Scrutable

badger wrote: ↑

Thu Jun 11, 2020 2:27 pm

We now have scientists breaking ranks and admonishing the Govt for not locking down earlier, but they were also involved in the process of advising the govt - possibly badly. Are we looking at an impending blame-game bloodbath between scientists and govt? And even within SAGE and sub-committees?

When it comes to blame, the person who makes the descicion is to blame. If they have been badly advised, it is till their fault as they got to choose their advisors.

In the specific area of scientific advice, there is no excuse of "following the science". Science is something you do, not follow. When you get scientific advice you can evaluate it by checking references, getting others to give you their opinion, and seeing if it is internally consistent. That process happened on this very forum and I think fairly consistently led to people saying the government was wrong even without hindsight.

And how hard do you interrogate people who know so much more than you about their subject? What questions do you ask them, beyond “is this your current best estimate?”

plodder wrote: ↑

Sun Jun 14, 2020 7:25 am

And how hard do you interrogate people who know so much more than you about their subject? What questions do you ask them, beyond “is this your current best estimate?”

In early March a Prime Minister could have started with:

Is your opinion part of a consensus?
Of all of the committee? Of scientists working in other relevant institutions (eg the WHO, US CDC, European CDC, Chinese, South Korean and Italian national institutions)?

If the answer to any of the above is 'no' ask them to explain what are the differences of opinion, and why they think their view is more reliable.

Woodchopper wrote: ↑

Sun Jun 14, 2020 9:33 am

plodder wrote: ↑

Sun Jun 14, 2020 7:25 am

And how hard do you interrogate people who know so much more than you about their subject? What questions do you ask them, beyond “is this your current best estimate?”

In early March a Prime Minister could have started with:

Is your opinion part of a consensus?
Of all of the committee? Of scientists working in other relevant institutions (eg the WHO, US CDC, European CDC, Chinese, South Korean and Italian national institutions)?

If the answer to any of the above is 'no' ask them to explain what are the differences of opinion, and why they think their view is more reliable.

It certainly seems to be the case that the UK's scientific advice occurred pretty much in a vacuum, without reference to what other organisations and countries were doing/saying.

Whatever advice was being given to ministers and whatever evidence it was based on, the public impression wasn't so much that they were in a vacuum ignoring what everyone else was saying (i.e. "test, test, test") rather it was that the rest of the world was wrong to imagine they could contain the outbreak, that even if they seemed temporarily to be succeeding it would only rebound with a much worse second wave, and the best we could hope to do was flatten the curve while we get it over with.

Boris Johnson's government has the worst approval rating in the WORLD for its handling of the coronavirus pandemic - below even Donald Trump despite the US having the highest death toll

Comes to something when even the "Fail" has headlines like this

Gentleman Jim wrote: ↑

Mon Jun 15, 2020 7:52 am

Boris Johnson's government has the worst approval rating in the WORLD for its handling of the coronavirus pandemic - below even Donald Trump despite the US having the highest death toll

Comes to something when even the "Fail" has headlines like this

Get better polling results soon Boris, mate

plodder wrote: ↑

Sun Jun 14, 2020 7:25 am

And how hard do you interrogate people who know so much more than you about their subject? What questions do you ask them, beyond “is this your current best estimate?”

Am talking about SAGE interrogating SPI-M, or whoever represents SPI-M on the SAGE committee. They may not know as much about modelling as SPI-M, but I assume that some of them know what they're talking about.

Martin Y wrote: ↑

Sun Jun 14, 2020 12:53 pm

Whatever advice was being given to ministers and whatever evidence it was based on, the public impression wasn't so much that they were in a vacuum ignoring what everyone else was saying (i.e. "test, test, test") rather it was that the rest of the world was wrong to imagine they could contain the outbreak, that even if they seemed temporarily to be succeeding it would only rebound with a much worse second wave, and the best we could hope to do was flatten the curve while we get it over with.

yes, but that public impression was married to the idea of herd immunity (derived from infection rather than vaccine) and that got reversed pretty quickly. I don't think the Govt meant to kill so many in care homes, and they did admit their strategy shifted when the numbers (in the model they were listening to) changed.

And no scientists flanking Raab at briefing now, because they "have a lot of other work to do".

Trouble at t'mill, lad?

