Black Box Thinking
I have been reading the book called ‘Black Box Thinking’ by Matthew Syed. This book has taught me that the more we fail in practice, the more we can confront and learn from our mistakes to help us grow and succeed. Lying to ourselves that the failure didn’t occur will only destroy the possibility of learning.
The first chapter in the book talked about the ‘logic of failure’, the example that really stuck with me is that during wartime, many of the planes that came back had bullet holes all over the wings and fuselage but were not hit in the cockpit or tail of the plane. With this information, the military command came up with what seemed like the perfect plan, they would place the armour on the areas where there were holes. However, Abraham Wald disagreed, he realised that the chiefs had neglected some key data. They were only considering planes that had returned and not taking into account the ones that haven’t. The information suggests that the cockpit and tail didn’t need armour because they were never hit but in fact, the planes that were hit in these places were crashing because they were actually the most vulnerable. The holes in the returning planes meant there could be damage but still return home safely.
Black box thinking is about creating cultures that enable us to learn from failures rather than being threatened by them. To utilise black box thinking, it is important to create a system where it’s safe to fail and to have constant feedback. When feedback is delayed, it is considerably less effective in improving judgment, Daniel Kahneman illustrates how quickly we learn to steer a car because it is instant feedback but it is harder to learn to steer a ship as there are long delays between actions and outcomes.
Black box thinking and marginal gains go hand in hand, the idea is to break down a big goal into small tasks, then improve each of them with constant feedback and when you put them together you will deliver a better product. To measure this, Matthew Syed talks about conducting Randomised Control Trials (RCT), this is where a number of people are split into two groups, one group with the new product and the other group with the existing product to compare then both outcomes are measured for impact.
The examples of black box thinking throughout the book were mainly about aircraft engineers and medical professionals, but it does not only apply to these professions. While reading this book, I thought about how Codeweavers follow an Agile workflow methodology which I believe is an element of ‘black box thinking’. We have an overall goal that we break down into tasks and within those there are sub-tasks that we continuously iterate and improve on. We also have a system where issues can easily be reverted back to its last working state via a feature toggle or a code revert.
The book recommended by Matthew Syed which I’m going to read next is: Little Bets by Peter Sims. This book promotes the idea of small scale experimentation as a key part of the process. If ten features can be launched while it takes a competitor to launch one, it means there are ten times the amount of experience to draw from to determine the success and failures.
Check back in a few weeks to read my next blog about Little Bets!