Storky and the perils of “survivorship bias”


    A tiny triumph, the colour of sand and driftwood, took off from Onetangi Beach a few days ago when no one was watching. For the first time in living memory a New Zealand dotterel/tūturiwhatu, appears to have fledged on Onetangi Beach, one of the most popular dog-walking beaches on Waiheke. Her survival seems even more miraculous given a tropical cyclone, raging tides and record-breaking rain. 

    How did she do it? 

    This summer was a bleak one for New Zealand dotterels on Waiheke. Predators dined on five out of six dotterel chicks at Whakanewha Regional Park, and only two chicks at Church Bay survived long enough to fledge. The nest on a field at Waiheke High School was also unsuccessful. 

    As dotterel parents have done for millennia, the mother and father of the Onetangi chick, soon nicknamed ‘Storky’, distracted predators by cheeping and running to distract them. The two birds were joined by a volunteer human army who used Facebook to share information, with daily posts, video and photographs by Anne Woodley, the Waiheke Dotterel Guardian Group and others. Onetangi residents helped guard the tiny ball of fluff from dogs and cats, and the dunes they planted provided crucial shelter. Islanders began to follow the chick’s cute daily antics online and worry about her survival. Storky was saved not only by the ancient instincts of her parents, but by modern technology, including social media.  

    Her survival was always going to be chancy. Although some people apparently think New Zealand dotterels are common, the birds they are able to see have survived thanks to intensive human intervention. New Zealand dotterels are endangered, and rarer than many species of kiwi. In areas where dotterels are unmanaged, their breeding success is low. 

    Thinking that New Zealand dotterels are common might be a classic example of “survivorship bias” – focusing on visible success and forgetting about hidden failure. And failure, for a dotterel, can be a bloody affair.   

    What do perceptions about Covid-19 have to do with dotterels? Both have something to teach us about misplaced optimism and survivorship bias. I often hear people say that the New Zealand government overreacted to Covid-19. After all, they say, “I don’t know anyone who died of it. Was it really that bad?” 

    New Zealand has low Covid mortality rates among high-income nations, but, like Storky’s survival, it could very easily have been different.

    (For a sense of just how different, watch This England, a British television docudrama about the first wave of the pandemic in the UK, based on the testimonies of staff at Number 10).

    Thanks to public health measures, when Omicron exploded in New Zealand it found a largely vaccinated population, and hospitals with treatments which had already been trialed overseas. “These measures,” wrote epidemiologists Michael Baker, Amanda Kvalsvig and Matire Harwood in The Conversation last month, “decreased the case fatality risk from about one in a hundred during the first two years to less than one in a thousand now.”

    In New Zealand, thousands of lives were saved by effective government public health interventions. Maybe yours. Maybe mine. Astonishingly, New Zealand had fewer people dying in the first two years of the pandemic – of any cause – than expected based on deaths in previous years. 

    Covid, of course, doesn’t go away just because it is no longer dominating the headlines. On 10 February, 8396 new (known) cases were reported, 171 people were in hospital, and 32 had died in a single week; more than in all of 2020 or 2021.

    The most famous example of survivorship bias comes from World War II, when US statistician Abraham Wald was asked to protect bombers from enemy fire using data about damage in returning planes. The US military wanted to reinforce the planes with armour plating in sections which had received the most hits. A no brainer, right? Wald’s insight was to see that this ignored the failures – the planes that had never returned. Where were all the planes with bullet holes in the middle of each wing and fuselage, the nose, propellers and cockpit? Wald told the army to armourplate those areas, because if a plane was hit there, it had crashed, removing itself from his data set. Like Covid minimisers and complacent dotterel optimists, the US military didn’t have the full picture – and hadn’t realised it.  

    When we think about helping next summer’s dotterel chicks, we need to remember all the Storkys who never made it – and figure out why. • Jenny Nicholls

    Subscribe and read Gulf News and Waiheke Weekender Online