UKH

Is there perhaps a bit of paranoia on UKC about new users automatically being labelled 'Sock Puppets', etc. if they start a thread about Covid-19?
I refer to this thread:
https://www.ukhillwalking.com/forums/off_belay/so_sweden_then-739006?v=1#x9516...
The OP seems genuine and the Link is to an article on the BBC Website and not some crank anti-Covid one.
Any thoughts?

Yes. You are OffWidth and I claim my five pounds.

I understand your post and point, but I wouldn't call it "paranoia". I might call it "justified scepticism"

Not sure who should be more insulted with that comment - Offwidth or myself...
 

Well, one day, someone is genuinely going to sign up to discuss climbing and then get a sudden urge on a whim to start a thread on a Covid topic that's aimed in a certain direction.

If they're genuinely new to the site and have not been lurking for a while, they're probably going to be surprised by the quantity of skepticism they receive.

I think in general there's a difference between the "sock puppets" and the accounts being signed up for to push views towards one side of the spectrum on Covid.  We had one earlier this week that started with a lost and found post I believe before going south and getting deleted.

I was planning to sit back and watch the thread to see which way it went.  One of the MOs I've seen a few times is to start with a couple of anodyne posts (some of which have been climbing related in the past) and then post a reasonable link for discussion, and then to try and nudge things in a certain direction from there. It's much more likely to survive than an initial link to outright nonsense.

Who signed up near this poster, anything suspicious looking?

aaronsmith823674 - https://www.ukclimbing.com/user/profile.php?id=322704
aaronsmith349857SS - https://www.ukclimbing.com/user/profile.php?id=322705

Nothing suspicious there, move along...

Would you say that the thread in question should have been Pinged?
It asked an interesting question and the information linked to came from a valid and trusted source.

How about banning the aaronsmith profiles instead of questioning the integrity of someone who happened to join at the same time?

I assume you're not accusing me of questioning the integrity of someone who happened to join after these two accounts, because I do not have the power to ban the profiles as you suggest and you are presumably aware of this.

I also see no signs of the site owners questioning the integrity of anyone on that thread.  They locked it - as I would if I was moderating and wanted to get a good nights sleep, not wondering what kind of crapshow might be awaiting me in the morning...

To be clear, I am not questioning their integrity.  As I said I was going to sit back and watch the thread and see how it went...  Holding off from jumping to conclusions for now.

• I have noted that the process of "new account > anodyne climbing post > start Covid topic" all in the space of a day is not exactly new around here, and tends to go in one direction.  
• I am noting that there are a couple of suspicious profiles immediately before the other account

People can draw their own conclusions (edit, some of which will doubtless be about me...)

Clearly the latter Aaron Smith is the boulderer Aaron Smith. It's given away in his username what with the SS at the end.

I think we should be able to have an open debate about this stuff and also welcome new posters. However, the unfortunate reality is that the site has been bombarded with bad actors for 18 months now and Albert's post fits an all too common pattern.

Maybe he's genuine, maybe it'd be better to hear him out before shutting him down. Then again, maybe it was just the opening salvo of yet another bullshit boreathon.

I think it's correctly graded at HVS but can see that some E1 leaders might find it a bit of a handful.

Okay, I went down the rabbit hole seeing as you suggested paranoia and I can't watch the thread pan out.

Exhibit 1: Their post here on learning to ice climb made today on UKC

• https://www.ukhillwalking.com/forums/winter_climbing/ice_climbing_skills-73898...
• Yes Rjukan I think, and have done plenty of summer climbing, mostly single pitch, and skiing. Never climbed on ice or worn crampons though. Obvs will have some gear to buy!

Exhibit 2: Their profile photo - of a Brocken spectre on what looks like burnt land with one yellow patch of grass

Exhibit 3: That photo is from a blog "Wee Black Dug" and the blog poster's linked Flickr account:

• http://www.weeblackdug.co.uk/?m=0 (See "Favourite Photo", right hand side and the Flicker link just below it)

Exhibit 4: From one of the posts on the blog a decade ago:

• http://www.weeblackdug.co.uk/2012/01/ben-vrackie.html?m=0
• We were armed to the teeth with crampons, ice axes, bothy shelters and the like, but in the event nothing other than a bit of wheezy footsore perseverance was needed to get us up to the top and back to the car in three & half hours.

These four exhibits are not all compatible with one person.  Something is fishy.

Edit: only a desktop view of the blog has the “photos” sidebar it seems, not a mobile phone view - I updated the link above this link for the desktop version: http://www.weeblackdug.co.uk/?m=0

I think you are too generous. Note peculiar name, US spelling, recent registration, Covid discussion in a particular  direction as well as Wintertrees points

It *could* be genuine but given these points and numerous similar posters with similar attributes whi have been shown not to be, I'd say they have to prove they are, not the other way round.

Brilliant!

Also great. 

1.  The mods have intel and can tell a fake.
2.  Fully get the idea of asking as many people as you can about a subject, but since Alberts friends have been to Norway multiple times before, why not ask your mates? Multiple times to Norway = plenty of experience in my book. 

Ed: and if Albert has been unfairly booted off the site, where’s his mates saying ‘WTF’ ?

Do you not think it’s a bit odd for someone to join to ask a reasonable climbing related question and then go on to start a thread about Covid which leans in a certain direction?

That’s a very good point about why not ask their mates. Probably because those sock puppet accounts haven’t been set up yet. They will be along shortly, offering advice on climbing WI7 and asking about the Covid situation in South Dakota or wherever. 

Over on the other Covid thread, I posted data taken directly from a UK government report and it has now been labelled with this warning.

"Misinformation. The numbers here meet no recognisable criteria for death rates, and are on the order of 50x lower than the case fatality rate for Covid and on the order of 20x lower than an annualised death rate for Covid across the whole population as based on current deaths."

The thread has been closed. This is a link to the report in question:

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/...

So yes, I'd say there is a disturbing amount of paranoia around at the moment.

Factually incorrect.

You posted a number you derived from a value in a table.  You posted it with a description that was different to that in the table you got the data from.  Context is key and an accurate description (yours was highly misleading) is as critical a part of data as the numbers.  Ask NASA about that one.

• Your description boiled down to: “death rates from Covid […] for unvaccinated 30-39 year olds”
• An accurate description the number you posted would be “percentage of the unvaccinated population of 30 to 39 year olds to die in a specific 3-week period”.  
• An accurate description of the source data would be “number of unvaccinated 30 to 39 year olds dying in a specific 3-week period expressed as a per 100,000 rate”
• You gave no citation to the source, no citation within the 6,000 word 33 page source to the bit of data you extracted, no method for how you got a percentage value when not given in the report, no qualification of what your “death rate” meant (when it was applying to a 3-week period which you hadn’t stated).

I can’t decide if you deliberately set out to find a way to give a “death rate” 20x to 80x too low, or if you’re so out of your depth you still don’t understand what you’re missing.

It filled up.  Those ones often do by Thursday or Friday.

Yes. A table published by the government.

There was no description. It's simply the most recent data published by the government. I provided the name of the person who originally provided the link in my original post, and then the link itself when I realised you were arguing without having read the report.

I'm glad you started this thread, as I didn't get chance to challenge a claim by the OP of that thread, or indeed the BBC article he linked.

"Sweden is doing better than its neighbours" - Not by a long chalk compared to Norway and Finland (factor of 10) or the perhaps more comparable Denmark (3 X worse).

"Sweden has some of the lowest infection rates in Europe" - Currently a third higher than Denmark, higher than Italy, Spain and loads of other places. That is ignoring the fact that they don't really do much testing.

I have explained this as clearly as I can. I’ll try again.

You posted a number with a massively misleading description and no source, you just alluded to something said by another poster, without linking to the specific post out of their many posts.

Do you seriously expect most people to figure out which post you mean out of 200 odd posts, follow the  link to the report and then figure out which number in which table over 33 pages you derived your number from, and then read all the details to understand what your number meant?

You gave a number bereft of context with an incorrect description.  

• Did posting their name tell other readers that your “death rate” was nothing of the sort?
• You got their name wrong….

> and then the link itself when I realised you were arguing without having read the report.

• I read the report.  That is clear from several posts.
• I eventually dragged the “in 3 weeks” qualifier out of you but only after stating it first.  I almost think you didn’t actually understand it until that point….

If I hadn’t dragged it out of you, your first post would have remained grossly misrepresentative, giving as it did a death rate between 20x and 50x too low by any usual interpretation.  

Yes.  I understand that.  I am not a simpering half-wit, you see.

I also note that the government gave a detailed description of their data in their table which says X, and your description of it was Y, where Y is not even a subset of X.

Say, do you remember the last time you posted a number from an official source but in a highly misleading way by changing the context around the number?  You got banned from that particular thread, so you started a new one with a name like “Banned for posting misinformation”.  By the end of that thread you wished untreated cancer on those disagreeing with you and got yourself fully banned and the thread deleted.  Now you keep misrepresenting the cause of that ban as your “inconvenient opinions” on covid, rather than your wishing death on other posters.

