Tuesday, January 27, 2009

Shark Attacks and Relative Risk

In another example of how our village primate brains don't process data from large sets well, witness the reaction to three recent shark attacks in Australia:

SYDNEY (AFP) – A spate of savage shark attacks in Australia has sent a shiver through summer holidaymakers bombarded with graphic details and claims that the razor-toothed predators are increasingly targeting humans.

Three attacks on swimmers within 24 hours over Sunday and Monday -- just two weeks after a snorkeller was killed -- have fuelled a fevered debate over whether overfishing has put man on the menu.

"Humans are next in line on the food chain," veteran shark hunter Vic Hislop told commercial radio. "It will definitely get worse."


It is always handy to remember when interpreting one's emotional reactions to such things that our brains did most of their evolution in a very different world from our modern one, particularly with regard to population. Our brains developed in an environment where we'd only into contact with a few hundred people at most. We are not geared to digest information on millions of people. Thus, we tend to overstate the risks of very low-probability events, especially the ones that inflame our most basic fears (from an evolutionary perspective, the value of an abject fear of spiders, snakes and sharks is clear). So, a death rate that rounds to zero causes a primal reaction in us, as if the cause posed a real increase in risk. The cause of the death of three people in a band of 25 would.

Luckily, we humans haven't been slacking these last 11,000 years, and we've developed scientific and statistical techniques to improve on our gut reactions.

Experts say there is no scientific evidence to support his claim that reducing the shark's natural prey through overfishing has produced a spike in attacks.

Three attacks in 24 hours might be unusual, but John West, curator of the official Australian shark Attack File held at Sydney's Taronga Zoo, dismisses claims that the number of attacks on humans is increasing.

"The human population is increasing and more and more people are going into the water, but there has not been a corresponding spike in fatalities from shark attacks," he told AFP.

"There is still an average of 1.2 fatalities a year over about the past 50 years -- if anything the fatality rate for shark attacks is dropping in comparison to the increase in the human population.

"Humans are not part of the shark's diet, otherwise there would be nobody safe in the water."

A total of 194 deaths through shark attacks have been recorded in Australia over the past two centuries, leading researchers to point out endlessly that more people die from bee stings and lightning strikes.

But there is something about being eaten that resonates with swimmers.


Damn skippy there is. I recall as a child when "Jaws" came out, it was so terrifying I found myself fearing sharks even in fresh water lakes where I innocently thought then(thanks a lot bull sharks) that sharks couldn't be.

People who say you should always follow your instincts haven't really thought the issue through. Our instincts lead us astray sometimes. Following them blindly is like always taking advice from your grandpa. He lived in another time.

4 comments:

alex said...

"A total of 194 deaths through shark attacks have been recorded in Australia over the past two centuries, leading researchers to point out endlessly that more people die from bee stings and lightning strikes."

Comparing the frequency of shark attacks to that of bee stings is not responsible, and these "researchers" should rethink that strategy.

ScienceAvenger said...

Why not? What would be a more responsible comparison IYO?

alex said...

You missed it too? Tsk, tsk.
You need to compare things of /equal exposure/.

ScienceAvenger said...

Nonsense, you're too anxious to be the contrarian and missed the point entirely. People are far more concerned about shark attacks than they are about bee attacks, despite the fact that they are far more likely to die from the bees. The exposure makes no difference at all in that regard, because we aren't talking about a concern/unit of exposure here. We're just talking about general anxiety. Were we completely rational about it, we'd worry more about the bees on average, and we don't.

It's best not to assume people who do studies like this are idiots, it makes ignoring evidence too easy.