Here’s a remarkable story that no-one in the media or social media has picked up on (although you can bet that everyone in the field knows).
So we have two epidemiological teams, one at Imperial, and one at Oxford, with differing ideas about Covid-19.
The Imperial team, led by Prof. Neil Ferguson, is the team whose study has led to the UK shutdown (Ferguson is a current member of SAGE, the UK government’s Scientific Advisory Group for Emergencies). They’re the current big guns in this world, and have been for years, ever since the 2001 foot and mouth epidemic when they were led by Professor Roy Anderson, and Ferguson was his protege.
Then there’s the Oxford team, led by Sunetra Gupta. They recently brought out a study [HD: text now below] trashing the Imperial’s study, and claiming that half the country may already have had the virus.
“I am surprised that there has been such unqualified acceptance of the Imperial model,” said Prof Gupta.
There was an immediate attempt to trash the Oxford study in the media (clearly orchestrated by Imperial), for example, here.
Now, do you accept the traditional image of scientists as sober, serious, disinterested seekers of truth? Or do you have more of a Biscuit Factory sort of view of them, where quite a lot of them are very flawed human beings, egotistical shits bent on climbing the greasy pole and treading on people to get to the top? Bullshitters and networkers and operators? Actually, I think the former types do exist, there are good, serious scientists out there (including some of my personal friends, and quite a few readers of this blog), but there are an awful lot of the latter types, especially at the top, and it’s rare to hear of a science department that isn’t full of bitter hatreds and jealousies and vendettas, where every Professor turns into an arsehole no matter how nice they seemed when they were a graduate student.
You may think I’m exaggerating, so let’s take a bit of a closer look at the Imperial-Oxford situation. I’m only going to pick out a few details now, because the full story is very large and I don’t know it all, plus a lot of it is very murky and undocumented.
The Imperial team was originally led by Professor Roy Anderson, leading luminary in the 2001 foot-and-mouth disaster, which is a whole other story that I’m just starting to put together now with the help of a brilliant colleague (and any help on that in the form of recollections and inside knowledge or links would be appreciated). It was Anderson who established Imperial, allegedly in an underhand manner, as the government’s go-to team on communicable disease crises. This was a world rife with intense rivalries.
Anderson had recently come to Imperial from Oxford. Why did he leave Oxford? Turns out it was, allegedly, mainly because of two things. One, he had allegedly not declared to the Wellcome Trust the fact that he was receiving income from a scientific firm, even though he was a Trustee of the Wellcome trust, and a director of a Wellcome Trust Centre.
Secondly, he had allegedly publicly claimed that a woman in the Zoology department was only appointed to a Readership (ie. above Senior Lecturer but below Professor), after her five-year Fellowship ended, because she had slept with the head of department, who was on the appointing committee. There were also allegations that he had been a bit of a bully, but his allegation against this lecturer was the main problem, and it got him suspended for two months. In the end he decided to leave for Imperial, which offered him a very good position. He took many of his team with him, including Neil Ferguson.
I should stress that I have no idea whether any of these allegations, on either side, are true, although I note that the woman won her legal case against Anderson. Yes, it actually went to court, and it was a big deal at Oxford, it wasn’t just a little inter-departmental spat. What I am pointing out is the soap-opera nature of the whole thing. This sort of thing is not at all rare in University science departments (and other departments too), and sometimes it’s worse in the more high-powered ones.
The crowning glory in this story, though, is this. Who was the woman who Anderson allegedly accused of sleeping her way into an Oxford Readership? Her name was … Sunetra Gupta. Who is now the head of the Oxford team engaged in the bitter struggle against the Imperial team that Anderson set up, and which is still run by his protege, Neil Ferguson. You couldn’t make it up. At least, even I wouldn’t have made that up for my novel, it’s just too perfect to sound like real life. But it is real life. Real University life, at least.
