Cricket's darkest statistics, and why assumptions about suicide might be wrong
By Nick Harris
SJA Internet Sports Writer of the Year
15 November 2011
The untimely death at the weekend of the former cricketer and writer Peter Roebuck, who took his own life by jumping from the sixth floor of a Cape Town hotel on Saturday night, has again prompted debate about the apparently high prevalence of suicide among men who’ve played the game professionally.
The operative word is ‘apparently’ - but we’ll come back to that.
Roebuck (left), 55, has been widely hailed as one of his sport’s finest scribes. For anyone unfamiliar with his unconventional life and times, it's worth a look at some of the myriad pieces published in recent days and linked here in The Guardian, The Telegraph, The Independent and the Daily Mail.
This extract from a piece by Greg Baum in The Sydney Morning Herald (linked in full here) gives a small flavour of the man if not the player.
“He was complex, intense, taut, edgy, opinionated, a little manic, mostly cheerful, sometimes broody. He was a contrarian, not for the sake of it, but because he always had another view. He spoke quickly, in a clipped tone, needing to get the thoughts out so that more could follow; his broadcast voice was his street voice. He did not do small talk, ever.
“Cricket was his metier, but it did not confine him. He was widely read and supremely intelligent. He was also self-possessed, yet drew people to him. Women liked him, but often he was awkward in their company ....”
Many of the obituaries and reflections hinted at least that suicide was not an unexpected outcome for Roebuck, that somehow there was always a darkness in the soul.
We will never know exactly why he jumped, what drove him to it on that night and none before, although his state of despair in his final hours, arising from a police investigation of alleged sexual assault, has been described.
But can cricket really play any role in a man deciding to end his life?
Certainly there is a theory that it does, one explored in two books by David Frith, the journalist, author, historian, former editor of The Cricketer magazine and founding editor of Wisden Cricket Monthly.
Frifth, 74, was born in London but spent his formative years between the ages of 12 and 27 in Sydney. He still speaks with an Australian twang, and I talked to him on Monday about his books on cricketing suicides.
The first, ‘By His Own Hand’, was published in 1991. The second, an updated, expanded edition of the same material, and written in partnership with the former England cricket captain turned psychoanalyst Mike Brearley, was published in 2001 as ‘Silence of the Heart: Cricket Suicides’.
The first book chronicles the lives and deaths of 80 cricketers who killed themselves, or are believed to have killed themselves.
By the second book, Frith had amassed details of 151 cricketing suicides, among them 23 Test players. Of those 23 players, six were men who had played Test cricket for England.
Frith tells me that he set out with the intention of trying to ‘get cricket off the hook’ for this appalling loss of life; that he wanted, through his research and the exploration of individual sportsman’s lives, to show that actually, and possibly always, they killed themselves for non-cricketing reasons.
Some of the 151 had long-standing depressive illnesses and mental health issues. Others had been left devastated by marital break-ups, financial problems, terminal illness or other incapacity. Many of them took their lives in the gloaming years when the thrill of competition was gone and middle age was setting in.
Over time, Frith tells me, he has been left in no doubt that something about the life of a cricketer affects some men. “It’s a creepy game,” he says. “And I should know because I played it for 50 years. And the statistics are there that show there is an issue.”
This is where, for the time being, that my own analysis diverges from Frith, a world expert in this field. In fact the world expert.
Ironically, given cricket’s extensive usage of statistics, it is statistics that I believe fail to make the case for a definitively stark link between suicide and cricket.
To my knowledge - and to Frith’s knowledge - there has never been any large-scale, statistically relevant study of the relationship between professional cricket and suicide, or indeed suicide and any one sport.
Frith’s own work is the most extensive in the field, and as he himself readily concedes, “it would take 20 of me” to compile the large amount of information to reach firm conclusions.
That didn't prevent a headline-grabbing statistic, arising from Frifth’s 2001 book, from making in into the national and international media. It was the claim that English cricketers are ‘almost twice as likely to commit suicide as the average male’.
This was based on two statistical 'facts':
That suicide accounted for 1.07 per cent of British male deaths.
That the rate among English cricketers is 1.77 per cent.
There are problems with both these claims.
Firstly the 1.07 per cent (based on 1998 UK data) was wrong, for reasons that Frith is not sure, but perhaps because inaccurate information was provided in the first place.
The Office for National Statistics website has a free database that carries all the relevant information, much of it available to download in searchable spreadsheets. And a search today showed there were 264,707 male deaths in the UK in 1998, of which 4,039 were suicides, which equates to a 1.53 per cent male suicide rate in the UK in 1998, not 1.07 per cent.
Secondly, that 1.77 per cent rate of suicides among English Test cricketers was arrived at by looking at 339 English Test cricketers - across all time - who had died by July 2000. Six of those 339 (or 1.77 per cent) had killed themselves.
To base any fixed conclusion on such a tiny sample pool of a few hundred people spread over more than a century is risky. The data pool is just too small to be statistically reliable.
Then consider the individual cases of the six men who killed themselves. They were:
William Scotton (died 1893, age 37, after suffering depression).
Andrew Stoddart (died 1893, age 52, after falling into bad health and debt. He shot himself).
Arthur Shrewsbury (died 1903, age 47, after shooting himself in the mistaken belief he was terminally ill).
Albert Relf (died 1937, age 62, following depression).
