[marty (OP)]

I listened to the ‘COVID-19: Will lock-down work?’ Radio NZ podcast with Chris Smith.  The interviewer, Kim Hill refers to the Imperial College disease modelling study by Neil Ferguson which has been highly influential in changing government approach to the disease.

However, in a series of tweets today, Neil reveals the background - “I wrote the code (thousands of lines of undocumented C) 13+ years ago to model flu pandemics…”

One of the replies on Twitter says:

“So many concerns.

Modelled on a flu pandemic.
13 years old.
C.
Thousands of lines of code.”

Are they right to be concerned?

[set fair]

What has changed in 13 years? +ve It has probably been test for example with ebola, -ve less hospital beds. I doubt it took account of people widely ignoring social distancing. It should have been published. My concern would be how degenerate the model is.

[alancalverd]

Yes.

The general form of a pandemic is well known from previous experience of human and animal pandemics. The particular progress of a new one depends on a whole lot of unknowns such as public response to instruction (stay at home or buy toilet paper?), infectivity, mobility of asymptomatic carriers, symptomless incubation period, mobility of symptomatic carriers, lethality of disease, distribution of treatment facilities, availability of preventive interventions,.....none of which is known with sufficient accuracy to make optimal decisions until its effect is already obvious.  So the only sensible decision is potential overreaction.

Best summary I've heard was from an American epidemiologist who said "If, when it's all over, people say 'what was all the fuss about?', we will have got it right". 

[Bored chemist]

Best summary I've heard was from an American epidemiologist who said "If, when it's all over, people say 'what was all the fuss about?', we will have got it right". 

Just like Y2K

[marty (OP)]

Edit: Should we be concerned that 'SAGE' means: wise, especially as a result of great experience?

[Bored chemist]

I don't know how old the idea of counting is, but I'm more concerned that my government can't count tests than about a 13 year old mathematical model.

[marty (OP)]

Agreed and app plans.

On Twitter: /FryRsquared/status/1257397805003538437