The question you should now ask is, if I no longer think that these scientists are all unimpeachable examplars of rectitude, shouldn’t I perhaps be at least a little bit more sceptical of their work? If this is an ego-driven world of power politics, with a lot of glory and funding at stake, and feuds galore (and there are many more stories, especially about Anderson and some of his mates, although I stress that everything in these stories are allegations only), perhaps the shining light of truth isn’t always the end result of the research? So perhaps we should reach for a hefty dose of salt whenever some glamorous set of results is revealed? At least, we should be asking questions like, ‘What reason do we have to think this is true’, other than the appeal to authority? Because the appeal to authority isn’t going to cut it.
Update: If you can’t see the FT article, here’s the text:
——————–
Coronavirus may have infected half of UK population — Oxford study
New epidemiological model shows urgent need for large-scale testing
Clive Cookson, Science Editor
March 24 2020
The new coronavirus may already have infected far more people in the UK than scientists had previously estimated — perhaps as much as half the population — according to modelling by researchers at the University of Oxford.
If the results are confirmed, they imply that fewer than one in a thousand of those infected with Covid-19 become ill enough to need hospital treatment, said Sunetra Gupta, professor of theoretical epidemiology, who led the study. The vast majority develop very mild symptoms or none at all.
However, the modelling by Oxford’s Evolutionary Ecology of Infectious Disease group has been challenged by other scientists. They have pointed out that the study presents possible scenarios — based on assumptions about the nature of the virus, its virulence and its arrival from China — that contradict those supported by most epidemiologists.
The Oxford research suggests that Covid-19 reached the UK by mid-January at the latest and perhaps as early as December. It spread invisibly for a month or more before the first transmissions within the UK were officially recorded at the end of February and the epidemic started to grow exponentially.
“We need immediately to begin large-scale serological surveys — antibody testing — to assess what stage of the epidemic we are in now,” Prof Gupta said.
The research presents a very different view of the epidemic to the modelling at Imperial College London, which has strongly influenced government policy. “I am surprised that there has been such unqualified acceptance of the Imperial model,” said Prof Gupta.
However, she was reluctant to criticise the government for shutting down the country to suppress viral spread, because the accuracy of the Oxford model has not yet been confirmed and, even if it is correct, social distancing will reduce the number of people becoming seriously ill and relieve severe pressure on the NHS during the peak of the epidemic.
The Oxford study is based on what is known as a “susceptibility-infected-recovered model” of Covid-19, built up from case and death reports from the UK and Italy. The researchers made what they regard as the most plausible assumptions about the behaviour of the virus.
The modelling brings back into focus “herd immunity”, the idea that the virus will stop spreading when enough people have become resistant to it because they have already been infected. The government abandoned its unofficial herd immunity strategy — allowing controlled spread of infection — after its scientific advisers said this would swamp the National Health Service with critically ill patients.
But the Oxford results would mean the country had already acquired substantial herd immunity through the unrecognised spread of Covid-19 over more than two months. If the findings are confirmed by testing, then the current restrictions could be removed much sooner than ministers have indicated.
Although some experts have shed doubt on the strength and length of the human immune response to the virus, Prof Gupta said the emerging evidence made her confident that humanity would build up herd immunity against Covid-19.
To provide the necessary evidence, the Oxford group is working with colleagues at the Universities of Cambridge and Kent to start antibody testing on the general population as soon as possible, using specialised “neutralisation assays which provide reliable readout of protective immunity,” Prof Gupta said. They hope to start testing later this week and obtain preliminary results within a few days.
This article has been amended since original publication to clarify the fact that the modelling is controversial and its assumptions have been contested by other scientists.
——————–
60 thoughts on “Soap-opera science”
“and any help on that in the form of recollections and inside knowledge or links would be appreciated)”
No direct knowledge, but I was involved in the clean up of the decomposing carcasses they left as a legacy. They turn to liquid after a few years in the group and it was escaping into places it was not welcome. I clicked early last week it was the same yahoo’s behind the numbers for C-19 as the foot and mouth response. I suspect if you look into the code, you will find an output on how many ‘cattle should be culled’, with cattle crossed out with pen and people written above it!
I am struck by how credible people seem to find the phrase “computer model”, when the reality of it is little more than a whole bunch of random numbers punched into a random number generator that is pre-programs to output a Big Scary Number.
