Back in the height of World War II, the United States faced a huge problem. Bomber pilots were getting picked off at a 45 percent clip. In an effort to find solutions, the Air Force started to collect and examine returning planes. If they could draw themes about the survivors, they could figure out how to design the best aircraft that had support and agility. A balance had to exist.
Too much protection meant the planes couldn’t maneuver. Not enough prevented them from surviving.
What they didn’t realize, however, was their process to solve the problem was completely backwards. Plotting bullet holes on returning aircraft didn’t give them information on areas of need. It communicated areas of strength. Their big breakthrough came courtesy of Hungarian mathematician Abraham Wald. It completely changed military aviation.
Wald’s ability to maneuver survivorship bias was a critical turning point in the war. The problem, however, is we haven’t quite learned from it. If you’ve ever prescribed a sleeper stretch to a thrower with this specific “limitation,” you were probably thinking like the Air Force originally was. It wasn’t due to the data you were looking at.