A second thread to explain how the #COVID19 modelling works. https://twitter.com/peripatetical/status/1247402435858944002?s=20

(Stephen Hawking was once told that each equation in his book would lose him half his audience. If the whole Australian population were potential readers, there'd be 6 people still reading to the end of the technical appendix)

The model is based on a traditional SEIR model. This means people flow between different states - Susceptible, Exposed, Infected/infectious and Recovered.
(the doubling up of the E and I compartments is a mathematical trick)

The flows between compartments are governed by parameters. The average number of secondary cases resulting from a case (R0). The durations of incubation and infection. The risk of hospitalisation or ICU admission by age group.

These are all based on the Australian population structure, which takes into account the size of age groups and people at increased risk (such as Aboriginal and Torres Strait Islander peoples).

There is a "base case" starting point, then various changes to parameters to show the impact of interventions

The four scenarios that are modelled here are "unmitigated" (no interventions), "quarantine/isolation" only, quarantine/isolation + some social distancing, and quarantine/isolation + more social distancing.

A second model looks at the care needs of those who have infection. Cases may have mild infection (only requiring outpatient care) or severe (presenting to hospital, but then requiring ward or ICU admission)

There are many more parameters that govern the flows between these states. The proportions that have mild/severe disease and where they present. The proportion that require ward and ICU beds and how long they need it for.

As you can see, there are a lot of simplifying assumptions that need to be made, and a lot of parameters that need to be estimated. In the end, modelling is just a tool, and one of many considerations that are used to inform decisions.