Just to add, great work in tracking all this down. A quote I use quite frequently;
“Science is the process by which we free ourselves from the tyranny of experts”
Not sure where I got it from, but it’s never been more applicable. There is not a lot of science going on, it’s all just opinion from the ‘credentialed’ classes.
I’m fast coming to the conclusion that Neil Ferguson will need to be put up against a wall and shot when this madness finally concludes. He will have been personally responsible for the impoverishment of millions, maybe tens of millions of people, including the deaths of all those who will die as a result, either from diseases caused by poverty, or suicide. Plus the misery caused by mass unemployment and economic depression. The government have acted pretty much on his advice alone and if it proves to one of the most egregious cases of scientific over-reach ever, he needs to pay the price.
The trouble is that if this does prove to be a f*ckup of immense proportions, its so big that no-one in power will be able ever admit it, and the whole thing will be deep sixed, not least because the MSM are balls deep in the whole thing too, and thus will have no incentive to expose their own role in the catastrophe.
…not least because the MSM are balls deep in the whole thing too, and thus will have no incentive to expose their own role in the catastrophe.
They’ll need more clicks and will want somewhere to shift the blame. They’ll be the first to crack. Especially with the pressure from alternative media.
President Trump seems to have a similar problem with experts. He’s now extended restrictions until the end of April. And that means we, too, have at least another month of this.
I think you may have the wrong end of the stick here: it is the Oxford unpublished draft ‘study’ that is in question, not the published Imperial paper.
The Oxford study may be in question by the establishment.
The Imperial study (original and reverse ferret versions) is in question by the outside world
The real world mortality stats from ONS seem to be tending towards backing up the Oxford study.
Obviously the govt will seek to massage the figures in an Italian manner to justify positions taken.
Thankfully such base motivations, rivalries, and disputes over basic science and computer modelling happen in relation to climate change where there is – we are repeatedly told – a consensus among scientists.
I believe that Neil Ferguson has been asked to publish his codes so the modelling can be checked and has refused.
Which sounds to me that it hasn’t been open to outside scrutiny.
He says he will release them in a week or so, but apparently he has had some big teams in to work on them, supposedly to help them be usable for other teams, so I doubt we’ll be seeing the original code, which will probably be a mess.
Still, that will still be good, to let some other people take a look at it.
The evidence where pro-active testing (testing of whole groups and not waiting for people to get ill) suggest fatality rates of about 0.5%. That broadly fits with Imperial’s calculation and their statement that about 1 in 10 to 1 in 20 show significant symptoms (of which about 1-4% die)
This I agree with wholeheartedly
Mike, would appreciate a source for this.
Your first sentence is unacceptable in a serous discussion on the issues
Not at all surprised. When I was an undergraduate scientist in Oxford in the 60s, the book in vogue was “The Struggles of Albert Woods ” written I think in the 50s. It is clear nothing has changed in the intervening half century. What is disturbing is that computer modelling is now regarded as fact and not hypothesis.
Not sure who this is addressed to, but I’m questioning all such work, from both sides.
President Eisenhower foresaw that scientists would willingly prostitute themselves for the Almighty Government Dollar. Remember that Eisenhower knew wereof he spoke — he had been President of Columbia University after WWII.
In Eisenhower’s famous 1961 Farewell Address, he warned about the “Military-Industrial Complex”. Immediately after that warning, he went on to warn about the politicization of science.
“Akin to, and largely responsible for the sweeping changes in our industrial-military posture, has been the technological revolution during recent decades.
In this revolution, research has become central; it also becomes more formalized, complex, and costly. A steadily increasing share is conducted for, by, or at the direction of, the Federal government.
Today, the solitary inventor, tinkering in his shop, has been over shadowed by task forces of scientists in laboratories and testing fields. In the same fashion, the free university, historically the fountainhead of free ideas and scientific discovery, has experienced a revolution in the conduct of research. Partly because of the huge costs involved, a government contract becomes virtually a substitute for intellectual curiosity. For every old blackboard there are now hundreds of new electronic computers.
The prospect of domination of the nation’s scholars by Federal employment, project allocations, and the power of money is ever present and is gravely to be regarded.