I’m out.

You appear to be criticising the BBC article as opposed to the person that posted a link to it and asked questions pertaining to it.
At the point of the thread being pulled, there was absolutely nothing contentious within it. 
It appears that the thread was pulled simply because a new user started a thread about Covid and I assume someone reported it for that reason - Panic: A new user has mentioned Covid!
They must be guilty, burn the heretic...

I'm responding to both the OP of the other thread and the link he posted. I'm refuting both, I hope that is clear, I doesn't seem a difficult concept to me.

Evidence appears to suggest that they are somewhat guilty.

All a bit OC12 and you claim that there's no paranoia involved...
Also, what you posted doesn't prove anything.
 

Here's the actual posts from the original thread:

I thought Sweden was generally rubbished by us here in the UK for having a disastrous approach to managing Covid? But according to the BBC they are doing just fine despite no lockdowns and a slightly lower vaccination level than we have in the UK.
Anyone know anyone in Sweden who can verify if this is true?

Far better at the moment apparently according to the BBC.

Neither of those appear to claim anything controversial and both appear to be asking genuine questions.
Maybe you read it differently?

IT shows the account holder is using a photo from a 12 year old flicker post of a person who has worn crampons many times, whilst they claim never to have worn crampons.

I did not claim it proves anything, I noted the various parts were not compatible with one person.  I can think of many different possibilities, I have not claimed any of them to be proven:

• The UKC poster is the blog poster and is telling fibs about never having worn crampons.  I consider this pretty unlikely
• The UKC poster has re-used the blog poster’s photo as their profile photo, with their permission.  Seems both unlikely and deceptive.
• The UKC poster has done a google image search and ripped off the blog poster’s image for their profile photo.  Which is exactly what a bad faith poster seeking to create a minimally viable profile would do, long with a couple of other posts.
• The UKC poster is mates with the blog poster, and it’s a photo of the UKC poster that the other guy has on his blog and flicker.  Odd that the UKC poster never asked his mate about winter stuff, ey?

The UKC poster fell exactly in to a pattern.  Nobody other than you has called “burn the heretic”.  Given how many other threads have gone - and how some have served as springboards for more outrageous new account posters - I can entirely see why the thread was locked on a Friday night assuming the site owners don’t want to wake up to a crap storm tomorrow.  It looked to me like regular posters were going to engage in good faith on that thread.  I did not ask for it to be closed or report it or the poster in any other way.  I can see why it was locked.

You seem very tetchy over this - are you suffering from the heat and humidity and failing to sleep as well?  

Edit: To jump in on another bit of the thread:

I read it as you did, but I also doubted the OP’s intentions for the thread given that they fell in to a really obvious pattern we’ve seen many times before.  I also note the comment from another poster that the site operators have access to more information than thee and me on poster identities and can make a more informed opinion.

I never claimed that there’s no paranoia. 

I haven't claimed controversial. I have hopefully demonstrated that it is wrong. Paul or whoever from UKC can answer why they think it is controversial, but of course being wrong will be a contributing factor. 

The evidence that Albert is a bad actor is pretty clear, but of course I only suspected that at the time.

For once, I’m sort of glad you’re back. Let me clarify something for you. As I understand it, you took the number of Covid deaths per 100,000 from the table and claimed that this is the CFR. However that is not the case.

It is the number of Covid deaths in a 3 week period per 100,000 people in a given age range. Not per 100,000 cases. Clearly the vast majority of those notional 100,000 people didn’t have Covid in that particular 3 week period or in the few weeks preceding that 3 week period.

Therefore the % does not represent what you claim it to represent. It is a fairly meaningless % on its own - its only real use is for comparative purposes, eg against other age groups for the same period.

The % you alluded to is the number of deaths per 100,000 cases in a given age group, ie the CFR. The table does not show this. 

I thought it was quite a poor piece of journalism. It’s true that their case rate is lower than some EU countries and the UK but it’s also broadly similar to a few and quite a lot higher than some. It’s also fairly meaningless to look at case rates at a given point in time in isolation because different countries have followed different trajectories, each with their own ups and downs. Better to look at cumulative case and death rates.  

Yes, it’s claiming something controversial - that the BBC is saying that Sweden are “doing just fine”. Firstly, the article doesn’t say this. It just says the current case rate is relatively low and the government is looking at removing remaining restrictions. So the poster is claiming the BBC is saying something which it isn’t actually saying.

Secondly, “doing just fine” is something you might reasonably say about New Zealand or China or other countries which have got away relatively lightly in terms of deaths, hospitalisations and economic damage. I think saying that about Sweden is rather stretching the definition of “doing just fine”.

How about if I posted something like: “This article says Sweden is only now looking at removing restrictions, whereas Boris did that almost two months ago. Just goes to show that England is doing just fine. Anyone in Sweden able to explain why it’s taken them longer to get on top of Covid?” Would you have an issue with this? I’m not claiming anything controversial and just asking a genuine question, right? 

FTFY

As someone who has lived in Denmark throughout the pandemic, I have suffered from a small number of Covid-sceptic friends/acquaintances in the UK endlessly telling me about how Sweden "got it right." Without fail this has simply been cover for their own anti-lockdown views. 

It's oureed. 100%, all in.

To elaborate a little, I spotted this account yesterday after its first post and thought "hmm... Bet the 2nd or 3rd is about covid". It's getting that predictable now.

One or two inane generic posts about something vaguely climbing, usually plagiarised text to sound genuine, then straight into the covid stuff. Their copy is super distinctive, and the attitude and tone is always back to "what, so you don't agree with this thing I'm pointing at when taken out of context?" very quickly. It's oureed/alyson30/rawn1962 or whoever. I don't understand how you could think otherwise. The only thing different from the last 20 or so incarnations is that this one has a profile pic (robbed off Google images)

Mr Ross just left the building.

It was only recently I came to understand that “this user is currently restricted from posting” is something a user can trigger themselves as well as the site owners. 

They made one more post on the ice climbing topic before checking out or being checked out this morning.  I wonder if they read some other threads and realised their credibility was blown.

One bit of good news is that it seems all the sleeper accounts leant more credibility by dint of being many years old have been burnt through this summer.  Who sets up half a dozen sock puppets 5 years in advance?

Well, quite.

It’s been an education to me how widely apart different peoples’s suspicion-o-meters are calibrated.

Still lost when it comes to the motivation; he/she/it keeps getting nowhere. I can only think, like we've said before, they've already completed Mumsnet on expert mode and UKC is the end boss fight of super trollio bros. Kinda nice to think it is, in a way.

Lot of similarities in language and the way they argue between the covid denialism pop up poster and the rom-like poster.

One consistent interpretation is that they are compelled to troll for their kicks.

The waters are muddied by a different person who also keeps either getting banned or restricting their own accounts and coming back to covid topics. 

Your patience and doggedness in exposing and refuting these tw@ts is outstanding! Please keep on but also look out for your own state of mind. Most people on here aren't taken in by the anti-vaxxers so it's remarkable that they keep on piling in.

I was also pondering this. UKC is pretty niche and tiny by reddit/mumsnet/FB standards. I can only assume there are one or two weirdos or bored ex climbers who are just using UKC as a bit of sport.

That may be some of it, but I don't think it's all of it.

Some of the pop-up poster's messages have had precise and unique sub-string matches (e.g. a whole sentence) between their content and similar posts made elsewhere.  For example, from investigating after one particular post, this timeline emerged:

1. A post appears on a specialist US board with nothing to do with climbing
2. Within 24 hours, a UKC post is made containing a whole, exact sentence from the US post in an otherwise similar message.  
3. This shared sentence has no other occurrences anywhere on the internet.

In the instance I did the most digging in to, the poser appeared to be a US based person who who advertised as a "social media influencer".

The problem with looking for conspiracies is one can always find one, be they real or not. I think the Star Trek TNG episode "The Drumhead" is an excellent deconstruction of this and of what happens when the anti-conspiracy side loose track of reality.  But,  there's more than a whiff of malign intent to a limited subset of the pop-up posters we've had...

It's deaths after a positive test. "Death within 60 days of first positive COVID-19 test or where COVID-19 is mentioned on the death certificate by week of date of death between week 32 and week 35 2021". I posted the link above for the 3rd time. Maybe take a look.

You will also see this: "In individuals aged 40 to 79, the rate of a positive COVID-19 test is higher in vaccinated individuals compared to unvaccinated."

I didn't include this in my original reply to you as you'd specifically asked for data on 'health consequences' for young people. I gave you the percentages for the most recent period available.

I have found the reactions on here to this latest government report quite troubling. It's almost like some people aren't happy unless unvaccinated people are catching Covid and dying in large numbers.

I think that Misha is correct in their interpretation and that you are wrong. 

The table gives the absolute number of people dying in that 3 week period by the criteria you give.  

The table then gives that number as a per-100k rate with reference to the cohort to allow unbiased comparison between the different rows in the table.  That is all that rate is good for - a population normalised comparator between rows in the table.  

