Heading into autumn - what next?

[cuccir]

So it seems overall things are better at the end of september than they were at the start. Cases and deaths are rougly the same. Hospitalisatoins on the other hand are down by about 28%.

Yes I think this is a fairly accurate assessment.

My sense, from reading what both more-inclined-to-restriction AND less-inclined-to-restriction (very loosely classifying people into two opposing camps when I know most are more in-between!) experts have been saying, is that there's a little bewilderment at the picture in England and Wales in particlar at the moment - cases (and hospitalizaitons/detah) remain relatively high compared to other countries with similar vaccination rates, but have been osciliating since early August, even though that osciliation is creeping up marginally (higher peaks, lower troughs) due to seasonality.

But the received wisdom is that we should be expecting one or two more 'waves' of infeciton before we reach this sort of endemicity (the height of the waves essentially being determined by vaccination rates and behaviour). So the question is whether we reached endemicity earlier than anticipated, or if something else is going on (eg different peaks and troughs passing through different age groups or regions which, on a national-scale, looks like oscilation?).

 

[35B]

I was at a (primary) school governors meeting yesterday. Medical absences this term are above the norm, with attendance in the low 90%s compared with our (and national) expectation of 95%+ pre-Covid. No patterns, except possibly Covid running through families, are obvious, but non-Covid absences are also up on the usual pattern. Our clerk, who deals with several schools, commented that reported experience elsewhere varies considerably, and that other schools in the general area are significantly more affected.

This fits with my childrens' (secondary) school experience, where Covid absences are higher, and including teachers.

 

[initiation]

A reminder that the modelling experts have got it hopelessly wrong again. Grey lines are what was published at the start of September and SAGE described as the most likely scenarios for hospitalisations (with 95% confidence intervals shaded). The outcome in red is approx. 50% below even the lowest confidence interval.

They got schools returning wrong, they got final unlocking wrong, they got this wrong. How much more belief can people have in them? Remember these forecasts will be used to influence government policy.

 

[MikeWM]

I was at a (primary) school governors meeting yesterday. Medical absences this term are above the norm,

Unsurprising. Even setting aside the damage we've done to children's immune systems by keeping them isolated for much of the past 18 months, parents aren't going to send their kid to school with a mild case of the sniffles or a cough, when they may have done so before, for fear that other parents (or their teachers!) may accuse them of trying to kill grannies.

but non-Covid absences are also up on the usual pattern.

Also unsurprising - if you tell people consistently for 18 months that attending school isn't vital, some parents (and teenagers) are going to start believing it and acting accordingly.

 

[cuccir]

A reminder that the modelling experts have got it hopelessly wrong again. Grey lines are what was published at the start of September and SAGE described as the most likely scenarios for hospitalisations (with 95% confidence intervals shaded). The outcome in red is approx. 50% below even the lowest confidence interval.

They got schools returning wrong, they got final unlocking wrong, they got this wrong. How much more belief can people have in them? Remember these forecasts will be used to influence government policy.

View attachment 103666

The data is not that far off the 95% confidence of the lower projection is it? Both are projecitons that Sage modellers made - but of course the media headlines report that they only made the top one, and then they are accused of wildly over-estimating when in fact they offer multiple plausible scenarios.

 

[philosopher]

Yes I think this is a fairly accurate assessment.

My sense, from reading what both more-inclined-to-restriction AND less-inclined-to-restriction (very loosely classifying people into two opposing camps when I know most are more in-between!) experts have been saying, is that there's a little bewilderment at the picture in England and Wales in particlar at the moment - cases (and hospitalizaitons/detah) remain relatively high compared to other countries with similar vaccination rates, but have been osciliating since early August, even though that osciliation is creeping up marginally (higher peaks, lower troughs) due to seasonality.

I have just had a look at Google Mobility data on OurworldinData and people in the UK seem more cautious than most other Europeans. In the UK people are generally staying at home more and generally visiting public transport stations, workplaces and entertainment places less than other European countries. This I think debunks the argument that the higher cases in the UK is due to people’s behaviour.