Yet, in holding scientific research and discovery in respect, as we should, we must also be alert to the equal and opposite danger that public policy could itself become the captive of a scientific-technological elite.”
Fascinating – why does Neil Ferguson even have the job ? The catastrophic miscalculations during foot and mouth were based on Imperial’s predictions and he also predicted 150,000 would die from BSE in the UK – In fact less than 200 did
We all know there is absolutely no penalty for total failure and rank incompetence in State employment, especially the higher up the pole you go – see Cressida Dick’s career. Lower down there may sometimes be consequences, mainly because the higher ups need a fall guy or girl. But once you reach a certain level you’re untouchable, possibly because by reaching that level you have learned over the years where the bodies are buried so no-one can ever dare throw you under the bus.
Ferguson is shameless; now trying to manoeuvre himself in case the apparent plateau becomes a trend.
The odd thing in this article is that Gupta seems to support the Government’s measures which, according to this article, are decided mainly by Ferguson. Not surprising as the quotation from Gupta is exactly what the government said was its aim in the very first public briefing. Also the Government is buying loads of antibody test kits so that the level of herd immunity can be estimated more accurately. Is Boris taking advice from both? Or are the two teams looking at different aspects and merely indulging in the usual but irrelevant bickering between rivals with big egos and never to be forgiven pasts?
Ferguson ineptitude knows no bounds, I suspect extreme politics are near the surface too.
When referring to test kits, it.would be immensely useful to define what is being tested for, there are loads of references on tinternet to kits, rarely do you see a precise definition.
1) A test for Covid 19 indicating you currently have it.
2) A test for the antibodies indicating you have had it and are now not a carrier / vector.
It would also be good to see some serious information on the accuracy of the tests — rates of False Positives and False negatives.
There have been a few media stories about Chinese-supplied tests giving the wrong result more than 50% of the time. But nothing about the accuracy of Western-supplied tests.
There’s a full description of the statistics on an antibody test developed in NC, USA in this comment with supporting link:
https://wattsupwiththat.com/2020/03/28/the-next-frontier-in-coronavirus-testing-identifying-the-full-scope-of-the-pandemic-not-just-individual-infections/#comment-2949320
Roughly, about 10% false positives and 10% false negatives (the spurious accuracy of quoted percentages on small sample sizes is not good statistics). At present it might be acceptable for testing for who hasn’t had the virus, but it leaves far too many false positives thinking they’re “safe” when they aren’t. Also, it is a poor test of whether you do have the virus while the prevalence in the population is low. You are more likely to be a false positive than a true one absent any other indications.
The contiguous cull advocated by the Imperial College team in order to wipe out FMD (in conjunction with the 3 km cull and Slaughter on Suspicion) was described by Anthony Gibson, then Regional Director of the NFU for the SW, as follows:
“The contiguous cull has been exposed for what it was; one of the most bloody, tragic and disgraceful misjudgements ever committed in the name of science,”.
He was right. The Imperial College policy willingly accepted and effected by the Blair government (Blair had an election in May 2001 to think about) resulted in the mass slaughter of over 11 million animals, the majority of which were healthy (as revealed by test results given in Parliamentary Answers subsequently).
The suffering inflicted on humans and animals was enormous, and the damage to the economy and livelihoods massive, much of it irreversible. 60 farmers committed suicide as a result of the trauma they experienced. Yes, the Imperial College team left an indelible mark on the countryside and farming. It was a shocking legacy, and in my view they (e.g. Neil Ferguson, who was part of Prof Roy Anderson’s team in 2001)should not be anywhere near the levers of power at present, as it appears they have not learnt any lessons.
The latest installment from Imperial College is out today looking at the number of infections across 11 European countries announcing their measures may have averted up to 120,000 deaths across Europe. No mention of Sweden. The range of the infected population their model predicts is quite something – Italy 1.9-15.2 million, Spain 1.8-19 million, Germany 0.25-1.5 million. Model does some fancy maths but basically extrapolates this from the observed deaths.