You are being purposefully obtuse.

It is astoundingly clear the reactions are not to the contents of the report - which various posters had acknowledged and discussed before you came along to egregiously misrepresent its content.

You are seriously delusional if you think anyone here wants people to be catching Covid or dying of it.

I have been interested in communicating the data clearly against an often piss-poor and confusing treatment of it by the media and against what started out as understandable misunderstandings (people faced with what for many is a very non-intuitive situation) and has clearly moved on to misrepresentation from some and perhaps delusion from others.

I would far rather no unvaccinated people were needing medical care of dying because of Covid.  Because:

•  I don't like the idea of people suffering needlessly as a result of them being suckered in by malign and/or delusional people consistently under-representing their risks by 10x to 100x 
• If we have to go in to strict control measures because of unvaccinated people later this winter, it is others who will suffer most from the control measures, and there's been enough heartbreak and major disruption from school closures already, for example.

OK. Assuming some are bored old members and some are bad actors, what to they stand to achieve? A cursory glance at the related history shows clearly that UKC is a fortress of reason so if they are trying to be persuasive are there not softer targets? Plus the sheer lack volume of people in here should put this into the 'why bother' pile. 

Which makes me suspect that its mainly bored mischief making. 

You will also see in the very same report it explains why those numbers are not to be used to deduce or infer vaccine effectiveness. But you've chosen to remove the context, which is something you'd only do for misleading and dangerous reasons. If you're too thick or lazy to understand the report and population estimates fully it could easily lead to false conclusions. And you're going out of your way to try to lead people to them. That's not ok. You need to f**k off with that.

"That, detective, is the right question"

Just because we don't understand their motivation doesn't mean we should write it off as a possibility.

That would certainly be a comforting thought, wouldn't it? That there's no malign intent or manipulation going on out there and that we have but a couple of sad little trolls being all this.

I suspect a large fraction of significant forum engagement goes on across hundreds of websites, any one of which could fairly claim to be insignificant.  

Someone from UKC could chime in with some statistics on country of origin of the banned posters...  

Edit: forgot to reply to this

I think a lot more people read than post, for what its worth.

Well it would be good to be sure. My interpretation seems more logical and useful.

That wouldn't allow unbiased comparison at all as the population numbers are massively skewed towards vaccinated people.

Yes, yes it does.  It explains why this is not appropriate in very clear, accessible language.

Bit of a pattern emerging here, no?

This is my independent verdict over their last 4 incarnations.  Over them they have variously lied, misrepresented, removed context and argued with 100% consistency in a direction that seeks to downplay risks from Covid and seeks to over-state risks from vaccination.  They also have their illogical argument against even vaccinating younger adults due to their misunderstanding of evolutionary biology that has been highly conserved over several incarnations.

Seconded.

This is not just trolling, this is M&S trolling

(Too obscure? Maybe someone will get it....)

Yes, it is good to be sure before spouting off repeated death rates that are 20x to 80x lower than any actual rate.  Funny how you just made up a definition (that is not in the report) that fits with your long standing and ill informed views on the situation.

Your interpretation has remained so vague (deliberately so?) that it's hard even know to understand what it is.  I have tested the number you gave against the two reasonable interpretations of what you may mean, and they all came out with rates 20x to 80x higher than the one you gave.

Nonsense.

The table has these rows:

• Rates among persons vaccinated with 2 doses (per 100,000)

• Rates among persons not vaccinated (per 100,000)

Your understanding off all this seems so incredibly poor I'm amazed that you've been asserting yourself with such confidence.

*Sucks air through teeth*

The report should be retracted imo. The population estimates are bollocks, so the per 100k numbers are necessarily bollocks, and extra mega bollocks in the unvaccinated columns. This simple fact is the reason why it looks like covid is more prevalent in vaccinated people, and why the line that lingeringfart quotes is in there. It's obviously not compatible with a vaccine with positive effectiveness.

I think I agree with you that the uncertainties over the unvaccinated cohort size in the oldest ages is sufficient that the per 100k data is potentially very inaccurate for those line items in the table.

But I haven't tried to get that across to said poster, because until they understand what the table purports to show, it's just going to confuse things even more to try an explain why some of the line items are very questionable.

It's really hard to arrive at a view that is fair and unbiassed on some of these "demographic, cohort normalised" issues; it's one of the reasons I've generally avoided plotting or interpreting them. It's one of the reasons I like the exponential rate constants so much, they tell us about change, and do so in a normalised way that is more "fair" to compare between different ages.  I get that they're not much help with unpicking the vaccination status however.

That many be hard; it is very easy to see someone with the stick, however.

I think we both look at that table and know that additional information not presented is used to determine the "per 100k" columns, and we know that it comes with error bounds (or CIs) on it.  The report doesn't need retraction, it just needs those error bounds propagating through to the per-100k columns.  Although as we've seen, posters motivated by some underlying view rather than a genuine intent to understand the data the just repeat the central values when making comparisons, without giving the bounds...

But it does make it really easy for shitty people to pull out that one line and make it look like vaccines give you covid. They shouldn't have put that per 100k column in the report imo, because they don't know if 100k is really 200k or 20k. It lends itself to being misrepresented in exactly the way our unflushable visitor has.

And it's because M&S used the track 'albatross' on those ads. Albert Ross...... C'mon people. Tough crowd....

Yeah, I’m too slow to get that sort of thing.  I think we have two different people here - although it’s hard to tell with so many accounts playing silly buggers over identity.  Why do I think this?  « Rom » has to work hard at playing the fool…

Albatross: a wearisome burden. 
God I’m slow today. Well spotted. 
BTW, a black bowed albatross has been resident off the Yorkshire coast all summer. 

Which is why I never considered that data this inaccurate could be published. If the death/hospitalisation rates were per 100,000 who have had a positive test it would be much more reliable. If it turns out Wintertree's and Misha's interpretation is correct, I agree that the table should be pulled or revised.

Obviously the case rates have to be as a proportion of the general population.

How does a penguin build its house?

And more people are killed in car crashes wearing seat belts than not wearing them.  What does that prove?

Male emperor penguins just snuggle up to their mates all winter. The females fck off to the sea and eat fish. Talk about equality!

Total misrepresentation. For the record I completely support the outing and removal of pop-up posters who have clearly come on UKC for the first time to talk about covid in a dishonest way. I know the moderators and trust their ability to use their rules, especially with the extra information they have that we don''t. I just think on the science thread scientists should be kinder to questioning  from established posters who get the science wrong (or just ignore the persistent annoying ones, like oureed)....any scientist confident in what they say shouldn't ever need to make personal attacks on such threads. I've also talked extensively about Sweden (the subject favoured by this latest banned OP) who had a lot more death than their nordic neighbours, got things very wrong about herd immunity  but they still had science driven restrictions, some of which like allowing outdoor small group meetings in the first wave, that the UK should have allowed under the science.

The data table in question shows rates per 100,000

When I first saw this I went ‘ey up, this don’t look so good’.  But reasons could include:

1. unvaccinated = Covid deniers = can’t be arsed getting tested (because this would prove they are full of sh!te)  

2.  unvaccinated = health issues so can’t take the vaccine so take loads of extra precautions. 

3. being jabbed increases your chance of catching Covid.

But given you stand a waaaaay higher chance of dying if you’ve not been jabbed, I’ll sack off 3.  and go with 1 and 2 (and no doubt plenty of other reasons). 

For the record, I actually think there is a degree of Covid paranoia, but don’t really see it stifling debate to any great extent.  

Yeah I’d have given up a long time ago. There must be lower hanging fruit to pick elsewhere. Some kind of delusional narcissist?

I love the smell of bullsh*t in the morning. 

“If it turns out” - You can tell yours was woefully incorrect because I gave two different worked examples with sources and methods, both of which came out orders of magnitude higher than your interpretation.  You dismissed these.

When faced with a 50x to 80x discrepancy between your interpretations and my worked examples, and with no support in the report for the conclusion you jumped to, why did you refuse to compare my workings and see if the discrepancy lay in your interpretation or my workings?

I was as constructive as I could possibly be.  It’s taken until now to get the penny to drop for you.

The process of independently checking something is a critical part of trying to be a fair player in this.  It seems you misunderstood something catastrophically in a way that’s aligned to your long standing stance and then refused to engage with any attempt to confirm it.  Incompetence or deliberate?  Who knows.

The way LSRH and I worry the oldest line items for the unvaccinated could be misleading is totally unrelated to how you misunderstood and so misrepresented the data.

As this thread shows, those columns are open to accidental or deliberate misrepresentation by those who are happy to understate the risks by 20x to 80x.  I feel like we’ve seen that before…  

It seems me me that you are conflating two totally seperate issues here to try and slough away your total failure to understand the table.

Well, not if broken down by vaccinated and unvaccinated status…

There are two completely different metrics.

Deaths within 60 days of a Covid test, ie broadly speaking deaths from Covid, for a given age range, in a 3 week period. I think we are agreed on that.

