisions that will CERTAINLY(and not theoretically)
8\\

affect people's lives. EVT is extremely useful, Epidemiological models too (they are similar) but real life is too short and some mistakes derived from a distorted precautionary principle is not said to be recovered.
:=)

Have a good day,

ps: @m_starnini, math is the same for everyone.

It can be right or wrong.

Keep this in mind Michele :=)