This article indicates that the Ferguson study is flawed: https://judithcurry.com/2020/03/25/covid-19-updated-data-implies-that-uk-modelling-hugely-overestimates-the-expected-death-rates-from-infection/#more-25884
Here is the twitter feed for NZ’s ‘expert’ who has led the calls for a total shutdown of life in NZ. Note her attacks on other scientists for not being ‘experts’.
Also, pink hair.
https://twitter.com/SiouxsieW?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1244712486181523457&ref_url=https%3A%2F%2Fyournz.org%2F2020%2F03%2F31%2Fswedens-different-covid-strategy%2F
In the latter part of the 1960’s into the early 1970’s the acronym GIGO shifted from Garbage In Garbage Out to Garbage In GOSPEL Out. Even then people were enamoured of the omnipotent, omniscient, “computer”. I think Alan Turing would be horrified!
I too would be interested in the coding used for these studies and I am very suspicious of any reluctance to reveal it, which is the case in many disciplines using computer models. I wonder whether there is nothing really secretive other than the fact that the analyst is using the standard modules and packages developed for R, Python, and other such programming systems. Such packages are easy to use and can generate forecasts from data collections. The packages are the same routines (algorithms) used in machine learning and so-called artificial intelligence programs. Of course, revealing that this is the basis for the projections would remove the mystique the analysts have built around themselves. It might even “damage” their reputations!
The difficulty analysts face, however, is that not only are the data collections grossly incomplete, but the data is also sparse. Which implies that any analyses based on normal statistical methods must be stated with cautionary caveats—caveats which appear to be left out in many cases such that the results, conclusions, and predictions appear to be accomplished fact.
There is no doubt in my mind that Gupta’s group are right with regard to the state of the data. Whether her theory about the arrival and distribution of the virus in the UK, or elsewhere, is valid, I do not know. It is plausible. No one will know until more extensive testing and investigation improves the data.
>I wonder whether there is nothing really secretive other than the fact that the analyst is using the standard modules and packages developed for R, Python, and other such programming systems
Possibly not, because Ferguson says the basis of it was written 13 years ago in undocumented C. I suspect he’s worried about how shoddy it all is, and full of fudges. Hence him bringing in teams from big companies to clean it up. Others reckon that he hasn’t a clue what’s in there because it’s all written by grad students and contract researchers.
>the acronym GIGO shifted from Garbage In Garbage Out to Garbage In GOSPEL Out
Good one!
Pity Tim Newman is now taken.
GIGO won first prize in a lapel badge competition for the annual meeting of what was the Australian Computer Society in 1972!
Too much of the bases for past prognostications have been lost or are undocumented. Some of it will be simply carelessness or broken good intentions to formally record the investigations.
But I am sufficiently cynical to suggest that some losses may have been deliberate.
Ha! I do believe Tim has slightly higher standards. Slightly.
Does anyone else remember ‘Harry Read Me’ from Climategate?
A sad tale of an IT pro trying to make sense (and failing) of all the undocumented, untested, unscrutinised crap the ‘climate scientitsts’ had flung together to publish their papers..
‘But this time its different’..say the Imperial scientists…..
Before my retirement I ran a specialist software house, some of our work involved modelling factory process flows etc (ie using thoroughly known parameters). I managed to get my hands on the code for the NASA GISS climate model in use at the time and I, plus another of my colleagues (we were the only two in the place old enough to have programmed in Fortran) had a good look at it. Suffice it to say that had one of my junior programmers written code that bad he would have found himself on a short waiting-list for a P45.
I’ve had absolutely no faith in “academic” computer modelling since that time. You only need to examine the UK predictions v reality of the last few pandemics:
Swine flu: 31,000 to 65,000 predicted, 138 actual.
Bird flu: up to 75,000 predicted, zero actual.
SARS: “thousands” predicted, actually zero deaths from only four cases.
BSE: 100,000+ predicted, actually 178 deaths from CJD.
As I see it, any form of predictive study is going to fail because of the uniformly low grade of the data being collected.
I remember it well. I howled with laughter.
Baron J: “As I see it, any form of predictive study is going to fail because of the uniformly low grade of the data being collected.”
