With Captain Hindsight we now know Vallance and Whitty we're right in September.

Could we have been Captain Foresight?

In this thread, I forensically take apart a "Great Barrington" proponent.

You need no foresight, just objectivity and integrity.

Journalism do you know that?

As a past AZT viral researcher, I know broad ideas drive innovation

But I expect us all to follow the Scientific Method.

What is IMHO dangerous is politicising science, while failing to confront theories lacking evidence.

That's a view
I will attempt to prove it objectively.

1
Here I analyse theories proposed by Great Barrington proponent; Sunetra Gupta.

In science, we judge theory by its ability to predict results.

Gupta's theories are questionable because they predict poorly.

Responsible Journalists look at the average balance of theory.

2
The source for Sunetra Gupta. is an @afneil interview: 2nd October []

The intent was to criticise the Whitty Valance Press Conference of 21st September [ https://www.gov.uk/government/speeches/chief-scientific-advisor-and-chief-medical-officer-briefing-on-coronavirus-covid-19-21-september-2020--2].

I'll look at how wrong Gupta was to show why theory must be averaged.

3
I chose this interview because it was subsequently used by many on the right to justify criticism of CMO/CSO.

The interview has no counter view. And Neil, famous for forensic questioning, uses Gupta to support his biases.

As a reference for anything it's questionable.

4
Two points to know before I get into it.

FIRST the famous 50,000 deaths was not a projection, it was an illustration of the dangers of exponential growth.

To quote Vallance,
“[the chart] is not a prediction, but it is a way of thinking about how quickly this can change.”

5

Even as a projection, the chart would not be sensationalist.

Full fact created this helpful illustration to show how viral growth changes if you increase time to double assumptions.

You gain a little time, but you don't negate the risk or disqualify the point.

6

SECOND. Death Rate

Cases don't matter. Ask instead how many will die? Whitty/Vallance DID project that.

"50,000 cases per day [in Oct] would be expected to lead a month later, so the middle of November say, to 200 plus deaths per day."

Sadly we got there already.

7

Objectively what do you say of Whitty/Valance?

a) 50k cases were declared as an illustration before any prevention
b) was likely at the high end of projections but was NOT unreasonable or scaremongering
c) enabled a death rate projection
d) which if anything was too low

8

Now I'll compare Neil and Gupta

I can't report a complete video on Twitter, so if you believe I've cherry picked, reply and I'll answer the point.

The approach
1. AN/SC statements.
2. I will answer as if I held SC's views. But I'll do so objectively & scientifically (OS)

9
On case illustration.

AN uses data points to claim it's inconceivable we'd reach 50K cases.

SE "I would tend to agree".

OBJECTIVE SCIENTIST (OS)
It's an illustration Andrew, not a prediction.

Even so we'll likely get to that level, though maybe a bit later than mid Oct

10
On modelling

SC: Projections must be challenged on probability and likelihood. Maths alone gets you to "extraordinary" numbers, and you miss the "why".

OS: The illustration may not look at all the factors. So it could be too low/high, but is probably not unreasonable.

11
Viral Growth Theory

SC [their models] apply to an epidemic that is just taking off. That's not the case for the UK.

OS: The risk is that we apply new virus projections to an existing virus. We'd have good reason to do so but we might still SLIGHTLY over-estimate risk

12
SV: We don't know how many people are immune to #Covid19 because the count of antibodies isn't a good indicator of immunity

OS We don't understand immunity. But antibodies are short lasting, and we see infection; Precautionary Principle applies https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30764-7/fulltext

13
SC As a result it's very hard to project the impact of this disease. The growth rate, especially in winter, could be wrong. Like seaonal flu grows.

OS: As a result we should proceed cautiously. We project conservatively and adjust if, as I hope, we see better immunity.

14
AN We were sold a lie by complicit scientists so politicians could put in a new Lockdown

SC it's not for me to say what the intention of the scientists is.

OS I can't speculate Andrew. But I see no evidence of a lie. Data is reasonable. I take issue with proposed action.

15

SC we must think broader than economy. Look at all factors affected by Lockdown.

OS we must think broader than the economy, all factors affected by Lockdown. But knowing if we crash the NHS, the consequences of that for economy, cancer, hunger is worse than any Lockdown.

16
AN The policy response to Covid is Lockdown. Is that the right approach?

SC No, it just delays things.

OS No, it's a disaster that we're relying on it. But it's crucial to delay the rate of viral growth until we have working alternatives.
Otherwise, we overwhelm the NHS.

17

SC if you could delay without cost, sure, do it. But you can't, and these costs are mounting.

OS: I fear the costs of lockdown are severe. But the costs of an overwhelmed NHS would undermine the economy completely.
We desperately need a lockdown alternative.

18
AN: We must learn to live with it. Agree?

SC Very sensible, from the start we should have let it become endemic

OS. I wish we could get to endemic equilibrium.
But we don't YET have a path to exposure that avoids calamitous damage.
And trust me, I've tried to find a way

19
AN what did you say to the PM?

SC I told him we should allow it to circulate until it's endemic, protecting the vulnerable

OS: I said, we must get to the point where, with less damage than Lockdown, it can be endemic. To do that we need [I don't know, she didn't say]...

19 (cont)
...or to do that we need to protect the vulnerable but we know at the moment to not overwhelm the NHS, protecting the vulnerable would look suspiciously like lockown for the over 50s and many conditions in younger.

So instead we must [didn't say].

20 Summary

I do not hold Sunetra Gupta's views. In science, that isn't a problem.

In playing an objective scientist holding controversial views, I've shown you how science allows detached answers while maintaining a challenging position

Gupta fails to do this
Every time

21

There was a time where we could rely on objectivity to represent an average consensus view while continuing to challenge it.

THAT is the nature of science.

Gupta and her parallels in Climate Change, economics, GM etc. show we can no longer rely on impartiality

22

The implications are significant.

Journalism must learn about the scientific method and why consensus exists.

And not by using the BBC's failed balanced approach.

Gupta holds (I guess) a 1% view.

Played against one other scientist, she would look like a 50% view.

23

A journalistic approach for a sole challenger interview is:

1. to challenge back representing the consensus view

2. make it clear a fringe opinion is represented

3. be severely critical, when proposed ideas cannot be implemented.

Like Gupta, Neil failed
Every time

24

It's not for me to say Journalists must adopt this approach, merely that in so doing they avoid bias and misrepresentation.

But if they choose not to do so, and cherry pick fringe theorists holding their biases, then it's not science, or journalism.

It's advertising

Postscript 1
This thread needed to stay relatively impartial.

If you'd like some more opinionated background reading:

[Why are we failing, how we could win?
https://twitter.com/atatimelikethis/status/1317409867527761920?s=19]

[Lancet best practice https://twitter.com/atatimelikethis/status/1316456962188611584?s=19]

[How they got there https://twitter.com/atatimelikethis/status/1318178096298807298?s=19]