Think this story might speak to why policymakers/modellers can be nervous about how modelling is handled in the public domain. The story uses the word "predicts". Some quick observations on that... 1/ https://twitter.com/normanswan/status/1246608215518789632

The ABC hasn't included the modelling it obtained on the story but my understanding is modellers say their work doesn't "predict" an outcome. Rather it projects a range of possible scenarios -- from worst to best case -- based on fallible data and assumptions. 2/

"Predicts" might still seem legitimate to us laypeople but one modeller I spoke to recently emphasised the difference. "We are not in the business of providing predictions. There’s too much uncertainty." 3/

Modelling is intended as one input to consider in the policymaking process. "It will keep changing because we are scientists, we watch what’s happening, take on board new data and update accordingly." 4/

Likewise, when the NZ government released its modelling, it cautioned: "It’s critical to understand that each of the models presents a number of potential future scenarios; there are no 'predictions'." 5/

. @peripatetical the other day: "Mathematical models simply tell you what could happen if you assume particular inputs. But there are simply too many uncertainties with COVID19 to have a single likely outcome." 6/

Not criticising the use of "predicts" in this story. The distinction between these words might seem trivial to many of us but it's an interesting issue. A partial taste of modelling like this might even speak to the merits of government opening the kimono. 7/7