Survivors can Kill you

Coffee was not as tasty and sophisticated as in Vienna. Abraham Wald was having these thoughts whilst sipping his first morning cup on a cold January morning of 1943. He was looking through the window of his Columbia University office as flocks of snow were falling on Manhattan’s streets. The cold winters had accompanied him throughout his life, from his hometown in Cluj (now Romania), to his college and doctorate years in Vienna to now New York City. This was the place that had warmly welcomed him and recognized his valued mathematician knowledge after having to flee Europe before the start of the war.

He turned his head to his desk. There, the most important challenge of his life - in the shape of a pile of papers with drawings, letters and numbers – was lying just waiting for him to solve it. He was part of the Statistics Research Group (SRG), a team of renowned professors (including Milton Friedman and George Stigler – future Nobel Prizes) gathered by the US government to support some of the military challenges encountered in the war with their scientific knowledge.

American bombers were suffering badly from the German air defence. The military commanders reached out to the SRG to consult where they should place additional armour on the planes. That was a critical question because armour increased weight and therefore lowered performance.

Wald, given his extraordinary statistics knowledge, had been assigned to sort this puzzle and to provide a recommendation to the commanders. They originally believed the best solution was to add armour to the fuselage area and not to the motors because the former was where most of the hits had been received. This was the no brainer solution and Wald was just expected to select the locations in the fuselage that were more prone to damage.

He had been thinking about this challenge for a couple of days and just as he drank the next sip of that tasteless coffee he had the eureka moment he would be remembered for generations. The formulation of the problem was valid. However, the sample taken was totally invalid. He had been asked to provide a recommendation only based on those planes that had been analysed, those that had survived. But, what about those that never returned? Those were the most damaged ones, probably the ones that had received hits in the motor area. Thus, that was the area of the plane to protect more.

Abraham Wald had almost followed a classical logical error to in the process of making a decision: the survival bias.

I am wondering how many decisions I make as a leader under this bias. Obviously, I do not count my team members as survivors or not survivors but I can count them as those who reach out more to me and those who reach out less, and I could even consider those that will sing me a tune that I like and those who probably not.

- How do you avoid the survival bias when making decisions on how to allocate your time/attention to your team members?

-        - How good are you finding out that your weakest leadership area is maybe not exactly where “the survivors” tell you it is?