Yes, but … there is a but — there is a high probability that official statistics will show a significant number of deaths from Covid-19, albeit less than the models predicted (thanks, of course, to the politicians’ impositions on the people).
On average, more than 1,500 people die every day in the UK. Did the individual die from heart disease, or flu, or lung cancer? Or was it the infernal virus? There is enough latitude to ensure there will be sufficient deaths ascribed to the virus to let politicians claim victory over the foe.
The real number to watch is the total number of people who die from all causes, and compare that to prior years. Because that number is more difficult for the Powers That Be to fudge.
Not to me it isn’t.
I’m sensing some nominative determinism at work.
Interesting articles, nice to hear some reasonable reasoning!
It might be worth looking at some of the fully-tested populations: Cruise ships.
As I understand it Princess Diamond had a population of 4,000 people, two weeks of good mixing in a small space, and an infected proportion of about 1/5. I’d imagine that UK’s would be higher (longer to mix).
Quick plug: I wrote about about this and death rates and stuff here https://www.linkedin.com/pulse/one-thing-you-should-know-coronavirus-statistics-martin-hill/
ferg needs to get his ducks in a row right now.
hockey stick hooker better go to Antarctica this summer if turns out to be wrong)
Not a prob. leading by example, especially when given stats and data.)
20 days head mask or none all will have been filtered. the meme from the mein. anyhow, its on rec as a response. kindly sent, sok.
He said that the model is thousands of lines of undocumented C. As an ex-ITer my first thought was to wonder whether he was telling a sick joke. Undocumented code could just be the tip of the iceberg and makes me wonder if it was tested and validated properly. Is this typical typical of academia’s standards?!
Hector, Thank you for this piece
I was looking at a Lancet paper on gene sequencing of Covid-19 (29 Jan 2020) and I checked the credentials of the listed academic authors and stumbled across (well it was info top page left of the search for the guy as per usual Chrome Google searches) an odd Google snippet that turned up in the first search iteration about one of the only non-Chinese contributor to the paper, Edward C Holmes of University of Sydney.
wtf?
– *more* soap opera science ?
elsewhere fwiw – a credible tale about covid-19 “patient zero”
Thanks tomo, will check out tomorrow, got to get some belated shut-eye now. (That sort of paper I would give zero credence to start with.)
Wading through that paper and simultaneously updating my atrophied applied biology knowledge (30 years ago…) it’s little more than an assay write-up with added conjecture…
Quite a targeted bit of sniping though – Dr Holmes likely ain’t too happy about that 🙂 Shades of “A Very Peculiar Practice”
Hence, the science is never ‘settled’.
Tell that to Greta Thunberg with her supposed superpower of black & white thinking who thinks climate science is settled!
Are such massive number ranges from their models any more useful than a layman saying there’ll be a lot of infections?
Ferguson will be bunged a knighthood to keep him quiet…
It almost certainly was written by a PhD student or RA. And it will be a mess; ‘undocumented’ code should NEVER be used for any serious work, even if a PhD student can get away with it for their thesis. Another way to interpret ‘undocumented’ is that there was no over-sight, or the supervisor was also inexperienced or poor at scientific coding…
As a zoology undergrad & postgrad in Oxford from 92-99 I can confirm that Anderson was a first-class shit. Dick Southwood had created a world class group of ecologists and evolutionary biologists in the department. Giants of the field. Bob May, Paul Harvey, Bill Hamilton, many others. Being taught there was simply extraordinary. Once Anderson arrived his main priorities were to bring as many of his cronies with him as possible (from Imperial, where he ended up going back to after he blotted his copybook) and to de-fund or force out as many people as he could that weren’t going to go along with him. The atmosphere became completely poisonous. I imagine he was hired because of his connections and his ability to schmooze and network, which are useful in winning funding. But he was hated by 70% of the department. There was graffiti about him in the toilets – neatly typed and printed out, stuck up with blu-tack. Sadly, academia is rife with people like Anderson – both men and women. It is said that academic fights are so vicious because the stakes are so low. In my experience, there are just some awful, horrible people at the pinnacles of our academic institutions. In the main, the nastiest, most spiteful people I have met were academics.
Comments are closed.