Sir David Spiegelhalter: ‘Risk is a very loaded term’


From the point of view of health risks, the raw oyster that Sir David Spiegelhalter is in the process of releasing from its shell and preparing to slide into his mouth probably wasn’t the safest bet. But from the perspective of sheer pleasure, after a year of restrictions and restraint, it feels like the perfect choice. He chews it a few times — the correct way to eat an oyster — and slurps it down, beaming. “Oh, lovely!”

Besides, his nickname might be “Professor Risk” but if there’s one thing the 67-year-old Spiegelhalter wishes he’d done more of over the years, it’s throwing caution to the wind. “My one regret in my life is that I haven’t taken enough risks,” he tells me wistfully. “I’ve been too cautious — in my career, in my travels. I wish I’d done far more adventurous things.”

Any excess of caution hasn’t stopped Spiegelhalter from reaching the very top of his field. In a year in which making sense of the numbers has become a life-and-death matter, Spiegelhalter, one of Britain’s leading statisticians, has stood out as a calm voice of authority. In the early days of the pandemic, he analysed the data to show how much of a risk Covid-19 posed to different age groups; later, he spoke presciently about why flu deaths would be virtually non-existent this past winter.

Spiegelhalter, whose formal title is chair of Cambridge’s Winton Centre for Risk and Evidence Communication, has also been a vocal critic of the government’s response to the pandemic. He made headlines last May when he blasted the daily presentations of Covid-19 statistics, in particular the daily test totals, as “number theatre”. “They were reeling out lots of big numbers which I knew were desperately unreliable, and it gave a spurious sense of precision and importance to these very flaky numbers,” he says. “It was an appalling lost opportunity, because there was a public hungry for proper detail, who were sacrificing so much, and yet they were getting fed this stuff.”

These days, like much of the rest of the UK it seems, he has slightly more positive things to say, thanks to what he calls the “massive success” of Britain’s vaccine programme and the rapid response to the B.1.1.7 Kent variant. He praises the fact that official briefings are now often given not by politicians but by scientists alone, allowing for better communication of information and statistics — such as on the risks of complications from the AstraZeneca vaccine, which the Winton Centre helped with. It is “too soon to tell” how we have compared against other countries as a whole, however, and Spiegelhalter still has some major gripes with the overall response, which he says comes down to “a lack of scientific understanding at high levels”.

“My real beef, my real anger, is that hugely expensive programmes like test and trace did not have built in, right from the beginning, a capacity for experimental evaluation,” he says. “They should have been running studies, different methodologies, different ways of tracing. As it is, we don’t know what benefit it’s had . . . It’s like rolling out vaccines without ever measuring whether they have an effect or not. For heaven’s sake!”


We are having a rather eccentric seafood feast on the tidal island of Mersea, in Essex. Spiegelhalter has brought along a picnic table, deckchairs, cutlery, crockery and glasses, a striped tablecloth, patchwork cushions and even some flowers from his wife’s allotment, which he’s stuck in a glass bottle with some water. So it all feels rather civilised, despite our scrappy environs — bits of fishing equipment strewn about the place, the loud squawking of seagulls, and a strong, fresh seaweedy smell. The spot we have picked is in the West Mersea Marine boatyard, between a pair of angling boats, which we hope will protect us from a rising wind and falling temperatures. “It’s got a sort of post-industrial chic,” he says. “I love it, I love boats.”

Menu

The Company Shed 
129 Coast Rd, West Mersea, Colchester CO5 8PA

Mersea oysters x8 £8
Spicy lobster soup x2 £10
Dressed crab platter £16.50
Salt & pepper chilli squid £7
Scallops £12
Sautéed potatoes £2
Mixed leaf salad £1
Total: £56.50

Waitrose 

Meursault Les Vireuils Domaine Prunier, 2018 £42

The Deli Downstairs
211 Victoria Park Rd, Hackney, London E9 7JN

Sourdough baguette £2.50
Demi-sel butter £4.10
Total: £6.60

Yabba-Dabba-Dough!
8 Yorick Rd, West Mersea, Colchester CO5 8HT

Coffee x2 £5

The boatyard is next to The Company Shed, which is currently just open for takeaway though it is usually full-service. To call it a restaurant, even then, would be a stretch — it is very much just a shed, as the name suggests, which just so happens to serve very good seafood.