Martin Y wrote: ↑

Sun Jun 14, 2020 12:53 pm

Whatever advice was being given to ministers and whatever evidence it was based on, the public impression wasn't so much that they were in a vacuum ignoring what everyone else was saying (i.e. "test, test, test") rather it was that the rest of the world was wrong to imagine they could contain the outbreak, that even if they seemed temporarily to be succeeding it would only rebound with a much worse second wave, and the best we could hope to do was flatten the curve while we get it over with.

Have you seen the interview with Rory Stewart from mid-March (when he was running for mayor of London)? It's aged very well in many ways, given subsequent events, but one comment he made was that as PM, he would heavily grill the advisors to try to understand why exactly they were recommending such a different strategy to other countries.

Woodchopper wrote: ↑

Sun Jun 14, 2020 9:33 am

plodder wrote: ↑

Sun Jun 14, 2020 7:25 am

And how hard do you interrogate people who know so much more than you about their subject? What questions do you ask them, beyond “is this your current best estimate?”

In early March a Prime Minister could have started with:

Is your opinion part of a consensus?
Of all of the committee? Of scientists working in other relevant institutions (eg the WHO, US CDC, European CDC, Chinese, South Korean and Italian national institutions)?

If the answer to any of the above is 'no' ask them to explain what are the differences of opinion, and why they think their view is more reliable.

These are all varieties of “is this your current best estimate” though.

AMS wrote: ↑

Mon Jun 15, 2020 4:46 pm

Martin Y wrote: ↑

Sun Jun 14, 2020 12:53 pm

Whatever advice was being given to ministers and whatever evidence it was based on, the public impression wasn't so much that they were in a vacuum ignoring what everyone else was saying (i.e. "test, test, test") rather it was that the rest of the world was wrong to imagine they could contain the outbreak, that even if they seemed temporarily to be succeeding it would only rebound with a much worse second wave, and the best we could hope to do was flatten the curve while we get it over with.

Have you seen the interview with Rory Stewart from mid-March (when he was running for mayor of London)? It's aged very well in many ways, given subsequent events, but one comment he made was that as PM, he would heavily grill the advisors to try to understand why exactly they were recommending such a different strategy to other countries.

I did. It's a remarkable interview.

plodder wrote: ↑

Sun Jun 14, 2020 7:25 am

And how hard do you interrogate people who know so much more than you about their subject? What questions do you ask them, beyond “is this your current best estimate?”

By having your incredibly clever chief advisor attend meetings so he can weigh up the opinions?

plodder wrote: ↑

Mon Jun 15, 2020 5:27 pm

Woodchopper wrote: ↑

Sun Jun 14, 2020 9:33 am

plodder wrote: ↑

Sun Jun 14, 2020 7:25 am

And how hard do you interrogate people who know so much more than you about their subject? What questions do you ask them, beyond “is this your current best estimate?”

In early March a Prime Minister could have started with:

Is your opinion part of a consensus?
Of all of the committee? Of scientists working in other relevant institutions (eg the WHO, US CDC, European CDC, Chinese, South Korean and Italian national institutions)?

If the answer to any of the above is 'no' ask them to explain what are the differences of opinion, and why they think their view is more reliable.

These are all varieties of “is this your current best estimate” though.

No they’re not. The answer to that question will always be ‘yes’. (They’re hardly likely to say “I’ve got a better estimate in the office but I kept it to myself, Prime Minister”.

PeteB wrote: ↑

Thu Jun 11, 2020 6:02 pm

I don't absolve politicians
Rory Stewart 12th March

https://youtu.be/pdeHUGalpaE

Jeremy hunt was saying similar things

I recommend listening to this

AMS wrote: ↑

Mon Jun 15, 2020 4:46 pm

Martin Y wrote: ↑

Sun Jun 14, 2020 12:53 pm

Whatever advice was being given to ministers and whatever evidence it was based on, the public impression wasn't so much that they were in a vacuum ignoring what everyone else was saying (i.e. "test, test, test") rather it was that the rest of the world was wrong to imagine they could contain the outbreak, that even if they seemed temporarily to be succeeding it would only rebound with a much worse second wave, and the best we could hope to do was flatten the curve while we get it over with.

Have you seen the interview with Rory Stewart from mid-March (when he was running for mayor of London)? It's aged very well in many ways, given subsequent events, but one comment he made was that as PM, he would heavily grill the advisors to try to understand why exactly they were recommending such a different strategy to other countries.