This number of deaths is then applied to the estimated number of people (not cases) in that age range to derive the % you referred to. That does not mean it’s a rate per 100,000 cases. It’s still a rate per 100,000 people.

You either do not have a fairly basic grasp of statistics or you are being deliberately obtuse to suit your agenda or both.

As regards the Covid positivity rate in vaccinated vs unvaccinated 40+, that is a more complex point, partly due to imperfect demographic stats (there are fewer unvaccinated 40+ than vaccinated and we don’t have perfect population numbers, so there is a larger error margin as discussed above) and partly due to behavioural factors (people who chose not to be vaccinated are probably less likely to get a Covid test as well). This has been discussed above and I’m not going to waste my time repeating it.

UKC, Fortress of Reason. Sounds like an old B-movie title.

Now we have the slippery weasel pinned down to admitting that their "death rate from Covid" for deaths (table 5) and  was indeed the case fatality rate (CFR) in their mind ("If the death/hospitalisation rates were per 100,000 who have had a positive test it would be much more reliable. If it turns out Wintertree's and Misha's interpretation is correct"), this raises an absolutely fascinating question about what they thought the corresponding "per 100,000" column in table 4 was showing.

Table 4 is is"COVID-19 cases by vaccination status between week 32 and week 35 2021" and presents an almost identical layout to table 5 including an identically labelled column "Rates among persons not vaccinated (per 100,000)".

It is unquestionably clear that the "per 100,000" normalisation is the same for both tables.  So if it was deaths normalised to cases (which it was not), what are cases normalised to?  If they were normalised to cases, all values would be 1…… Bit of a clue there about the scale of their misunderstanding.

They display such an abject lack of critical thinking in evaluating their understanding of the report, that they have failed on two really obvious points:

• Their interpretation of table 5's normalisation  is fundamentally incompatible with table 4's
• Their interpretation of table 5's normalisation is out by a factor of around 50x from an independent estimate of the numbers if interpreted in their way.

> You either do not have a fairly basic grasp of statistics or you are being deliberately obtuse to suit your agenda or both.

If they are being genuine, they have an incomprehensibly strong belief in their ability to communicate data onwards given their staggeringly poor grasp of what they're reading.  A familiar pattern.

I'm glad you put "or both" in there.  In the interests of fairness, it's important to keep an open mind.

Its a shame they didn’t engage with my first constructive attempt to show them a worked example of why I thought they had misunderstood it.

No, your interpretation is wrong and deliberately so to suit your agenda. I haven’t sought to interpret the numbers. I have simply explained how the numbers are calculated. That is factually correct. You are wrong.

Which is why the report says the numbers are not to be used fir determining vaccine effectiveness.

I also note another poster with a pretty strong bias stating most Swedes are happy with this approach.  Not the ones I know, nor the general opinion I see on Swedish TV (and yes I understand the language adequately thanks )

I've taken the rate per 100,000 as a subset of the cohort identified in the title of each graph. You have taken it as a subset of the entire population regardless of the cohort.

I think my interpretation seems the most logical but quite happy to accept yours. Can you point to anything in the report or elsewhere that will confirm this.

They usually spell advice as advise too.

These are tables, not graphs.

Let's look at table 4. 

• Title: Table 4. COVID-19 cases by vaccination status between week 32 and week 35 2021
• Final column: [Case] Rates among persons not vaccinated (per 100,000)
  • Values - between 367.5 and 1520.8

If the "rate per 100,000" was a subset of "cases by vaccination status" ("the cohort identified in the title of each graph") it would be 1.  Because 1 times the number of people who have cases (the title cohort) have cases (the row values)

No, Misha has not.  Nor have I.  We have taken it as what it is, the absolute value divided by the population of the cohort as identified by the chart title (number of people dying by the criteria for (a) or (b) within a specific 3 week period) divided by the appropriate cohorts of the 8x2 two dimensional 2D matrix breakdown formed by the non-grey rows (8 age ranges) and final two columns (± vaccine status)

You were not happy to accept my interpretation - which is the same as Misha's - yesterday. 

I can point you to this worked example which shows how your interpretation (CFR by cohort) is out by orders of magnitude.

https://www.ukhillwalking.com/forums/off_belay/friday_night_covid_plotting_42-...

You ignored it and continued to stick to your outright wrong interpretation.

Another way of deciding would be to take the absolute numbers, to get the cohort sizes for the 8x2 matrix from the government dashboard and normalise it yourself.  I did this crudely for a few of them before my first reply to check that it was a ballpark right figure.  Given your refusal to engage constructively with my previous worked example, it hardly seems worth my time.  You however could to it to see if it proves you right.

This is embarrassing at this point.

So you’re saying if you took 100,000 vaccinated people in that age group, and 100,000 unvaccinated, more of the vaccinated ones would rest positive?

You still haven't answered the question: 'how does a penguin build its house?'

Yes, it's gone beyond paranoia - this place has become hostile to neuro-diversity in general.

Any challenge to the entrenched way most of us think, being drip-fed 'truth' from the Beeb, by way of example, is met with shouty exchanges and ridicule.

I advice you all to "check your privilege" 😁🙃😀

Somewhat ironic statement, given this thread was started as a response to a locked thread where someone was using a BBC article to support a point of view and a number of us have pointed out problems with the article.

It was! Thank you for noticing 👍

I thought it was ironic.

Perhaps I’m being too neuro-diverse in my interpretation…. 

I got labelled “an English tw*t” the other day because of my apparent eagerness to chow down on government propaganda!  

Edit: jinx.  Too slow to type that out…

No. It's per 100k NIMS unvaccinated people. They each represent somewhere between a lower leg and a large wedding.

Oureed is of course well aware of this but choosing to ignore it.

I actually laughed out loud at this.

Do you agree that the first table shows cases per 100,000 population in each age band (separately for vaccinated and unvaccinated)? The second table has the same captions and it would be bizarre if they switched the metric from per 100k people to per 100k cases from one table to the next, particularly as they use the same captions.

Another way of looking at this:

First table shows 113k cases in 30-39.

Second table shows 58 deaths 60 days post test in 30-39.

58 / 113,000 x 100% = 0.05%.

I feel like I'm howling at the moon here. NIMS estimates are wrong. Number of people vaccinated isn't wrong. So the number of people not vaccinated (total minus vaccinated) is super wrong. So the per 100k numbers are horse shit.

Edit: this is not to say you're making the wrong point. This report is only good for deaths/case and hosps/case. Which it shows pretty damn well, and is what you did in the 2nd bit of that last post. Strongly recommend sticking with that method.

Facebook sells me adverts at about 80 cents for 1,000 views of the post I want to boost.

Your last Friday night Covid plotting thread got about 4,000 views according to UKC.

No professional 'bad actor' trying to influence public opinion could be arsed spending hours on fake profiles and convoluted cover posts for 4,000 views of whatever sh*t they want to post.  They could put a post on Facebook, target the demographics and spend $3.20 to boost it.

If somebody is putting in this level of effort it's because they are have the tools for some other reason and are having a laugh using them to troll on UKC or because they are angry and have a grudge.   

Yes, I'm well aware what the published viewing figures on threads are on here.  Because I have eyes.

I'm also aware of how some marketing campaigns are run below the top tier of social media.  My point stands - many deeply engaged forum users are spread over many different websites.  Any one in isolation can say "but we are small".  

Anyhow, I would bow to your superior knowledge and expertise when it comes sock puppet accounts, but....

Do you remember that time we had a really bloody obvious sock puppet, and I called them out on being a really obvious sock puppet, and you spoke in their defence, including calling me "pretty paranoid in your old age"?

https://www.ukhillwalking.com/forums/off_belay/vaccine_side_effects_ctd-737935...

Although, now I revisit that thread, you did then go on to talk about yourself in the third person, almost as if you'd forgotten to sign in to the sock puppet account, so perhaps I should listen to what you have to say...

In terms of "having a grudge", the pop-up anti-lockdown posts have been appearing since before the first lockdown, which makes it seem very unlikely to be a grudge that developed in relation to the subject matter, and it would be very strange indeed to start advocating for action likely to kill random people as a result of a specific pre-existing grudge.

Interesting that MikeDuddy - the other half of the pair of socks - has not been shut down.

Indeed. The issue is our friend is claiming the table shows the CFR, which is not what it shows.

Anyway, deaths / hospitalisations per 100k population per year (not per 3 week period…), even if accurate, doesn’t say anything about the risk of contracting Covid. Its main use is for tracking the situation over time, as with the dashboard map. I think it will be interesting how the numbers evolve over time but on its own the table is of relatively limited use.

Of course my 0.05% is only indicative as the deaths in the 3 week period mostly don’t relate to cases in the same 3 week period. I was just using it to demonstrate that our friend was mistaken.

There's lots of ways of running marketing campaigns but if you are doing it professionally the numbers need to make sense.  Hours of time for 4,000 page views does not make sense.

I have zero knowledge of sock puppet accounts and near zero interest in them.   

My guess is somebody has particular views, they got ridiculed for them, there was a pile on and then they got angry and retaliated with sock puppets.   Just another day on the internet.