After the first oyster, we move on to the spicy lobster soup — although we had a few minutes of sunshine when we first arrived, the food is now cooling rapidly. The soup is good, though rather heavy on the crème fraîche and light on lobster. Spiegelhalter has a swig of Meursault and tears off some of the sourdough baguette I’ve brought along from my local deli in Hackney. “I’ve been a bit indulgent in lockdown, though I haven’t put on much weight,” he tells me, spreading a thick layer of butter on the bread and folding it over on to itself.

For all his criticisms of the government during the pandemic, there are some theoretical positives that Spiegelhalter thinks could come out of all this — one being a greater level of data literacy among the public. He wants people to see numbers and statistics “as arguments that can change our emotions”, rather than simply as cold, hard facts. “They can be manipulated, chosen and framed . . . to support whatever arguments the communicator wants to make,” he says. “One has to have that critical ability, which involves examining one’s feelings and examining motivations, in order to understand numbers . . . It’s so easy to come to the wrong conclusions, or any conclusions.”

He gazes out over the estuary. “Sometimes you can’t come to a conclusion.”


Spiegelhalter was brought up in north Devon in what he describes as a lower-middle-class family, with little money — his father was an estate agent and his mother worked in a factory. Their family holidays consisted of the four of them — him, his older brother and their parents — squashing themselves into their Mini, strapping some camping gear to the roof and driving off to Europe for two weeks.

After grammar school, he was accepted into Oxford to study mathematics, but by the second year, pure maths had become too difficult. It was then that he moved on to statistics, and became enamoured of the subject under the tutelage of Sir Adrian Smith, now president of the Royal Society and chief executive of the Alan Turing Institute, Britain’s national centre for data science and artificial intelligence.

There were some downsides, though. “It’s not a great chat-up line, saying you’re a statistician,” he smiles, raising his thick black eyebrows under the woolly hat he has now put on along with a cosy-looking lined waterproof jacket. “But what it did was it made me think of ways of making it interesting. And you could say that’s what I’ve done ever since, and that’s my career now — making statistics interesting.”

His services to statistics have won him a knighthood, election to the presidency of the Royal Statistical Society in 2017-18, and a fellowship of the Royal Society. The last of these is his proudest achievement, as it was given to him by his peers. “I never thought I’d get that because I’m not — I know this sounds silly — but I’m not very clever.”

Spiegelhalter’s focus on risk was the result of a successful application to take the job as the first Winton Professor of the Public Understanding of Risk at Cambridge, a position he held between 2007 and 2018. But the word “risk” is actually one he shies away from. “It’s a very loaded term,” he says. “The problem with ‘risk’ is that it only addresses the downside — you say there’s a good chance of winning the lottery or of something good happening . . . I much prefer thinking in terms of potential benefits and harms, which is clumsier but really expresses what we’re faced with in every decision that we make.”

Spiegelhalter first made a name for himself in the academic world in the 1980s, when he developed ways for AI to handle uncertainty. Some of his most influential work was building algorithms for diagnosing congenital heart disease, in which he broke new ground by enabling probability theory to be used in complex computer models. But although he recognises the growing importance of AI in all parts of our lives, he is sceptical about anyone who claims it’s a quick fix for a problem. “It’s almost always a sign that people don’t know what they’re talking about,” he says. “It’s essential that they are held up to scrutiny. That means scrutinising carefully what the algorithm is saying, and also what is being said about the algorithm. Both need to be trusted.”

At a time of heightened anxiety, when people have longed to be given certainties and assurances, Spiegelhalter has had the difficult task of communicating the idea that everything is uncertain, and that “following the science” actually means very little. But he believes strongly in the importance of doing this. “What this whole thing has displayed is . . . firstly the need to acknowledge uncertainty, and secondly to admit when you’re wrong. And I try to do that, but it is difficult.” He grabs a handful of the salted crisps he has brought along.

