The longer you spend on Facebook, the more likely you are to come across a post that says “we didn’t have seatbelts/safety features/basic worries for our survival needs when I was a youngster and we still lived.” If you haven’t met them yet, imagine someone telling you “my grandmother smoked every day of her life and she lived to be 95” or Uncle Billy telling you “I used to pound 35 cans of Budweiser before my commute and I never died.”
These are all examples of “survivor bias,” which states that because you are one of the survivors, these activities appear to be less dangerous than they are. It’s because Uncle Billy (God rest his soul) is gone that you don’t hear from people with comparable experiences, such as “I used to pound 35 cans of Budweiser before driving and died instantaneously, day one.” Let’s get away from the dismal examples and look at a couple of examples that pop up on the Internet now and then: The plane and the helmet. Wars are an excellent area to look for survivors, which is why both of these stories take place during a conflict.
This one goes like this, albeit it’s probably apocryphal: Generals became concerned during WWI when hospitals began to see a high number of soldiers with head injuries. Given the influx, they reasoned that the problem could be that the newly adopted helmet – especially, the Brodie helmet – was causing brain injuries. In reality, the helmets were allowing significantly more soldiers to live, which is why they were seeing so many head injuries.
The Americans intended to limit the number of casualties in their air units during WWII. Many planes were found to have bullet holes in three major areas: the fuselage, outer wings, and tail. They devised a plan to reinforce the crap out of the locations that had been hit by enemy fire. Which sounds reasonable?
However, the fact that the only data they received was from the survivors is why this is such a good illustration of survivor prejudice. Before they could begin reinforcing the areas, a Hungarian-Jewish statistician named Abraham Wald examined the data and recognized the weakness in their reasoning.
In essence, bullet wounds in the fuselage, outer wings, and tails of the planes that survived demonstrated that being shot in those three regions and making it back suggested that being shot there wasn’t highly damaging to the planes. If you assume (and you have to assume because you’re not seeing all of the evidence because it’s lying somewhere in a war zone) that bullet holes are distributed fairly evenly across planes when you look at all of the planes that go out to fight, then the ones that didn’t make it back had bullet holes in other areas.