Maybe next time you could spend literally five seconds checking out a brand new poster, before calling me paranoid for suggesting they just might be a sock puppet.  Especially when they are the only person supporting your position...  As otherwise people just might get suspicious especially if you also suddenly start talking about yourself in the third person...

I feel like you're trying to tell me something...

Buying FB ads doesn't, on its own, generate the response a bad faith poster requires, though, does it? If the ads are mistargeted at a less-than-supine audience they'll just get blocked, mocked, or ignored.

What such posters want is engagement, i.e. discussion of the ad, hopefully with dissent, confusion, and angry division. Hence on FB, bad actors do exactly what you describe, creating fake profiles and using the comments to sow division. The same is true on every social media platform.

I think UKC/UKH and forums in general are more attractive to paid trolls than we realize. We think that they're insane to bother with us, a tiny community of niche-interest like minds. But I doubt they found us by deciding to go after climbers. Rather the attraction is that the forum is open, anonymously joinable, and posts show up quite prominently in Google searches. UKC is just another open door from the point of view of someone who's solely interested in finding open doors.

I've experienced this before, many years ago, when doing a live chat Ask Me Anything on a thoroughly obscure academic topic, hosted by an unsecured Uni server. 4chan trolls signed on and created mayhem - but they didn't find the conversation because of its subject; they found it because it was publicly-searchable.

I suspect this is true, except someone *both* already had the tools *and* is using them because they're encountering unusual opposition.

Yes, I'm telling you that you probably pissed somebody off who is now running sock puppets on your threads and your over-reaction is just filling the threads with even more off-topic nonsense.

There are far more efficient ways a professional bad actor could spend time and money to influence opinion.  Internet marketing is all about ROI and hours of work for 4,000 page views and if they are lucky 400 unique visitors - because most of the page views will be the same people coming back for more - is pointless.

Is it national irony day?  Moving beyond the issue of people adding noise to the Friday night threads...

The poster the OP refers to had nothing to do with the Friday night threads.  Nor did the last such pop-up poster this week.  Both signed up, made a token post or two, then went in to start a new thread on a Covid topic.

This has been happening since March 2020 when we had our first, carefully constructed and thoughtful looking post from a brand new poster that argued against lockdown with a lot of misrepresentation and basically eugenics.

That was something like 8 months before my plotting threads, and before I'd said much at all about Covid.  

So don't blame it all on me pissing someone off.  There are several separate bad actors here

1. The one(s) pushing against Covid control measures since March 2020 who I think are not UK based going off the observations made by others on spelling, and what their views align to.  You can't pin that on me.
2. Another individual who has been banned a few times and whose general posts and content suggest they are UK based.  They do seem to be making it a mission to fill the Friday threads with noise, but they're not the only one, are they?
3. "Rom" - someone with a 10 year history it turns out of abusing multiple sock puppets. They've lost all their sleeper accounts this summer (or so it seems....?) and are now left using new ones so look a lot like (1) and (2).  Their historic multi-account trolling has covered many subjects I don't weigh in on (Brexit for example) so I don't think you can pin that on me.  

The Friday night threads came together more as an organised opposition to the new accounts that would appear pushing highly loaded statistics and/or interpretations.  You've got your chicken and egg backwards with regards (1) at least which is the subject of this thread.  

I think your take on the "marketing" angle is not appropriate applied to posters falling under (1).  I think Bruxist is much closer to the mark. 

Robert F. Peston* has now also publicly done exactly what the report says not to and I've been screaming not to and oureed knowingly and malevolently did with a straight face.
I'm with Mainwood on this; we need at least to open the debate on whether NIMS numbers are becoming deleterious.

* you can guess what the F is short for

Except for your own, that you accidentally revealed in a previous thread, and (possibly for the first time ever) decided not to engage further in that thread when called out on it.

Yet, we regularly get commercial internet marketing SPAM posts.  I’m sure this one will be deleted soon.  They’re pretty common, and these are just the tip of whatever iceberg the site’s systems block I suspect.

https://www.ukhillwalking.com/forums/off_belay/swim_shirt_for_a_fat_people-739...

Edit: SPAM deleted.  I edited my screenshot to remove their URL as well.

I outed Rom about two years ago after noting weird patterns on threads where he was engaging with me in typical Rom fashion. He coughed to some of the sock puppets but claimed he’d been forced to create them because my “badgering” him was making him anxious that I’d somehow “track him down”.

I got no reply when I pointed out that the accounts in question pre-dated me joining UKC by quite a few years.

Unfortunately it looks like my sock puppets are all lazy b*stards who can't be bothered to agree with me or even give me a like.

Chapeau on the name by the way.  I'd have said it was a bit obvious to engage in a debate on sock puppets with an account named after a tool used by hackers but apparently not.

Sure, and how many seconds of attention and do you reckon something like that got.  Copying pictures off an ancient blog post on a relevant topic, creating a believable profile and engaging in a discussion on ice climbing is orders of magnitude more work.  It is completely out of proportion unless the motivation is personal.  

Wireshark is a network protocol diagnostic tool. Could be used by white hatters, black hatters, or just anyone wanting to diagnose network issues.

Its the defacto freeware tool for many network analysts.

I find it bizarre that the column headings don't correspond with the title of the table. Every time I read the rates per 100,000 I want to refer back to within x days of a positive test. It also seems bizarre that they choose to publish fairly useless data rather than more meaningful data. I agree the percentages seem ridiculously low, however, and am happy to accept I was misled by a poorly worded table in a government report. I apologise for replicating the data.

If Wintertree hadn't waded in with his unique brand of mockery and over-analysis (focusing on time period instead of population cohort), this issue could have been resolved in a couple of politely-worded posts. (See, for example, my exchange with Dave Garnett.)

I think you are probably onto something here. I am already anticipating Wintertree using my latest exchange with Misha as a means to delegitimise anything I have to say in the future, possibly even next year. Coupled with his jeering approach it makes for a very toxic environment for debate. Seems to be popular though, so I can't imagine anything will change.

1. Psychological warfare exists: "Our aim is to challenge the difficulties of modern warfare using non-lethal engagement and legitimate non-military levers as a means to adapt behaviours of the opposing forces and adversaries."
2. There's loads of reports of Russian social media influencing and disinformation.
3. UKC is insignificant compared to Facebook.
4. UKC is perhaps not insignificant compared to a Facebook group targeted for or set up for malign influence. 

I don't think UKC is targeted by Russia. I might be wrong and it might have been in the past. I vaguely recall posts with suspiciously consistent bad information. Now sock puppet accounts have been admitted, perhaps trolling. 

It does sound paranoid but 1-4 above have some basis in fact.

That wasn't context but a vaguely worded disclaimer which I'm hesitant to parrot.

If I had been trying to draw conclusions from the fact that Covid rates are higher in vaccinated people than unvaccinated people in age groups 40-79, I agree that some attempt at including context should have been made. As it is, I posted the raw government data as a basis for discussion. Frankly, any context anybody adds will only be based on speculation.

I really don't understand why I am the person whose attitude gets criticized on here!

I do. Many others do. It's been explained to you many times.

You seem to be the outlier in misinterpreting that.  

I am curious, and you haven't answered this yet despite it being asked, what did you think the /100,000 cases was in Table 4?  Case rates per 100,000 cases within x days of a positive test?  That should have clued you up instantly.

Excluding the oldest ages (issues over the accuracy of the unvaccinated denominator), the data is perfectly fair, useful and meaningful.  It allows comparison between rows in the table.  End of.

If they seem "ridiculously low", why were you so confident in them?  Something does not add up here.  And I don't mean in the maths.

That is your opinion.  It was clear to everyone else.  You can claim it, you can't accept it.  The issues LSRH mentioned and that you grabbed on to like a plank from a sinking ship are secondary to your misunderstanding.

You did not replicate the data.  You stripped it of context and you presented it with a context not present in the report to convert it into  something it assuredly was not.

You are misunderstanding or misrepresenting my response that you ignored.  I am surprised you did not engage with my good faith, patient and detailed reply given your admission that the false data you presented (because data includes the physical units - what you call it -  not just the numeric value) was "ridiculously low".

Here it is for others to read - I doubt they can be bothered - https://www.ukhillwalking.com/forums/off_belay/friday_night_covid_plotting_42-...

Re: "focusing on a time period" - the data that you misrepresented came from a specific 3-week time period.  I spent time and effort showing you how you could check the data you misrepresented by calculating it yourself from a corresponding 3-week time period. Using the same time period is the least biassed way of getting a confirmatory measure, given the demographic shifts over time.  I was not "focused" on a time period, it was part of an attempt at a minimally biased confirmatory measurement.

My post (linked above) was politely worded.   It was blunt, but it was not rude.

https://www.ukhillwalking.com/forums/off_belay/ukc_covid_paranoia-739007?v=1#x...

I don't see a resolved exchange.  I see Dave raising an objection to your interpretation/understanding and you dropping it quietly.

It reverse as another example of you being incompetent or dishonest when it comes to the data around Covid.  I can't say which.