I ask him if he thinks he has been wrong over the past year. “Oh yes, yes.” I pass him the scallops, which are served with thyme butter and bacon.

So what did he get wrong? “My whole attitude from the beginning was to play down the severity . . . I wasn’t arguing against [the lockdown] but I wasn’t arguing for it,” he says. “[Former government adviser Professor] Neil Ferguson was talking about 20,000 deaths and I remember saying: ‘Oh yeah, that would seem like a reasonable figure — like a bad flu season.’ So I was hopelessly wrong on that.”

Line chart of Weekly deaths in England & Wales ('000s) showing Understanding the data: why is there a 'death deficit'?

He takes another sip of wine. “My huge built-in optimism makes me unsuitable as a leader of men. Or women. Or anyone.” He grabs a sautéed potato and dunks it into the mayonnaise.

So given these admissions, and his embracing of uncertainty, does he have any time for the “lockdown sceptics”? Not really. “There’s always room for a range of informed scientific opinion — that’s admirable and I feel sometimes there hasn’t been enough of . . . a counterargument against the main thinking. But that lockdown sceptic group acquired some people without a concern for serious evidence, who clearly enjoyed their celebrity and who I think have been an actual danger . . . I’ve lost respect, essentially, for that group. Which is a shame because very often a sceptical view — there’s often something in that.”


Although he appears to be incredibly busy, Spiegelhalter formally retired at the end of 2018. His current role at the Winton Centre is an unpaid one, but he does still earn money from his book The Art of Statistics, published in paperback in February 2020. He says a newfound desire to understand data, amid the vast amount of statistics we have all had to contend with over the course of the pandemic, has been a key factor in the book’s success. “I’m earning slightly embarrassing amounts of money from it. I’m trying to give it away but I don’t seem to be able to give it away fast enough,” he says, laughing gently.

The tide has come right up, and the wreck of a boat that Spiegelhalter has been delighting over is now almost completely submerged. I later discover it is the remains of the Nanaloa, a military vessel used for air-sea rescue operations in the second world war. My hands have become so numb that I have had to put my thick, padded ski gloves on. I suggest this might be the point to find a coffee.

We pack up our things, and set off along the coastal road in search of a decent cup of something dark and hot. What we find is a place called Yabba-Dabba-Dough!. I ask for a decaf cappuccino and make do with a cup of caffeinated hot milk. Spiegelhalter orders an Americano, and seems to be given an Americano.

Despite his unerring optimism, Spiegelhalter has suffered huge personal tragedy in his life. In 1997, he and his wife lost their five-year-old son Dan to cancer. When I ask him about this, as we get our coffees, his tone is sombre for the first time. “That has changed me really, and has made me a strong advocate for the idea that the quality of someone’s death is hugely important, both for them and their family. And that’s why the deaths that so many people have had during this pandemic, I think, are particularly sad.”

I ask him whether the loss of his son changed his outlook on taking risks, but he says what it really taught him was the importance of luck. This was a cancer that was diagnosed when Dan was just 11 months old and, despite all the best treatment, could not be beaten. “Whatever you want to call it — unavoidable unpredictability, chance, fate . . . There are certain things that just happen . . . We shouldn’t overestimate the amount of control we have.”

By the time we get back to the boatyard, we’ve been talking for four-and-a-half hours straight and I think we’re both exhausted. As I drive away, I see Spiegelhalter is horizontal in his car — he’s put down his front seat to have a snooze. On his way home, he’ll be having a brief stop on the east side of the island to peep at a “mortsafe”, an iron cage that was sometimes placed over graves during the 19th century to guard against body-snatchers. I wonder about just how high the risks actually were of having your body snatched, and how you could work such a thing out back then with no superstar statistician around to guide you.

Jemima Kelly is an FT columnist and Alphaville reporter

Data visualisation by Ian Bott and Keith Fray

Follow @FTLifeArts on Twitter to find out about our latest stories first





Source link

Leave a Reply

Your email address will not be published. Required fields are marked *