So I wonder what is causing cases to be persistently higher in the UK then the rest of Europe.

 

[yorksrob]

I have just had a look at Google Mobility data on OurworldinData and people in the UK seem more cautious than most other Europeans. In the UK people are generally staying at home more and generally visiting public transport stations, workplaces and entertainment places less than other European countries. This I think debunks the argument that the higher cases in the UK is due to people’s behaviour.

So I wonder what is causing cases to be persistently higher in the UK then the rest of Europe.

That is an interesting question.

Do we still have more prevalence of the Delta variant compared to the continent ?

 

[MikeWM]

So I wonder what is causing cases to be persistently higher in the UK then the rest of Europe.

Testing too much? According to Worldometer, the average UK citizen has done 4.5 tests by this point (!). Of fairly sizable 'western' countries, only Denmark and Austria exceed that, and no countries with a population comparable to the UK come close.

 

[Cdd89]

I think this is likely; the OWID test positivity map suggests the UK is about the same as the rest of Western Europe (some better, some worse). Of course that’s an imperfect measure too since different tests have different probabilities of returning positive.

It’s getting harder to compare countries. With vaccination coverage the demographic distribution of those protected is more important than the final percentage. Even on deaths: imagine country A mandates vaccination with a high (involuntary) uptake, while country B does not and gets a lower uptake. Country B would have far more deaths, but it would be an oversimplification to say that A is therefore better.

 

[brad465]

Definitely some truth to this. All of a sudden the focus is going to be on cost of living issues.

That said, there's still "dragged back to the office kicking and screaming" types trying to play the "there was someone on my train without a mask, it's not safe" game. Not sure they're being listened to any more now though.

My organisation has optional office visits which are being taken up, but still capacity limits which is annoying, and seem to be rather slow on acknowledging how we are moving on.

A reminder that the modelling experts have got it hopelessly wrong again. Grey lines are what was published at the start of September and SAGE described as the most likely scenarios for hospitalisations (with 95% confidence intervals shaded). The outcome in red is approx. 50% below even the lowest confidence interval.

They got schools returning wrong, they got final unlocking wrong, they got this wrong. How much more belief can people have in them? Remember these forecasts will be used to influence government policy.

I wonder if it's possible that they were correct with R number estimations, but are overestimating how many infections will lead to hospitalisations. Perhaps SAGE have forgotten a rather significant thing started happening to the population from the end of 2020 that reduces severe illness?

 

[Darandio]

The data is not that far off the 95% confidence of the lower projection is it? Both are projecitons that Sage modellers made - but of course the media headlines report that they only made the top one, and then they are accused of wildly over-estimating when in fact they offer multiple plausible scenarios.

And it's SAGE who whipped up hysteria when these projections were made based on the worst case scenario with talk about needing a half term firebreak lockdown this month. It's SAGE members who said the government weren't listening to them over these projections so they went into every media outlet going to generate frightening headlines.

As their prediction of catastrophic doom hasn't come off again, if people call them out and concentrate on them getting it hopelessly wrong again then tough.

 

[35B]

Unsurprising. Even setting aside the damage we've done to children's immune systems by keeping them isolated for much of the past 18 months, parents aren't going to send their kid to school with a mild case of the sniffles or a cough, when they may have done so before, for fear that other parents (or their teachers!) may accuse them of trying to kill grannies.

Also unsurprising - if you tell people consistently for 18 months that attending school isn't vital, some parents (and teenagers) are going to start believing it and acting accordingly.

Interestingly, the qualitative feedback did not support any of those assertions. At that school, I got no sense of any pattern additional caution from parents of that sort, or of parents treating school as inessential.

It's also worth comment that no one present sought to downplay that school closures had been disruptive, and there was particular concern from teachers about the Year 2 cohort who have had both their years in school significantly disrupted, and where special efforts are having to be made to help some of those children.