Here's the interesting thing - over your accounts you have consistently been wrong in the same direction.  Mistakes from poor comprehension or sloppy work don't happen like that, sometimes something is too big, sometime it is too small. 

Let's be honest, by next year you're going to be under yet another identity.

Certainly seems a challenging environment for those seeking to push an entirely one sided view that consistently seeks to down play the risks from Covid by multiple orders of magnitude. 

Perhaps that's because I don't keep popping up under different identities consistently misunderstanding or misrepresenting anything I can get my hands on to push the view that things are 10x to 100x better than reality suggests, and I don't resort to wishing suffering through untreated cancer on those who disagree with me, and I don't deliberately solicit bans by lying egregiously about the risks from covid, and I don't keep popping up under now accounts, and I don't paint disagreement with me as some Orwellian nightmare.

You're not the other side of the debate.  

So many things about the government's Covid strategy seem ridiculous to me...

I explained the data and left him to check it. Just because I don't harass and insult people doesn't mean I'm dropping an issue quietly.

I've been correct about variant escape, waning immunity, Valneva and protecting vulnerable people. I also think I'm correct about the importance of sharing vaccines on a global scale although this is open to debate. The main reason that poorly-worded table misled me was because I am used to more rigorous presentation of data.

Are you claiming you don't understand the difference between "data" and "government strategy"?  That is stretching credulity behind breaking point.

No, you just wish untreated cancer on them or tell everyone to "go f*ck themselves" then come back with a different account don't you?

It's not exactly been a secret that viruses mutate Is it?  If you'd said the sun rises, you'd have been correct about that too...  I mentioned the possibility of worse mutations arising on March 21st 2020 [1].  It's also biology 101 that mutation and other variation goes hand in hand with evading immunity.  This is one of the reasons we're still catching colds and flu all the time...

You have been largely wrong on variant escape because you have consistently claimed it would be driven by vaccination but the evidence disagrees with the nasty variants emerging pre-vaccination or in very low vaccination cohorts, and with a lot of stuff suggesting chronic illness- which is much reduced by vaccination - drove that.

It looks to me that I was right on the ball 17 months ago in stating that letting the virus circulate uncontrolled would lead to worse variants emerging - much higher fatality rate with Kent, much faster spread with Delta.  You - over several accounts - have claimed that the virus would naturally "fizzle out" by becoming less lethal.  Robustly proved wrong by the evidence, but you get banned (deliberately it now seems) and dissociated yourself from your failures.

Which is why I don't think you are here in anything resembling good faith. 

You've been staggeringly wrong there, misrepresenting or misunderstanding the fade of antibodies as indicating an equivalent fade of immunity and claiming a rapid loss of immunity. Wrong.

Of course immunity fades, sun rises in the morning, that sort of thing.   It's about how rapidly it fades.

I don't recall you being "right" about that.  

No, you and your previous accounts have consistently misunderstood or misrepresented how vulnerable younger adults are by around a factor of 100..

Yes, you've found something to latch on to that has actually got strong moral and practical force behind it.

I needed a laugh today, thanks.  Your onwards presentation was so lacking in rigor it took half a dozen posts to get to the bottom of what you thought you were presenting, let alone what you thought the table shows.

I repeat: Here's the interesting thing - over your accounts you have consistently been wrong in the same direction.  Mistakes from poor comprehension or sloppy work don't happen like that, sometimes something is too big, sometime it is too small

[1] https://www.ukhillwalking.com/forums/off_belay/is_it_worth_it-717284?v=1#x9154...

It's happened - https://www.ukhillwalking.com/forums/off_belay/vaccine_its_going_wrong_already...

This was an obvious one, who knows if there are more subtle ones.  

I'm not sure that highly polarised argument against misinformation is much of a barrier to it being absorbed and repeated by those, mostly lurkers, who might be susceptible to the message. Indeed for some, being maverick, standing against main stream opinion is appealing in its own right. It's scattergun and it's hard to understand how such an approach might be motivated in the long run (is it paid work or delusional nonsense or weird hobby?) but for all that we shouldn't assume it's ineffective.

jk

Of course people know viruses mutate, but their understanding of how selective pressure affects population dynamics is often very poor. 

Oh, and another thing I'm hoping to add to my list is the greater breadth of natural immunity compared to vaccine-induce immunity, although more research is required to confirm this. 

So I've done a wintertree and scrolled back through my posts in the thread in question. I never mentionned an equivalence between the fade in antibodies and in immunity.

Requiring a 3rd dose after 6 months to boost immunity does suggest a rapid loss of immunity as compared to most other vaccinations

Here are some extracts from my posts in that thread:

"The drop in antibodies is considerable. Antibodies are not the only immune defense but this drop is enough for researchers from several countries to be concerned." 

"There may be a similar sharp drop in antibodies with regards to naturally acquired immunity but I haven't seen any reports on this."

"PM Bennet said: "Findings show that there is a decline in the body's immunity over time. The aim of the supplementary dose is to build it up again"

"Perhaps it will be limited to older people, or it may be that everyone will experience a significant drop in antibodies. We don't know yet. "

And from wintertree

"I do not think the emerging scientific research supports your stance."

"A third dose of the current vaccine targeted at pre-Kent variants is being considered for the most vulnerable specifically because it may provide extra protection against the new variants."

Wrong!

You’ve strongly and repeatedly hinted at “instead of” rather than “compared to” as overt justification to get infected first, disregarding the strong increase in health risks at all adult ages this brings.  Hint: natural immunity can “top up” vaccine immunity - so arguing a binary choice is wrong.

The differences between sorts of immunity have been discussed as nauseam on here many times, you seem alone in thinking you’ve been a lien voice here.  I’m not sure you always recognise the scientific language though?

Its there in the archives.  I didn’t specify which “posts concerned” - you seem to be picking one from after you tightened up your misunderstanding and/or use of language in response to comments.  Of course you still haven’t told us your past banned accounts so we can’t check those.  Convenient, no?

I can’t even understand what you’re trying to say in relation to this “wrong.” Anyhow you didn’t fully quote me (shock, horror).  Here’s what I said:

Further, there is more than a bit of confusion out there between fade of antibodies and fade of immunity, which are not the same thing - at all.  I do not think the emerging scientific research supports your stance.

I stand by that.  

Here’s the context to those quotes.  https://www.ukhillwalking.com/forums/off_belay/rogues_gallery-737543?v=1#x9498...  

Yes, I agree.  

• I have seen that very poor understanding in the confident assertion by a banned account - that I strongly suspect was you - that we should not be overly concerned, because coronaviruses tend to loose their virulence over time
• I have seen that very poor understanding when you have made similar statements about how its likely to "fizzle out" with future mutations.
  • I have seen you point blank ignore requests to evidence that statement, including from another poster who knows their beans professionally.
  • There are areas you never address, and if one pays attention they speak more loudly than all that you have to say.
• I have seen that very poor understanding reflected in your confident assertion on another thread today that "Fortunately Delta has proved to be no more virulent than the original virus." .  
  • That's what oureed says, what is the evidence starting to say?
  • https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(21)01358-1/... - Risk of COVID-19 hospital admission was approximately doubled in those with the Delta VOC when compared to the Alpha VOC, with risk of admission particularly increased in those with five or more relevant comorbidities.
  • https://www.medrxiv.org/content/10.1101/2021.07.05.21260050v3.full - Results: Compared to non-VOC SARS-CoV-2 strains, the adjusted elevation in risk associated with N501Y-positive variants was 52% (43-62%) for hospitalization; 89% (67-116%) for ICU admission; and 51% (30-74%) for death. Increases with Delta variant were more pronounced: 108% (80-138%) for hospitalization; 234% (164-331%) for ICU admission; and 132% (47-230%) for death.

Once again, you have posted something that is so seriously misinformed as to be embarrassing.

Once again, the misinformation you have posted seeks to downplay the risks in our current situation.  

Again.

A recurrent pattern across your various incarnations, no?

Thanks for accepting. I agree they could have worded the captions more clearly by adding ‘people’ after ‘per 100,000’.

Funny that because the CDC does not share your certitude. This is what it has to say:

"The Delta variant is highly contagious, more than 2x as contagious as previous variants."

"Some data suggest the Delta variant might cause more severe illness than previous variants in unvaccinated people."

They then refer to the 2 studies that you linked, one of which has not been peer-reviewed. Do you have more recent data to confirm? I couldn't find any.

Sorry, but what is the explanation for the last two columns in Table 4 of document linked above which suggests greater Covid rates among unvaccinated than vaccinated.  Adding "people" to the heading doesn't seem to change the odd numbers to my mind.  Robert Peston brought this up too.

Where in my post was I certain?  Let us review what we each have said:

• oureed: "Fortunately Delta has proved to be no more virulent than the original virus."
  • You draw a certain conclusion ("has proved") with no evidence.
• Wintertree:  what is the evidence starting to say?
  • I am questioning, not concluding, and I am giving evidence.
  • I am giving no certain view on the virulence.  You are misrepresenting me

> They then refer to the 2 studies that you linked, one of which has not been peer-reviewed. Do you have more recent data to confirm? I couldn't find any.

