We’re Terrible at Estimating Risk

Click the picture to see it at full size.

We’re terrible at estimating risk, especially when it involves things for which we have strong emotions or biases. Consider the photo above, which was posted to Facebook. It features someone’s kid, posing in front a collection of serious firepower, with this caption:

I just wanted to take a minute and remind everyone to keep their outlet plugs in if they have small children in the house.

Poe’s Law rears its head again: it’s so ridiculous that it has to be a joke, yet there’s no shortage of people — especially in America — who are pretty certain that a gun is the most important piece of safety equipment you can keep in your home, in spite of the overwhelming evidence to the contrary.

By the bye:

• I think guns are still pretty fun, but I also think they’re a privilege that’s all-too-often mistaken for a right, and
• this doesn’t invalidate what the Facebook poster says about electrical outlets.

Maybe part of it is because we’re not all that hot at teaching math to laypeople, but even those of us who took statistics and probability courses get risk estimates wrong all the time. We’re just not wired to have a good “gut sense” for what’s risky and what isn’t.

Consider sharks and vending machines. Which kill more people? (The following fun fact photos come from Buzzfeed.)

I assume that these get killed in their attempts to tip vending machines in order to get free stuff, only to be crushed to death.

Which kills more: vending machines or roller coasters?

Which kills more: roller coasters or high school football?

High school football ain’t got nuthin’ on hot dogs:

If I placed a wager on what killed more people — slipping in the tub or falling out of bed — I’d have lost my money:

Does Black Friday shopping kill this many people?

And finally…

I once joked about an idea for a mobile app that would’ve saved the lives of people like David Carradine. If you were into autoerotic asphyxiation, you’d check in with this app before commencing the jerky-choke, and then go about your business. If you didn’t check in with it again within some predetermined time (30 seconds? 60 seconds? I have no idea how long it’s supposed to take), the app would call for help. I remember thinking that even if the App Store would approve such an app, there wouldn’t be a big market for it.

I think I might be wrong: