Sept. 6, 2019
By Gerry Turcotte
Recently I read a wonderful LinkedIn entry by Aron Laxton about the U.S. Navy’s efforts to study and reinforce aircraft based on planes that had been damaged from the front. Engineers studied and mapped the bullet holes that peppered the “wounded” planes and determined that additional armour needed to be added to the wingtips and to the central body of the aircraft.
The bullet holes were proof of where the aircraft was vulnerable. Or so they thought.
A statistician on staff, Abraham Wald, however, deeply disagreed. He proposed, instead, that additional armour be added to the nose, engines and mid-body. His colleagues thought he was crazy. None of the planes had showed any such evidence of damage.
As Laxton explained, though, “Wald realized what the others didn’t. The planes were getting shot there too, but they weren’t making it home. What the Navy thought it had done was analyze where aircraft were suffering the most damage. What they had actually done was analyze where aircraft could suffer the most damage without catastrophic failure. … They weren’t looking at the whole sample set, only the survivors.”
It’s a wonderful example of misperception — or of studying a question from the wrong point of view and missing the obvious. In one of the many comments to this post, a respondent connected this to customer service. “We get input from customers on where our products/services don’t meet their requirements and then use that input to improve our processes. Unfortunately this is biased information. It’s ‘survivor bias.’ What about getting input from the ones who left — the ones who gave up on doing business with us?”
Note: This is an Abuse Tracker excerpt. Click the title to view the full text of the original article. If the original article is no longer available, see our News Archive.