 

[MikeWM]

Interestingly, the qualitative feedback did not support any of those assertions. At that school, I got no sense of any pattern additional caution from parents of that sort, or of parents treating school as inessential.

It's also worth comment that no one present sought to downplay that school closures had been disruptive, and there was particular concern from teachers about the Year 2 cohort who have had both their years in school significantly disrupted, and where special efforts are having to be made to help some of those children.

Well, fair enough, you were there so you should be able to judge that better than me You did say that 'non-Covid absences were up' though - is there a suggestion of why that is?

 

[35B]

Well, fair enough, you were there so you should be able to judge that better than me You did say that 'non-Covid absences were up' though - is there a suggestion of why that is?

It was hard to tell, and in a school of 350-odd, the differences are small enough that it's hard to infer patterns from what will be single digit numbers of additional absences - as was said, a single multi week absence due to chickenpox will have a measurable impact. My personal gut feeling, based on the figures that I saw, is that there are a few more autumn bugs doing the rounds than usual which would be lost in the rounding were it not also for what are definitely a number of Covid absences, with the combined impact taking attendance from "normal" to "below normal".

 

[Eyersey468]

And it's SAGE who whipped up hysteria when these projections were made based on the worst case scenario with talk about needing a half term firebreak lockdown this month. It's SAGE members who said the government weren't listening to them over these projections so they went into every media outlet going to generate frightening headlines.

As their prediction of catastrophic doom hasn't come off again, if people call them out and concentrate on them getting it hopelessly wrong again then tough.

Agreed. I have no sympathy for SAGE if they feel they are hard done by with people calling them out over yet again incorrect projections

 

[DelayRepay]

Agreed. I have no sympathy for SAGE if they feel they are hard done by with people calling them out over yet again incorrect projections

I have no sympathy either. I appreciate that models will never be perfect, and I appreciate it would probably be irresponsible not to understand the worse case plausible but extremely unlikely scenario, which I think is what the top line in the graph posted shows.

The problem is that the media pick up on the worse case scenario as being a forecast and report it. And when they do, individual SAGE members go on the TV and tell us we all ought to wear masks and prepare for more restrictions. In fact they should be explaining that their 'worse case' scenario is just that - something to help the government and NHS put in place contingency plans but something that is extremely unlikely. Rather like we have contingency plans for nuclear fallout and all sorts of other things that we never expect to use but it would be irresponsible not to have.

Anyway, I am very pleased. I was a bit worried about September and the impact of schools returning. I think if things were going to go wrong, it would have happened last month.

I think it's almost exactly a year since the 'Tiers' were introduced. What a pain that was. Thankfully there doesn't seem much prospect of them returning.

 

[yorksrob]

I think the problem with SAGE's modelling is that the best and worst case scenario's vary so much as to make them useless. It would have been better for them to admit that they don't know.

 

[DustyBin]

I think the problem with SAGE's modelling is that the best and worst case scenario's vary so much as to make them useless. It would have been better for them to admit that they don't know.

Exactly.

I once terminated a contract with a cost consultancy who were doing exactly the same thing. They’d present figures at feasibility stage ranging from unfeasibly low to ridiculously high; it was literally useless information that even a layman could have provided. It begged two questions; firstly why not just say you don’t know and secondly, if you don’t, why are you marketing yourself as a cost consultancy?!

The parallels with SAGE are rather striking, they however seem to get away with it.

 

[initiation]

I think the problem with SAGE's modelling is that the best and worst case scenario's vary so much as to make them useless

Yes. And the higher curve in the previous graph isn't even the highest they presented! The actual worst case they presented has hospitalisations going off the chart!

I have no sympathy for modellers claiming that the media or public is treating them poorly or not seriously. They have consistently provided poor forecasts of what is going on and then used their new found 'fame' to try and terrorise the public via the media.

 

[Eyersey468]

I also accept models will never be perfect but what has annoyed me instead of making it clear that there has been a range of possible scenarios SAGE has gone running to the press with the worst case scenario every time and made it out that it was the most likely.

