Read on Twitter
How @unimelb design experts and epidemiologists built a new type of #COVID19 modelling that was instrumental in eliminating COVID-19 in Victoria; a thread.
While previous mathematical models were able to tell us how quickly the virus was spreading, they couldn’t tell us how different scenarios (ie different policy settings) would impact the virus spread.
In other words: prior to COVID-19 no modelling existed that could be agile and responsive to the specific & various scenarios needing to be tested by Government.
The COVID-19 model created by an interdisciplinary team was able to do just that.
Take a bow @Agent_Jase, @TonyBlakely_PI, @nat_carvalho1, @laxmanb, @mark_transport and others on the team. 
Their model accurately 
predicted that if Victoria didn’t go immediately to Stage Four in early July, there was a 48% chance of 400+ cases per day (which would then require a longer lockdown to contain.) 
https://www.mja.com.au/journal/2020/probability-6-week-lockdown-victoria-commencing-9-july-2020-achieving-elimination
predicted that if Victoria didn’t go immediately to Stage Four in early July, there was a 48% chance of 400+ cases per day (which would then require a longer lockdown to contain.) https://www.mja.com.au/journal/2020/probability-6-week-lockdown-victoria-commencing-9-july-2020-achieving-elimination
The model also predicted:
The day Victoria would get to a running average of 5 cases per day.
The week 
NZ would achieve elimination.
Cool! And people go wow. And we like them to go wow. But that wasn’t the hard bit.
The day Victoria would get to a running average of 5 cases per day. The week NZ would achieve elimination.Cool! And people go wow. And we like them to go wow. But that wasn’t the hard bit.
It was the model’s flexibility and its ability to test different policy options, that made it so successful to achieving elimination.
@DanielAndrewsMP, @VictorianCHO and @VicGovDHHS approached the @unimelb team in August to assist them to model and think through the strategy for reopening.
The model was rebuilt to reflect how stages or steps (including lockdowns) would look, using data and combinations of things:
Evidence on case numbers.
Contract tracing performance.
Policy settings and community adherence.
Mask wearing.
School attendance.
Evidence on case numbers. Contract tracing performance.Policy settings and community adherence. Mask wearing. School attendance.
It was all geared towards easing restrictions in a safe way so that we would avoid a third wave. 
The model then gave an estimate of how #COVID19 would spread through this simulated scenario.
In August, the #COVID19VIC rolling average was just below 200 cases per day. The model suggested a 5-case-per day (14-day rolling average) threshold for loosening restrictions, which gave us a 3% chance of having a COVID-normal 
Christmas.
Christmas.
That threshold was savaged in the local and national media as a ‘guess,’ ‘proven wrong already’ and ‘impossible.’
It was a challenging time for the team, who continued to stand by the science.
On November 26 Victoria officially reached the elusive elimination threshold and logged its 28th day of zero COVID-19 cases.
The new modelling was responsive, agile & accurate.
The roadmap worked.
Today we stand with all Victorians to celebrate ELIMINATION OF COMMUNITY TRANSMISSION.
The roadmap worked.
Today we stand with all Victorians to celebrate ELIMINATION OF COMMUNITY TRANSMISSION.
