“Leicester were 5,000-1 to win the league but ended up as champions. What are the odds of that happening?” – Neil Lennon
Les Croupiers Casino, Cardiff, June 1998. For the previous three years I’ve been working as a schoolteacher, but I’m about to head off to music college, so tonight I’m keeping my cash in my pocket: I’m going to need every penny of it. My colleagues Dave and Iwan, by contrast, are on a roll. Come to think of it I never saw them lose, but I suspect they did so in private, or perhaps my memory has become selective.
Either way, for once they’ve decided to quit while they’re ahead. Having cashed in, Iwan finds a £10 chip in his breast pocket – “for emergencies”. Rather than bother the cashier again, he gives it to me and tells me to see what I can do.
I’m a cautious gambler when it comes to my own money, but this is a free hit, so rather than hedge my bets I saunter up to the nearest Roulette wheel and place it boldly on Zero.
Someone else’s money, someone else’s winnings.* It was worth it for the reaction when I returned only a couple of minutes after setting out, carrying £360 worth of chips. What were the odds of that happening? Well, that’s an easy one – assuming a fair wheel, precisely 36 to 1, or a smidgen over 2.7%.
Watching how people decide to grapple with numbers, and how they apply to our everyday lives, is endlessly fascinating, and the reams of statistics being spewed out during the Coronavirus era so far has provided plenty of opportunity for doing just that. What’s the fatality rate of Covid-19, and how might that compare to the chances of dying from other causes, we all want to know? Reading between the lines, I guess we mean, what are my chances of dying from it?
It’s really the wrong question. The probability of me dying eventually from some cause or other is 100%. Beyond that, the application of population-wide statistics and probabilities to individual lives is a tenuous affair at best: it’s simply not what they’re designed for. The half-life of an element can tell you pretty much bang-on what proportion of a large sample of its atoms will have decayed during a certain period. But if you’re sat there looking at any particular atom on its own, you’re back at the Roulette wheel.
Let’s say you’re facing a serious medical operation and the surgeon – as they are wont to do these days – informs you that it has a 99% survival rate. That sounds good, you think. 99 out of every 100 patients make it through. But it’s hardly any consolation if you’re the 1 who doesn’t – in fact, it’s probably even more annoying knowing how improbable your death was.
A newspaper columnist this week, as part of an argument to send children back to school, quoted 0.03% as the likelihood of dying of Covid-19 for under-18s. Let’s for the moment assume that’s roughly right. As an individual parent you might think that sounds fine, a chance worth taking. But the only concrete meaning of that number is that, given 10 million or so children of school age nationwide, it translates to the racing certainty of 3,000 deaths (and that’s without factoring in the health risk to their teachers). So as a national decision-maker, you might well view that percentage quite differently. Grieving parents would hardly be consoled by the reassurance that their child’s death was statistically exceedingly unlikely.
But hang on, I hear you cry – around 15,000 children are killed or injured in road accidents in the UK each year. We don’t re-organise our entire lives around that, do we?
Putting aside the question of whether we should look again at the idea that this is a price worth paying for the freedom to drive our own cars, the suggestion that we make no allowance for road safety in our everyday lives is clearly nonsensical. In fact the layouts of our cities are in huge part devoted to allowing for and containing the risk from human-operated vehicles. Imagine for a moment what a town centre without traffic would look like; or perhaps as a more realistic exercise, visualise what a city without human-operated, individually-owned traffic could look like. No need for car parks or parking spaces, far safer junctions and crossings, cleaner air. If we invented the modern automobile today, I wager no-one would dream of putting one under the fallible, volatile control of anyone who could pass a short test, and allowing them to plough through our midst largely unsupervised.
So we’ve made about as many concessions to road safety as we’re willing to make, and presumably we’ve decided that 15,000 dead or injured children a year is something we’re happy to live with as a result.
Terrorism? The Daily Telegraph recently claimed that “over the last decade, the annual chance of being murdered in a terrorist attack on British soil was about one in 11.4 million per year.” I suppose that gives us an average of six deaths a year, and a significant degree of disruption to our way of life.
What about flu, you ask. An average of around 17,000 people in England and Wales have died from influenza in recent years – it’s worth noting that this has varied greatly from 2,000 to 28,000 (during the particularly deadly season in 2016). Why are we fretting so much more about the Coronavirus?
In fact, if and (let’s all pray) when there’s a vaccine for SARS-CoV-2, the precautions against it may well end up looking very similar to our current regime for flu: free vaccines for the most vulnerable, and available to all, and the NHS geared up for a huge increase in hospital admissions in the event of a bad season. Of course, that might change as our medical scientists gather more data and come to a better understanding of how this new virus operates.
That latter point is crucial, and should underpin every discussion currently taking place on the topic. If you hear anyone speaking with anything resembling absolute certainty regarding any aspect of the current science around Covid-19, please approach with care. The point about all these numbers – overall mortality rates, probability of dying from it, “R0” and so on – is that they require an awful lot of data to establish to any great degree of accuracy, and the gathering of that data in this case is still in its early stages. Plus, they’re still strictly speaking only a measure of what’s happened in the past, not necessarily an accurate prediction of what may come next.** Science is the process of establishing and quantifying the degree to which we don’t know things. The upshot of that is that there is no such thing as a “scientific fact”.
I’ve left air travel till last. There hasn’t been a fatal commercial air accident in the UK since 1989 – so to put it another way, the probability of you dying in a plane crash in this country is currently my lucky Roulette number: zero. The number and complexity of precautions we take when it comes to commercial air travel is immense compared to most other forms of travel.***
Grasping the connection between these two statements is the key to understanding the dilemmas surrounding our current crisis.
* – In fairness to Iwan, he bought us all dinner at Charleston’s afterwards. This was the occasion on which I invented the rare fillet steak with onions, mushrooms, peppercorn sauce and a pineapple slice on top. It didn’t catch on.
** – I’ve heard researchers suggest that one of the effects of SARS-CoV-2 might be to turn every carrier into a zombie after 12 months. Presumably it’s a reminder that as yet they, and consequently we, have little idea what might be around the corner.
*** – This article provides a lot of food for thought on how we respond to air and road accidents.