Was it the Boogey Man in Orlando? Why did we not identify him before it became tragic for so many?
Tuesday, 7 June 2016
The Orlando horror has added to our desperation to identify those who will do bad things. We want to believe it is possible to catch them before something terrible happens and the crass industry of risk assessment, management and mitigation uses every opportunity of this kind to reinforce the belief that it is possible. Our fear invites faith in those who say they can. The industry players say they can extract the golden threads of knowledge running through examples of spree killers who have tormented schools, college campus's and elsewhere to educate all others on identifying the likely culprits before they do something awful. I have sat through a number of these workshops. They empower organizations to think they know what they do not know and give well meaning administrators the capacity to claim due diligence has been done. The story telling is often evocative, but mostly they are spoofery. You cannot work backwards from a single example, or even a series of single examples, because you never know how many people shared the golden thread but did not do anything.
Some claim to have scientific tools that allow for accurate assessment of risk of those already suspected of being risky. There is some value in developing systems so an organisation can respond to obvious concerns. Seeking objectivity in the assessment of the risk that individuals might constitute is also worthwhile, but it is mistaken to rely on the people or the methods they use to estimate the risk of violence in any one person. I can say this having been one of these people and having used these methods. Here is why confidence is misplaced.
- Assume that 1/1000 men will commit an act of serious violence in a given time period.
- A risk assessment procedure is created that differentiates, with 95% effectiveness, those who will be violent from those who will not. (This is a vastly superior accuracy than any method currently available)
- Assume now that 100,000 people are tested and of the 100 who would commit violence, 95 would correctly be identified by the risk assessment instrument. 5 violent men would be overlooked.
- Of the remaining 99,900 who would not be violent (ignoring the 5 who have been overlooked), 4995 would be identified as likely to be violent with the same confidence as those who were actually violent.
How can we distinguish between the 100 we know will be violent from the 4995 we identify as likely to be violent but who will not be? The answer is, we cannot.
Risk assessment tools and workshops are generally insufficient for the task of identifying the Boogey Man, but they are lucrative for the providers and reasonably good at finding the sub-group of a population in which the Boogey Men are. This is as good as it will ever get.
In The Last Truth I trace the evolution of characters that explore this problem. There is someone slowly becoming a Boogey Man, someone who is accused of being a Boogey Man, and another who denies being one, but which of them is dangerous?
Get updates via email
Join 120 other awesome people who subscribe to new posts on this blog.