The bit in bold should have been a bit of a warning. Here, we did a pretty reasonable job of predicting what would happen (at least better than Johnson was doing) based on very naive and simple sanity checks. "Oh look, a lockdown worked in Wuhan - Italy has gone for a lockdown, most countries except Japan seem to be having a 35% per day growth rate until lockdown - that's scary."

What we weren't aware of, was Exercise Cygnus, and the fact that it had highlighted a lack of PPE as an issue. But then Gove didn't read that report until the 4th week in April.

And nor did the self-proclaimed superforecaster Cummings.

Johnson didn't attend the first 5 COBR meetings on the pandemic. I'd have thought that by the time you got to the second, one might have thought the subject was quite important.

Failing to educate oneself is not an excuse. Hancock himself proclaimed the wonderful "diversity of thought" in the Cabinet, which must be in the running for one of the stupidest things said by a British Cabinet minister in June 2020.

Woodchopper wrote: ↑

Mon Jun 15, 2020 6:48 pm

plodder wrote: ↑

Mon Jun 15, 2020 5:27 pm

These are all varieties of “is this your current best estimate” though.

No they’re not. The answer to that question will always be ‘yes’. (They’re hardly likely to say “I’ve got a better estimate in the office but I kept it to myself, Prime Minister”.

I'm not sure I agree. Dealing with experts, especially when they don't all agree (which is the case here) is difficult at the best of times. When you add time pressures etc it's very difficult for a manager to filter out who is talking shite.

I think the idea of SAGE is that the managers are also experts in evaluating expert evidence (if I'm following the discussion, we're talking about internal discussions, not presentation to Cummings to decide UK policy)

Bird on a Fire wrote: ↑

Wed Jun 17, 2020 3:23 pm

I think the idea of SAGE is that the managers are also experts in evaluating expert evidence (if I'm following the discussion, we're talking about internal discussions, not presentation to Cummings to decide UK policy)

It really doesn't need much expertise to plot the data on log-plots and see that most countries were increasing at about 35% per day. I mean, if they were too lazy to do it themselves, the FT had been producing the graphs anyway.

Bird on a Fire wrote: ↑

Wed Jun 17, 2020 3:23 pm

I think the idea of SAGE is that the managers are also experts in evaluating expert evidence (if I'm following the discussion, we're talking about internal discussions, not presentation to Cummings to decide UK policy)

I thought Plods and Chops were talking about the PM interrogating the experts, while Badgy and Birdy were talking about SAGE interrogating SPI-M.

(Well, Woodchopper kept referencing the PM and Plodder kept not correcting him anyway.)

Report 41 - The 2020 SARS-CoV-2 epidemic in England: key epidemiological drivers and impact of interventions - the effect of lockdowns: report on fatalities in the first wave - only national lockdown brought the reproduction number below 1 consistently: introduced one week earlier it could have reduced first wave deaths from 36,700 to 15,700 (95%CrI: 8,900–26,800).

an thread on teh twitter

shpalman wrote: ↑

Wed Dec 23, 2020 12:12 pm

Report 41 - The 2020 SARS-CoV-2 epidemic in England: key epidemiological drivers and impact of interventions - the effect of lockdowns: report on fatalities in the first wave - only national lockdown brought the reproduction number below 1 consistently: introduced one week earlier it could have reduced first wave deaths from 36,700 to 15,700 (95%CrI: 8,900–26,800).

an thread on teh twitter

Thanks.

And when the exponential decay time constant is far longer than the risetime time constant, not only is a week's delay responsible for most of the cases, it needs far longer to get cases down as well.

I've been talking to lots of Dad's college friends who learned this viscerally in 1967. You don't have time to wait for certainty.

And that is consistent with what we knew in March - https://www.theguardian.com/world/2020/ ... f-covid-19

shpalman wrote: ↑

Fri May 22, 2020 11:17 am

Well akshully you can't lock down too early because then it will go on too long and people will get bored of it (this seems to be based on Whitty's personal experience of patients in medical practice who do not see drug prescriptions through to their completion, and not supported by the behavioural scientists appointed by the government itself to Sage’s subcommittee, SPI–B)

Sunak wrong to say scientists given too much power over lockdown policy, Sage experts say

Well, one "scientist" seems to have been given too much power right there at the beginning.

China has proven that lockdown early and often is the key.