Survival bias is a concept that is often overlooked but has far-reaching implications for our decision making and data analysis. It describes the logical fallacy of focusing only on the surviving elements of a process, ignoring those that have been eliminated. This bias can lead to a distorted perception of success and has profound implications in areas such as business, the military and research. In this post, we'll take a closer look at survival bias, provide real-world examples, and discuss how to account for it in your analyses.

What is survival bias?

Survival bias occurs when an analysis is based only on the "survivors" or successful instances, while those who failed are not taken into account. This leads to a distorted view, as the reasons for failure are ignored and the remaining data is no longer representative of the original group or situation.

Examples of survival bias

1. Military strategy in the Second World War: A classic example of survival bias comes from the Second World War. The Allies wanted to better protect their bomber planes and analyzed the bullet holes of the returning planes. The original assumption was to reinforce the areas with the most bullet holes. However, statistician Abraham Wald pointed out that the planes analyzed were the survivors - and the true critical areas were those without bullet holes on the returning planes, as hits in these areas were likely to result in the loss of the plane.
2. Startup success stories: In the world of startups, we often hear about spectacular success stories such as Google, Amazon or Facebook. These stories can lead to the belief that success in the tech industry is widespread. What often goes unnoticed are the numerous startups that fail and never reach the public eye. This selective perception leads to a distorted picture of the actual risk and success in the industry.

How to take survival bias into account in analyses

• Search for complete data sets: Try to build analyses on complete data sets that include both successes and failures. This may mean actively looking for data on failed projects, companies or experiments.
• Consider context: Understand the context of your data thoroughly. This includes knowing how and why certain data was collected and what factors may have led to the exclusion of data.
• Diversify your sources of information: Do not rely on a single source of information. By diversifying your sources, you can obtain a more balanced and complete picture of the phenomenon under investigation.
• Critical self-reflection: Always question your assumptions and considerations. Is it possible that you are unconsciously overlooking information that does not fit with your previous understanding?

Conclusion

Survival bias is a subtle but powerful force that distorts our perception of success and failure. By being aware of this bias and applying careful analytical practices, we can make more informed decisions and reach more accurate conclusions. In a world driven by data, understanding and accounting for survival bias is critical to developing realistic strategies.