Extract from an email that I’ve received about Grattan’s bad report

This extract is self-explanatory.

==EXTRACT==

Seeing my fellow Victorians in collective delusion has been very upsetting. I am ‘on the left’ and often am scared to speak.

The misuse of agent based modelling, throwing around terms like ‘the science says’ and backing it up with ‘this super computer has a fair bit of grunt’ has infuriated me.

One thing that really troubled me was that the Grattan Institute ‘Towards COVID zero report’ (https://grattan.edu.au/wp-content/uploads/2020/09/Go-for-zero-how-Australia-can-get-to-zero-COVID-19-cases-Grattan-Report.pdf page 20) ignored local mental health hotline effects and instead turned to google search trends indicating affect from other countries to conclude that “They find that the pandemic and spread of COVID-19 has a clear negative impact on mental health in the UK, but the introduction of restrictions and controls was associated with a rebound in mental wellbeing.”

The reliance on top down data to justify personal bias is out of control. The agent based model is a joke. It’s great for understanding how systems impact on each other, or perhaps for prediction where you have a few flu seasons worth of data in context. Or perhaps if you are modelling something like bee migration. But to set thresholds for the country?