Re: [113attendees] IETF 113 Update: 2022-02-10

["Scheffenegger, Richard" <[e-mail]>]

Am 18.02.2022 um 18:51 schrieb Alexandre Petrescu:
> That potential R interpolation function, applied to that Austrian data,
> could be valuable help in predicting more precisely than the current
> predictions that can be visible at
> https://www.sozialministerium.at/Informationen-zum-Coronavirus/COVID-Prognose-Konsortium-2022.html#februar-2022
>

That is very unlikely, that you get to better predictions using
simplistic models. The ones published above are done like weather
simulations, with agent (single individual) behavior modelling, feedback
loops and ensemble forecasting. The sum total of all these various runs
are accumuated and contribut to the confidence intervals depicted.

Human behavior is probably at least as complex as weather patterns, thus
the limited time range (only 7 days, with a population of 9 mio in austria).

> Probably this would predict that the next peak would be in 7 months time
> (I dont know?), which means peak in August 2022.  August is much far
> away than March.

Do not forget that these peaks were all driven by different variants of
the virus, and when/how fast a new, more evolutionary fit subtype
emerges, is a stochasic process - which can not be modelled on an
individual basis.

It is not impossible for Covid to become more contageous but even less
problematic (like the two other endemic coronaviridae types, which cause
seasonal flu symptoms - mostly), which is the more probable outcome.

Or more deadly (CoVID is said to share a lot of lineage with MERS/SARS
coronavirus, which had a 10% and 30% mortality. The worst case would be
a type that is as contageous as Omicron, but as fatal as MERS at 30%
after 14 days, such that infected individuals have had plenty of time to
spread it around before becoming sick)