No one could believe how Badly the Planes were damaged upon return.

How statistical analysis helped pilots & their planes to survive in WW2.

Data is an interesting thing. We collect so much of it. Everyone one is talking about it. ‘Big Data’ is mentioned all the time, and although the terminology has become commonplace, the application is not always discussed, hence ambiguity can arise on how to collect, analyse, and apply findings around data.

Let’s consider Abraham Wald, a Hungarian mathematician born in 1902. He is arguably one of the fathers of statistical analysis, and his methodology is still used and taught today when looking at problem solving with data.

The work he did in WW2 was infamous. Not so much for what he analysed, however more so around what others ‘didn’t’ analyse. When WW2 planes returned from battle, they realised that the planes were being hit primarily in the fuselage and fuel system, and the damage was great. Planes were still getting hit in the engine, however the bullet holes were far less.

Those in charge were addressing what they thought was to be a logical risk management strategy. If the planes are being hit in these areas, lets ‘armour up’ so to speak, and add extra protection around both the fuselage and fuel system, as these areas were collecting most of the bullet holes. Makes sense, right? Well, not to Wald.

He looked at the data somewhat differently. Everyone was focusing on the planes that returned, however Wald focused on the planes that didn’t. The data wasn’t wrong. It was the way that the data was being interpreted that was wrong. Everyone was simply looking at the wrong set of data. Wald then focused his attention on the planes that weren’t returning at all, and instead of suggesting adding extra armour to the fuselage and fuel system, he suggested adding extra armour to the engines.

This didn’t make sense to the powers to be, as the data was clear. However, as Wald pointed out, the planes that were being hit in the fuel system and fuselage were returning, therefore making these areas less of a risk. He decided to go where the bullet holes weren’t. By armouring up the engines, he saved countless planes & pilots during WW2.

The moral of this story is to think about the way you look at your own data. Don’t just look at the data that is in front of you, think about the data that is missing. If your existing customers give you a low rating on one aspect of your business, is this critical? Are there maybe bigger issues that you haven’t considered because you are focusing on the customers you have, instead of the ones that simply never return.

Rudi Tartaglia

www.tmarketing.com.au