rare events: there is an app for that

Is my flight going down?

The Economist has a story about an iOS app called “Am I Going Down?” that estimates the odds of your flight going down based on the departure and arrival airports, the airline, and the type of plane used [Link]. The methodology isn’t available, but presumably it’s based on past performance (# crashes / # of flights)

This app highlights our terror associated with rare events. The Economist story came out yesterday and there was a plane crash today. I’m trying not to overestimate the risk associated with rare events, but it’s hard when they happen and the endless news cycle begins.

People are really bad at assessing risk. There is a huge risk analysis literature on how we tend to overestimate rare events, and this overestimation drives policy decisions. I like two papers by Chauncey Starr and Chris Whipple about how we perceive risk and what we do about these perceived risk. I’ve included 2 figures from their 1980 Science paper on how to identify which technological risks are acceptable to society.

Perception of risk. We tend to overestimate rare risks. Unless they are incredibly rare, in which case we are able to ignore these risks completely.

Perception of risk. We tend to overestimate rare risks. Unless they are incredibly rare, in which case we are able to ignore these risks completely.

risk_perceived

Here are a few estimates of risks. We tend to overestimate rare risks (like those associated with getting vaccinated) and underestimate common risks (like Type 2 diabetes).

 

Starr, C., & Whipple, C. (1980). Risks of risk decisions. Science, 208(4448), 1114-1119.

Starr, C., & Whipple, C. (1984). A perspective on health and safety risk analysis. Management Science, 30(4), 452-463.

I’ve blogged before about rare risks. Here are a few of these blog posts:

What rare risks do you over-estimate?


2 responses to “rare events: there is an app for that

Leave a comment