 

[duncanp]

I also accept models will never be perfect but what has annoyed me instead of making it clear that there has been a range of possible scenarios SAGE has gone running to the press with the worst case scenario every time and made it out that it was the most likely.

Exactly.

I have used the lottery analogy before on these forums, but it is worth repeating here, in view of the news about the Euro Millions draw.

I could buy a ticket for next Tuesday, and a possible scenario is that I win the jackpot of £184 million. But it is far more likely that I will win nothing.

Fortunately, I get the feeling that the government has (finally) picked up on the reality of SAGE scaremongering, with their "Winter COVID Plan" mentioning a full lockdown as a very last resort, as evidenced by the fact that the furlough scheme has finally ended.

And have you noticed there are far fewer headlines along the lines of "<insert scary prediction of armageddon>....warns SAGE scientist.." in the papers these days, along with a very welcome period of silence from Devi Sridhar, Susan Michie and all the other locktivist pains in the backside.

 

[Cdd89]

Susan Michie

She pootled off to Portugal a couple of weeks ago. I don’t recall her saying travel rules should be loosed, so presumably she must have thought they should be relaxed a few weeks after the government did so.

 

[35B]

Agreed. I have no sympathy for SAGE if they feel they are hard done by with people calling them out over yet again incorrect projections

I found Tom Whipple's review in today's Times of the new book by David Spiegelhalter and Anthony Masters "Covid by Numbers" quite interesting (behind paywall, sharing link provided). Respecting their copyright, I won't share the whole piece, but in the context of the discussion on modelling, this passage about Neil Ferguson's modelling stuck out:

And to prove the lack of an agenda they are also critical of that most maligned of Covid figures, Neil Ferguson. The Imperial model he produced arguably kicked off the lockdown. To this day people point out that we have not come close to the 510,000 deaths it predicted.

Yet this is not, Spiegelhalter and Masters say, the right metric to judge it on. The 510,000 was what would have happened in the “highly implausible” event that we would have done nothing — not even made voluntary changes. And the number it came up with in the event of full lockdown? 48,000. They point out that 150,000 deaths later that figure “has turned out to be rather optimistic”. Who would have thought it? Professor “Lockdown” Ferguson, the Covid optimist.

Given that pandemic is still unfolding, I'm reminded of what Zhou Enlai allegedly said when asked about the effects of the French Revolution - "it's too early to tell".

 

[yorksrob]

I found Tom Whipple's review in today's Times of the new book by David Spiegelhalter and Anthony Masters "Covid by Numbers" quite interesting (behind paywall, sharing link provided). Respecting their copyright, I won't share the whole piece, but in the context of the discussion on modelling, this passage about Neil Ferguson's modelling stuck out:

Given that pandemic is still unfolding, I'm reminded of what Zhou Enlai allegedly said when asked about the effects of the French Revolution - "it's too early to tell".

Or alternatively one could argue that he (like many) had an over inflated opinion of the efficacy of lockdown.

 

[yorkie]

Anyone defending Neil Ferguson has zero credibility.

 

[bramling]

Exactly.

I have used the lottery analogy before on these forums, but it is worth repeating here, in view of the news about the Euro Millions draw.

I could buy a ticket for next Tuesday, and a possible scenario is that I win the jackpot of £184 million. But it is far more likely that I will win nothing.

Fortunately, I get the feeling that the government has (finally) picked up on the reality of SAGE scaremongering, with their "Winter COVID Plan" mentioning a full lockdown as a very last resort, as evidenced by the fact that the furlough scheme has finally ended.

And have you noticed there are far fewer headlines along the lines of "<insert scary prediction of armageddon>....warns SAGE scientist.." in the papers these days, along with a very welcome period of silence from Devi Sridhar, Susan Michie and all the other locktivist pains in the backside.

I think it's also the case that people are now, finally, simply getting bored of this. Nothing too horrific has happened as a result of July, and with the schools going back it would probably have done so by now if it were going to. More and more people are back to work so are no longer living the zombie lifestyle of the last 18 months, and being back to the reality of work they no longer wish to have their leisure time curtained by petty restrictions like having to pre-book attractions, or spoiled by having to wear masks. Many will have been mixing with others in various settings without anything awful happening.

