This week marks my fourth year teaching case-control studies. Each new cohort of students brings fresh perspectives, and it is an honor to teach them the foundational methods of epidemiology. I am continually humbled by the importance of these approaches, developed in the mid-20th century and still widely used today, often in the analysis of large-scale population health data.
In epidemiology, we use control groups alongside cases to identify factors that contribute to disease risk. Ironically, in many other areas of a researcher’s life, whether mentorship, grant writing, or entrepreneurship, we tend to focus exclusively on “success stories.” We assume that if we replicate the actions of these success cases, we can achieve similar outcomes. Yet this approach neglects an essential truth: not all success can be tracked down objectively, as parts of it are attributable to random chance.
Nassim Taleb, in his book Fooled by Randomness, highlights how easily we can be misled by fortunate outcomes. While causal factors (such as personal skills, hard work, and good planning) certainly matter, chance also plays a significant role in many achievements. If we only study those who succeeded (the “cases”), we might ignore how many other attempts failed under nearly identical conditions.
Without comparing both groups, it is impossible to separate factors that are associated with success from those that merely coincide with it. This principle is as relevant to grant writing and entrepreneurship as it is to epidemiology. If we only examine successful grant proposals, we risk attributing their triumph to certain strategies, overlooking rejected proposals that used similar methods. Likewise, if we only analyze thriving startups, we miss lessons from those that started under similar conditions but did not survive. By failing to include a “control” group, we may mistakenly conclude that certain factors caused success when they were, in fact, incidental.
A concise way to capture this idea is:
Desired Outcome=α×(Causal Factors)+β×(Randomness)
where α and β represent the respective weights of causal factors and chance. These factors vary greatly depending on context, but it serves as a reminder that outcomes are rarely due to causal factors alone.
Focusing exclusively on success stories can blind us to the significant role of chance and obscure the factors that drive outcomes. Recognizing the element of randomness in any venture reminds us to diversify our risks and remain open to multiple pathways. In doing so, we increase the likelihood of landing on a favorable outcome, no matter how unpredictable its origins.
Feb 2025
