[chbot] Modelling virus spread...

[Bevin Brett]

I built a computer simulation of the spread of a virus through a group of people who had three tiers of interactions - basically those they met daily, weekly, and yearly.
I tried varying the likelihood of an interaction transmitting an infection.
Provided that the numberOfDailyContacts * likelihoodOfTransmission was greater than 1, there were two differences
1) how long it took for 97% of the population to get infected
2) whether it was 97% or 99% of the population before the virus died out
So, while I don't claim my model matches a particular virus, I do not agree with Carl Bergstrom's argument that flattening the curve has a demonstrable significant benefit other than not swamping the medical system - in particular in my model it took 150 days to infect 99% or (with a tighter lockdown) 750 days to infect 97%.
Two year lock-down to save 3% (the mortality rate) of 2% (the number that flattening the curve stopped from getting infected) of the population from dying a few years earlier is not worth it. There are lots of much cheaper ways to achieve the same effect, that we have already said we are not willing to do.
Anyone else tried modelling this ? What results did you get?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ourshack.com/pipermail/chchrobotics/attachments/20200509/ad7b2ff0/attachment.html>