I spent the afternoon at an attraction yesterday, and whilst there were the customary laminated signs about masks on the door as well has hand sanitizers round the place, I didn't see a single person doing either. Everyone was interacting normally, no dirty looks, the only thing different was people slightly more spaced out, which isn't a bad thing anyway.

Unless the increasingly marginalised pro-restriction lot manage to whip up fear over a variant or something, I think for many this is over now, it's gone on too long and it has lost salience as an issue. In any case events are overtaking, and we're now getting into the territory of the longer-term consequences, and it's not looking pretty. The inflation is a concern, but the powers that be are frit of raising interest rates because it could trigger a pack of cards falling down. This is why they shouldn't really have been reduced to the extent they have been.

 

[DustyBin]

I found Tom Whipple's review in today's Times of the new book by David Spiegelhalter and Anthony Masters "Covid by Numbers" quite interesting (behind paywall, sharing link provided). Respecting their copyright, I won't share the whole piece, but in the context of the discussion on modelling, this passage about Neil Ferguson's modelling stuck out:

Given that pandemic is still unfolding, I'm reminded of what Zhou Enlai allegedly said when asked about the effects of the French Revolution - "it's too early to tell".

What exactly was meant by “full lockdown” though? And over what period of time were these deaths forecast to occur? And have 150,000 people really died of covid?

This is, in my opinion, a case of attempting to defend the indefensible. Ferguson isn’t fit for purpose and neither is SAGE as a body. If they’re the best we have god help us.

 

[35B]

Anyone defending Neil Ferguson has zero credibility.

What exactly was meant by “full lockdown” though? And over what period of time were these deaths forecast to occur? And have 150,000 people really died of covid?

This is, in my opinion, a case of attempting to defend the indefensible. Ferguson isn’t fit for purpose and neither is SAGE as a body. If they’re the best we have god help us.

Fair questions, and the time horizon is my major reservation; I would say a truer comparison would cut off in summer 2020.

More broadly, Spiegelhalter was not defending Ferguson; he was criticising his statistical modelling. But the point made stands - Ferguson is criticised for scaremongering, yet the Covid death toll of the last 18 months far exceeds what he proposed as likely with “full lockdown”. That is - as the review makes clear is true of the whole book - a bracing reminder that all is not as we conventionally assume.

 

[yorksrob]

Fair questions, and the time horizon is my major reservation; I would say a truer comparison would cut off in summer 2020.

More broadly, Spiegelhalter was not defending Ferguson; he was criticising his statistical modelling. But the point made stands - Ferguson is criticised for scaremongering, yet the Covid death toll of the last 18 months far exceeds what he proposed as likely with “full lockdown”. That is - as the review makes clear is true of the whole book - a bracing reminder that all is not as we conventionally assume.

It suggests that lockdown didn't work (at least as well as it was portrayed as).

 

[yorkie]

All lockdowns do is buy time; if a vaccine is in development then in theory you could have very few deaths if you locked down as soon as the virus was known about, until all vulnerable people were vaccinated. A very harsh lockdown that was well observed would still have resulted in transmission as people would have still had to carry out certain essential tasks and some jobs clearly have to continue during a lockdown. But that would have been over a year; it would have been unaffordable and untenable.

The sort of people who call for lockdowns are either the type who have well proportioned homes, work from home jobs (or are retired etc), gardens, are able to visit the countryside, and are more interested in trying to stop other people doing things than anything else; they will still go out on a regular basis themselves, while urging others to 'stay home' or tend to be far-left authoritarians who are keen to fall into line with their far-left authoritarian comrades and allies.

People who use terms like "full lockdown" are probably indicating that they want other people to be further restricted, with more businesses closed, more debt, more jobs lost etc, but of course they would still have indulged in whatever activities they saw fit themselves.