Those are the studies I have reviewed, one of which has been peer reviewed.  Both are at odds to your statement.  

I note that the CDC do not appear to have referenced any papers giving evidence lethality has not increased.

Let us recap:

1. You made the unqualified statement that goes against emerging evidence.
2. I drew attention to  evidence that suggests (1) is looking to be on thin ice. 
3. You implied I have "certitude"

I have claimed no certainty.  I have shown your certain, evidence and reference free statement is opposed to peer reviewed evidence and evidence emerging from a credible source.

I am keeping an open mind, and I am not making a one sided statement without any evidence.

[Screams into pillow]

That doesn't help.  On the face of it the numbers *are* very odd. What's going on?

Please don't make me type it out again. Just scroll up. Or ctrl-F for NIMS.

Googling suggests NIMS is "National Immunisation Management System"?   Your posts suggest it is something to do with population numbers but it's hard to believe they are so far out.

My understanding:

• V=The number of vaccinated, very old people is very well known (precisely known?)
• N±n=The number of very old people is not so well know with an uncertainty of ±n
• U±n=(N±n)-V = the number of very old people not vaccinated and is calculated as I show.  The uncertainties in N translate directly to U.

For old people:

• N is quite large, U is quite small.
• n/N is small.  Anything normalised to N (whole population) is subject to a small error.
• n/U is small.  Anything normalise to U (unvaccinated cohort) is  subject to a large error.

This report does not recognise the uncertainty in the U denominator for old people and does not translate this into an error bound or CI for the values in the table.

The uncertainties aren’t normally that significant, but because one large number is being subtracted from another large number of the same order of magnitude, the uncertainty on the result is large, proportionally speaking to the result.  That result with its relatively large uncertainty is then being divided in to another number, giving a relatively large uncertainty on the result.

Didn't mean google, just read my posts above. Never mind, I'll type it out again, because I don't think you'll be the last to skip over the discussion so far and come in with that question.

Believe it. They're miles off, and then when you subtract a well known number (vaccinated) from the total (fairly crap) and divide by what's left, you get, unsurprisingly, complete crap.

https://www.ukhillwalking.com/forums/off_belay/friday_night_covid_plotting_42-... 

https://www.ukhillwalking.com/forums/off_belay/ukc_covid_paranoia-739007?v=1#x...

https://twitter.com/JamesWard73/status/1436017485262823424 

Remove every instance of 'old' from that and you're about there. 

Edit: Here's a chart. It shows.... probably not what you think it does. Nobody knows which is more wrong, ONS or NIMS, but everyone knows we don't know how many people there are. https://mobile.twitter.com/paulmainwood/status/1407256437286375425

Edit again: and my earlier explanation from t'other thread reproduced here for completeness:

If unvaccinated=Wild ass guess - vaccinated, then the huge error bars get slapped onto a much smaller number. Further clarification for the sake of the thread, let's take an example: If you have 1m +/- 5% people, and you vaccinate 900,000 of them, how many are unvaccinated? Could be 50k, could be 150k.

What if the initial population estimate was off by 17% instead of 5%? Do the sums.

My dear Wintertree

Communication is complex. Do not underestimate the subtleties of the English language. Take these 2 sentences.

- ‘Research has proved that Delta is no more virulent.’

- ‘Delta has proved to be no more virulent.’

Although similar, the first sentence requires evidence to have been found whereas the second sentence simply requires that something has been shown to be the case.

I do agree with your remark about my use of the word certitude, however. I originally used ‘confidence’ and should have stuck with that. The CDC does not share your confidence about Delta being more virulent. 

I’m just staggered.  You give every sign of wanting to be taken seriously.  You consistently misunderstand or misrepresent some things and apparently just make others up, always in the same erroneous direction.

You’re the one that made a certain statement without evidence.

I’m the one who questioned that with some evidence.

I never said I am certain that delta is more transmissible.  Nor am I confident in that.  You keep misrepresenting my very clear post.

I am confident that you made an unqualified statement without evidence presenting as certain something that is at odds with emerging evidence in the form of a paper that has been through robust peer review and another from a credible group that has been prepared for peer review. We are not at the time for certainty in this, we are at a time for keeping an open mind and listening to emerging evidence.

> Although similar, the first sentence requires evidence to have been found whereas the second sentence simply requires that something has been shown to be the case.

One requires “evidence” and the other “something has been shown to be the case”.

1. I feel you are deeply confused here about what “evidence” means….
2. You have brought nothing to the table to show your statement “to be the case”.  If you had, and if it was credible, it could be labelled as “evidence”…

Did you mean to say n/U is large?

I did, thank you.

What a mistake to make.  I am a twit.

Well OK I can see all that.  But why calculate (let alone publish) things that way? To get a rate, you don't need to know the population.  I cant see any comment in the report about this issue either. It looks like a pretty big mistake to make given the concerns over vaccines.

[Screams into pillow]

https://www.ukhillwalking.com/forums/off_belay/ukc_covid_paranoia-739007?v=1#x...

https://www.ukhillwalking.com/forums/off_belay/ukc_covid_paranoia-739007?v=1#x...

The issue of uncertainty over population estimates came as a real shock to me when I started diving into Covid data.  I don't think I'm alone..

I believe their noble intent was to give a fair way to compare the rates of infection, hospitalisation and death between vaccinated and unvaccinated individuals.  

• As they don't have data on exposure events that can't calculate an "exposure > infection > detection as case" rate, so the only option is to normalise per-cohort population within each vaccine status and age cohort 
• They could do based table -  case > hospitalisation and case > fatality rates to normalise those tables, but the infection > case rate will have strong confounding factors such as demographics, and this would introduce an asymmetry in to the report between cases (per-population) and deaths & hospitalisations (per-case).
  • A dead giveaway that it was normalised to population is that per 100,000 is a very common unit for population-normalised values but to the best of my knowledge has never been sued for conversion ratios (CFR, CHR etc).

> I cant see any comment in the report about this issue either. It looks like a pretty big mistake to make given the concerns over vaccines.

The whole issue over different population estimates seems to be a bit of an elephant in the room in various reports.  I've not followed the dissections and debate by others that LSRH has, they may chime in with more...

I think you are right and I see no sign of the report acknowledging this issue, however it gives other reasons why the data should not be used to infer vaccine effectiveness under the heading "Interpretation of the data".  Emphasis below is mine.

These data should be considered in the context of vaccination status of the population groups shown in the rest of this report.

The vaccination status of cases, inpatients and deaths is not the most appropriate method to assess vaccine effectiveness and there is a high risk of misinterpretation. Vaccine effectiveness has been formally estimated from a number of different sources and is described earlier in this report.

Indeed.  I think it was Si dH who suggested a few weeks ago that perhaps more detailed broken down data was not being routinely released because it could be open to bad misunderstanding or misrepresentation.  Recent discussion here confirms that - thankfully someone else has been following the whole population saga and clued us up to the sub-issue over the denominators.  

I think it's really poor form that they've not made some attempt at qualifying the bound of uncertainty in these columns.  If they had, and if they had any basic QC processes on outputs, they'd probably have pulled the columns.

I am not calling for data to be censored to avoid accidental misinterpretation or deliberate misrepresentation, but this does seem to illustrate the pitfalls of not anticipating such shenanigans and not testing the quality of their methods, bounds and presentation against the risk.  I suppose they felt they retired it with the rider I quoted above, but I think even with that these columns are a disaster.  

I'd quite like to see some ONS bods sent over to PHE on a training mission.

I don't really have anything to add beyond the links and words I've thrown in, and highest recommendation of Mainwood. 

• NIMS is crazy-wrong, in all ages.
• We don't have anything 'better'. 
• The method presumed used to infer unvaccinated numbers amplifies the error by a big multiple
• I don't understand why I'm the only one screaming about this

A distinction without a difference in common usage I think, although I wouldn’t use either in a scientific paper.

Anyway ´virulent’ isn’t a very precise term in this context.  The evidence seems to be that the delta variant has a higher multiplicity of infection, with infected individuals becoming infectious two days earlier than those infection with alpha variant and having about a thousand-fold greater viral load.  This adds up to it being about twice as transmissible.

https://www.nature.com/articles/d41586-021-01986-w

That doesn’t seem to imply that it’s more pathogenic per se, so what does virulent mean?

A distinction without a difference in common usage I think, although I wouldn’t use either in a scientific paper.

Anyway ´virulent’ isn’t a very precise term in this context.  The evidence seems to be that the delta variant has a higher multiplicity of infection, with infected individuals becoming infectious two days earlier than those infected with alpha variant and having about a thousand-fold greater viral load.  This adds up to it being about twice as transmissible.

https://www.nature.com/articles/d41586-021-01986-w

That doesn’t seem to imply that it’s more pathogenic per se, so what does virulent mean?

