World War II, survivor bias and risk based decision making

World War II, survivor bias and risk based decision making

Sep 5th, 2013

World War II, survivor bias and risk based decision making

During World War II bombers crews knew their chances of coming home were extremely low because of anti-aircraft and enemy interception. Airmen were described as “Ghosts already” by the historian K. Wilson. The Air Force asked a panel of scientists, engineers and statisticians to develop a plan to enhance the chances of success of air missions, which in the mind of the military was equivalent to reinforcing the planes, building more armor, etc. Obviously the more armor, the less bombs could be carried, so there was a compromise to be found.

Based on the location of the bullet holes on the planes returning damaged to their bases, the military wanted to reinforce the armor of the planes in the most damaged areas. The panel, however, opposed to that course of action, rightly claiming that was a poor decision.
The military were falling to the survivor bias: the planes coming back with bullet holes were actually displaying areas were the planes were hit, but survived. Most likely planes that were hit in other location did not come back! The military’s logic was flawed, at least as much as our logic when we look at “highly successful” role models.

Sometimes, in technical applications, you see papers on failures. In a prior blogpost we taught about Pugsley  and how some companies actually foster “failure”.

In our day to day practice, our reports always have a section on “Past accidents/failures history”, and we never neglect looking at “near misses”, as they can bring interesting information on “what could have happened, but did not, that time”. In those cases, we develop a lot of effort to understand the factors that made the situation evolve towards a benign ending. As a matter of fact, unfortunately, each time things pan-out favorably (let’s call this a “survivor situation”), the non-survivor are easily forgotten. We all recall with ease situations that panned out nicely, possibly because “we were so lucky”, and we do not spend enough time analyzing that detail that transformed a potential catastrophe into a success. Most of the time we totally remove that detail from our “inner records”.

As risk managers we owe to reopen the book of accidents, failures, near misses, and draw from them as much unbiased and uncensored information as we can.

Tagged with: , , , , , ,

Category: Consequences, Crisis management, Mitigations, Probabilities, Risk analysis, Risk management

Leave a Reply

Your email address will not be published. Required fields are marked *

Riskope Blog latests posts

  • Landslides risk assessment and monitoring
  • 8-03-2023
  • During the first couple decades of our professional life we worked extensively with Landslides risk assessment and monitoring in the…
  • Read More
  • Societal risk acceptability in the Canadian House of Commons
  • 1-03-2023
  • We are very pleased to see the notion of societal risk acceptability in the Canadian House of Commons, emerge publicly…
  • Read More
  • Multi dimensional Consequences
  • 22-02-2023
  • Multi dimensional Consequences approaches take into consideration the varied losses that a failure can generate. Thus they allow for a…
  • Read More
  • Get in Touch
  • Learn more about our services by contacting us today
  • t +1 604-341-4485
  • +39 347-700-7420

Hosted and powered by WR London.