Read on Twitter
Watching Lockdown 1.0 now on BBC2, looks to be shedding some new light on scientific advice and how it informed policy.
Highlighting how NERVTAG identified the risk as 'low' in January 2020. It would be interesting to know more about how this judgement was reached.
Medley: "The whole point of SPI-M is to generate a consensus view". A familiar idea in science...but how is consensus reached? The UK government provides on guidance on this, and minutes tend to be opaque
Jombart (paraphrased): "The data wasn't good and that creates a lot of uncertainty in the predictions". Yes, but was that uncertainty discussed in public. Evidence suggests that uncertainties were downplayed https://www.nature.com/articles/s41599-020-00612-w
"Modellers didn't know how care homes were connected to communities - this was a failing" An open question: to what extent should epidemiological models strive for 'realism'?
Early March: modellers start to look at "non-pharmaceutical interventions". Johnson: "the country is extremely well-prepared" .........
Johnson: "I shook hands with everyone in a hospital the other day". The parallels between this and Gummer's BSE burger incident are striking. A high-profile, ill-fated attempt at public reassurance. See Jasanoff on 'civic dislocation' http://mittalsouthasiainstitute.harvard.edu/wp-content/uploads/2020/04/Civilization-and-Madness-Sheila-Jasanoff.pdf
"There is no consensus on immediate action on lockdowns" Interesting to here the dissenting view from Riley, but how does SPI-M and SAGE deal with dissensus. Is it in the minutes?
A clip from the Tomas Pueyo interview, a layman challenging the scientific received wisdom. More on that encounter here http://ihuman.group.shef.ac.uk/what-does-covid-19-mean-for-expertise/
Susan Michie not pulling punches on the unscientific nature of 'behavioural fatigue' i.e. that citizens would get bored with stringent measures. Where that idea came from, and what it says about some scientists' understanding of the public, is another 'intriguing' question...
I am mainly tweeting about the science-policy angles, but the families' stories of personal loss are heartbreaking and are really helping to ground the wonkery
Michie thought we needed to keep trust in the leadership. She weighed this when defending government policy on sporting events on Newsnight. Days later, the country was in revolt and voting with its feet. The Premier League closed down of its own accord.
Fascinating clash between modellers and NHS about what data quality, and how 'up-to-the-minute' it is
Programme goes along with the narrative that Ferguson's model changed the policy. See https://www.nytimes.com/2020/03/17/world/europe/coronavirus-imperial-college-johnson.html
Now we get to a crunch point. The public presentation of the virus doubling rate was that it was five days, but then there was a realisation within the modeling community that the rate was much faster than that. This led to full lockdown. But hang on...
What this programme does not say, is that there is evidence that this was not just a miscalculation caused by 'poor data'. Rapid doubling rates were within the credible range of modelling from LSHTM. So why was it being presented as a single number of '5 days'. Me again 
I don't think it was up to the modellers to check care home policies. But again, this programme is excellent on the personal toll of this crisis.
No government spokesperson on camera, but their written statement makes reference to learning from SARS and MERS. It's not clear to me that this really happened, particularly with policy. A Wuhan-style lockdown initially 'could not be imagined'
And that's it...good programme, shed some new light on the story with some voices we haven't heard from before. But didn't grapple with the fundamental issues about how to grapple with uncertainty, and that the faster doubling rate was already in the models. 