Thanks for the post. I'm still rather astounded this was presented and also that it hasn't been urgently clarified. Robert Peston is a bit of a prat but he seems to have unfairly maligned for highlighting  this.

Why do you need the whole cohort? Why not take a 1000 (say) randomly selected unvaccinated in a cohort and count how many get it? Aren't there regular surveys doing almost exactly this?

I've not really had very high expectations of the PHE reports for some time.  I always count it as a bonus when they manage to embed a PDF of the Excel plot instead of a low-res screenshot.

I agree that it should be clarified. I don't understand why it's not been more officially addressed.  I've only recently twigged the scale of problems it makes from LSRH's hi lighting of the issue. I've long been a fan of looking at the change in a measure over time rather than looking at absolutes, it's a more robust way of understanding if things are getting better or worse against all sorts of unknowns.  With these tables being new, there is no historic data to compare with yet; once there is, the normalised columns aren't needed for that purpose.

I think that is an organisational issue around data sources, briefs and so on.

•  PHE have full access to whatever longitudinal formats the mass individual data exists in, and they produce summary reports of that for private and public consumption.   
• The ONS have a brief to conduct random sampling surveys in association with other partner institutes, and they produce and publish those.

There are pros and cons to each kind of analysis, and there is a lot more confidence in taking them together than either in isolation.  

PHE are doing what PHE are supposed to be doing, they're just explaining their data source (for age/vaccine status cohort populations) or qualifying the error bounds to a professional standard.  

I used to know a couple of really good data people in public health.  They both left shortly after the regional health observatories were subsumed in to the then-new PHE and now apply their talents elsewhere in the world.   

I haven't looked yet.  Peril sensitive sunglasses have jumped in.  I should look.

Am I on mute???

People are a slippery bunch and difficult to count.

In the absence of a national ID card, requirement to immediately register/deregister where you live, requirement to register/deregister with GP/NHS, recording of those entering/leaving UK including Irish border breaking GFA and rapid deletion of the deceased from stats etc etc etc the population and to a lesser extent NIMS is a bit* uncertain.

*a bit may be a million or two

Can you see any way round that?

If not, all that can be done is accept, minimise and estimate uncertainty.

I still reckon a million is almost neither here nor there when there is not a single person in the UK known to be fully immune.

The way round it is to avoid doing what I illustrated in https://www.ukhillwalking.com/forums/off_belay/ukc_covid_paranoia-739007?v=1#x... and then deriving stats from the massively erroneous results and publishing them in official reports. And if you call yourself a journalist, to check suspiciously non-credible conclusions before publishing.

I don't think you're picking up what I'm putting down here.... When it comes to cases/100k unvaccinated people, when you're working out the number of unvaccinated people as wintertree and I variously illustrated, a million is very much here, and indeed, there.

Was never specified; I was taking it to mean more health damage, higher hospitalisation and fatality rates and so on.

To jump off topic into some pointless musings that probably aren’t in the right language for a subject bod…

When a virus itself is pretty harmless but it’s total novelty when presented to an adult means their immune system goes destructively bonkers, is it even pathogenic at all?  

If the virus had been in circulation for decades, young children would be exposed and sneeze it off, moving on with immune protection.  Fatality rates in adults - who were once children sneezing it off - would probably be much lower.  

The virus would be the same as what we have now, so it’s innate properties would by definition be the same.  But the pathogenicity would be very different,  This means pathogenicity against humans can not in this instance be an innate property of the virus, but is an emergent phenomena of the interacting system of humans and virus.  I suppose that is always the case, but it seems particularly so here?  Or am I over complicating this and pathogenicity is a broad church based on the medical history of individuals?  Is there anything “special” about a virus that subverts the immune system instead of some other part of the host?  

Statement of the obvious - relative error rises a lot when you subtract the majority.

A million is almost neither here nor there for population level immunity or rise/fall in the number of cases when not a single person we know of is fully immune.

Sure, it doesn't matter if you're not talking about what we were talking about.

If you are, which we were, it really really does.

Here: https://theconversation.com/covid-19-why-infection-rates-among-double-vacci...

Now fingers crossed maybe we won't have to explain it again. Maybe??? 

New Zealand's latest Covid wave is Delta. There have been around 1000 new cases. At the start of the outbreak only 20% of the population was fully vaccinated. There has so far been 1 death.

"Delta has proved to be no more virulent than the original virus" seems a reasonable statement to make. 

Observational studies in France and Israel suggest the same thing, especially with regards to young people. 

It's about the seriousness of the disease. The most obvious parameters to look at are hospitalisations and deaths.

I think your explanations weren't as clear as you perhaps think, and they only looked at the maths.

I still can't see a real explanation for why the approach was used when apparently the problems are well known, beyond Wintertree's point about different organisations not talking to each other. PHE have basically given antivaccers promotional material and then said nothing.

Only if you ignore controlled scientific studies and instead take uncontrolled data (where you are ignorant of the demographics) from an in-progress wave and divide current deaths by current cases.  This neglects the absolutely critical impact of lag on fatality rates.  It’s like it’s March 2020 all over again.

You are so full of it I’m surprised It’s not coming out of your ears. 

You’re either here in bad faith or you’re wilfully delusional.  

Sort yourself out before someone believes you and ends up dead.

Apart from wasting eons if your time on the internet, what is your central point?  What are you advocating? What do you want to get from these discussions?  Keep it to the point and simple please.

The explanation kinda is maths, isn't it?

Me neither

Yes. This. 

Still not convinced you've read the posts where I said basically that. Here are the links again: https://www.ukhillwalking.com/forums/off_belay/ukc_covid_paranoia-739007?v=1#x... and https://www.ukhillwalking.com/forums/off_belay/ukc_covid_paranoia-739007?v=1#x... 

The death was on 3 September and there have been none since. For sure there are probably more to come, but remember we're talking about virulence relative to the original virus. If Delta were significantly more virulent I would have expected to have seen more deaths already. The wave started mid-August.

This is just weird. Why would Delta being no more virulent than the original virus make anyone want to change their behaviour?! In any case, if people act on what I say they will be getting double-jabbed and tested regularly.

Yes, which will change your deeply flawed result by a factor of 2 with the next death, then 3 etc.

It was an example of one factor you have not controlled for in your deeply flawed interpretation.  There are many, many more obvious factors that mean a top level number - especially in a small and intensely suppressed outbreak - can’t be compared with the fidelity needed to prove your point in virulence.

Thats why the peer reviewed manuscript I showed you and the pre print prepared for peer review by a credible institution are worth something and your Ill informed view is worth less than the dog much I just stepped around.

Because, like every other misunderstanding or misrepresentation you and your former accounts have made, it seems to downplay the risks.  A bit here, a bit there, always the same direction.  If you confidently proclaim such crap in other places online or offline without check you will be pushing your deeply flawed view over the risks.

The reason I brought up virulence is that it's possible for you both to be right, depending on how you define it.

From a clinical and epidemiological point of view, it can be difficult to distinguish a low dose infection with something really pathogenic (because it targets critical cells or tissues, like the nervous system) from a high dose of something less specifically nasty but which causes a lot of damage by infecting and killing a lot of cells (including ones in important places), releasing a lot of toxins or stimulating a massive inflammatory reaction.

I tend to think of the former as more virulent because they are highly pathogenic at relatively small numbers (rabies virus or Clostridium botulinum, say) , but lots of people die every day from overwhelming infections from common, weakly pathogenic organisms (E coli, candida) that are in the wrong place and not being controlled effectively.

So, in principle, the SARS-CoV-2 delta variant might not be any more pathogenic per virion than the alpha variant, but if the typical viral load is 1000x higher (because it replicates much faster) then not only does it have the potential to infect more cells (and stimulate more systemic immune and inflammatory response) but it is more likely to infect more critical tissues, like the lower respiratory tract. 

Thanks for the detailed response.  "per virion vs cell" or "total infection vs host" gives a lot of wiggle room  - is pathogenicity assigned to one infectious agent or to the eventual multiplicity of them?  One nuclear bomb vs a hundred conventional ones.

I totally agree that the damage per virion could be no worse with this variant, but that it's greater development rate within a host could mean more host level damage.  This certainly fits with the idea it's still going through a rapid host adaption phase.  It's still nowhere near the host level damage of the other nCovs...  (ominous music rumbles in the background)

Jumping back on topic:

The poster oureed has clarified which interpretation they meant: "It's about the seriousness of the disease. The most obvious parameters to look at are hospitalisations and deaths."

So, a bit late for them to claim they were talking about the per-virion pathogenicity...

I don't think changes in fatality rates are so "obvious" when comparing delta with the original virus (March 2020 era, the claim from the other poster), as clinical care has come on dramatically through the lessons learnt and through the repurposing trials - individual prognosis has been improved significantly since the start as a result of improvements to care.   Hospitalisations should be less affected as there's much less change to pre-hospitalisation treatment.  

That's a really important point.  It would probably take a really significantly nastier variant to overcome the better outcome statistics resulting from dexamethasone and other improved treatment, plus better outcomes from second infection following vaccination and/or previous infection.