Harold Gimblett (died 1978, aged 63, who overdosed on prescription drugs after years of mental health problems)
David Bairstow (died 1998, aged 46, hanging himself a few weeks after surviving an overdose. He’d suffered from depression. The coroner actually returned an open verdict, not convinced of true intent of suicide above a ‘cry for help’).
The cases of these six tragic men, three from Victorian times, but all with depressive histories and / or health and financial worries in their post-playing days, were the basis of the ‘English cricketers twice as prone to suicide’ claim.
In fact, had there been five of them and not six, that would have been 1.47 per cent of the 339-man sample pool, and hence a lower figure than real 1998 British rate of 1.53 per cent of male deaths by suicide. English cricketers might easily have been presented as less susceptible to suicide than the typical person.
This is not to disparage Frith’s work in any way because the incidence of suicide by players of other Test nations is apparently high: a startling 4.12 per cent in South Africa, 3.92 per cent in New Zealand and 2.75 per cent in Australia.
But again, the sample pools are small, statistically insignificantly small, and most of the cases, like those involving English cricketers over vast stretches of history, are not obviously connected to the business of playing cricket.
That South Africa figure is based on seven deaths, the Australian figure on five deaths and the NZ figure on two, one of them by somebody with terminal cancer. It should also be noted there has never been a confirmed suicide of any Indian, Pakistani, Sri Lankan or West Indies Test cricketer, of any era, which might lead to the assumption that Test cricketers in those countries are wholly immune from the phenomenon. That wouldn't be a statistically valid conclusion any more than saying English cricketers are twice as likely to die at their own hands as 'normal' English people.
Somewhere in the region of 2,700 people have ever played Test cricket. The 23 suicides identified by Frith by 2001 equate to a rate of 0.85 per cent of that total. To be necessarily macabre to illustrate a point, there need to be almost as many suicides again amongst living Test players (at any stage of their lives) for the universal Test player suicide rate to become even slightly 'abnormal'.
And that 1.77 per cent Frith calculated in English cricketers has already fallen since 2001 - because there have been a number of deaths of English Test players but no suicides since Bairstow.
Professor John Aggleton is a specialist in behavioural neuroscience at Cardiff University’s School of Psychology. In 1994 he co-authored a paper (linked here) on the differing life spans of right- and left-handed people, and used cricketers as his research pool because there was a clear record of handedness for the individuals.
The study involved a Who’s Who (literally) of first-class cricketers, 5,479 of them, of whom 3,165 had died.
Unfortunately, that data pool is too unreliable in terms of causes of death to assess the significance of suicide in cricketers, but Prof Aggleton did offer a theory as to why a perception of high suicide rates in cricket could endure even without any significant data to back it up.
“In psychology there is a term known as the availability heuristic,” he told me. “This is where something seems common because an example comes to mind readily, rather than because it is actually common.”
In other words, the death of a well-liked public figure such as Bairstow in 1998, and of a well-known writer such as Roebuck this past weekend provokes a link in the public consciousness between cricket and suicide - but it might well be illusory. And when we hear about the troubles of cricketers from Marcus Trescothick to Michael Yardy, it is all too easy to assume there is some oddly irregular relationship between their sport and mental health.
But perhaps there isn’t. Perhaps depression and its most dire consequences aren't any more oddly common or uncommon in cricket than anywhere else. Just because we're not hearing that Fred Blogs from the car plant or Tom Brown from the local solicitor's office is off work with long-term depression and we're not seeing news of Bob Smith the local butcher's suicide in the paper doesn't mean they're not happening.
There might even be a case for saying that it is counter-productive - perhaps even self-fulfilling - to draw conclusions on cricket and suicide on partial evidence.
Does cricket have any more significant link to suicide than football?
German goalkeeper Robert Enke killed himself in 2009, the year after English goalkeeper Tim Carter killed himself and the year before Dale Roberts, a goalkeeper with Rushden & Diamonds, took his own life. Three goalkeepers in three years: how freaky is that? Probably not freaky at all. Justin Fashanu killed himself, as did Dave Clement and Alan Davies. And the list could go on.
Leary: died in 1988
Actually Bairstow also played football (with Bradford City) as well as first-class cricket, as did Stuart Leary (pictured left), a 1950s and 1960s hero with Charlton Athletic and Kent CCC who threw himself off a cable car on Table Mountain in 1988.
If this grim casualty list of sportsmen from just two sports tells us anything (and it's a partial list), then it’s likely that death at one’s own hand is not necessarily as uncommon as some might think - in any walk of life. And we just do not know, frankly, whether there is a statistically valid link between any sport and suicide, let alone why.
No major academic study has been done, although there is a dedicated Centre for Suicide Research at the University of Oxford.
Extensive studies there, for example, have found that the profession with the highest suicide rate in the UK is vets - who are four times as likely to kill themselves as the norm. And we know some contributory factors: stress from training onwards, long hours, high psychological demands, access to lethal drugs and knowledge how to administer them, a philosophical acceptance of euthanasia as a way to alleviate suffering, and exposure to suicides among colleagues that may result in 'suicide contagion'.
The medical profession in general has a suicide rate at twice the norm for some similar reasons, while rates are also high among farmers for different and carefully identified reasons.
As for those who play games for a living? We can only guess. But we don't know for certain that cricketers are any more likely to kill themselves than anyone else, let alone why.
Roebuck never saw himself becoming a statistic that would continue the debate, even thought he actually wrote the foreword to Frith’s first book, in 1991.
He concluded a review of the same title with the line: "Be gentle with yourselves, my friends, and do not expect more of life than it can give."
.