How Not To Be Wrong: The Power Of Mathematical Thinking: Summary & Key Insights

One of the first things I emphasize is that our intuitive grip on chance is fragile. The law of large numbers tells us that random variation evens out as we take more samples—the average converges to the expected value. But human instinct rebels against this: we see patterns in small samples and draw reckless conclusions. That’s what happens when investors misinterpret streaks in the market or when policy makers change course because of a few atypical data points. The larger the sample, the more stable the truth it reveals, but most real-world decisions are made under small samples. We need mathematics to remind us that what we call luck might often be statistical noise.

In one example, I discuss how small counties sometimes top cancer-rate statistics. It’s tempting to infer that something special—perhaps an environmental factor—lurks there. Yet small population size itself magnifies apparent variation: with fewer people, random clustering looks dramatic. The very same counties can be both the healthiest and the sickest in different years merely because of sampling fluctuation. Understanding this humbles our desire to find causes where none exist. The law of large numbers doesn’t just state that averages stabilize—it warns us against overreacting to fluctuations, a lesson as much psychological as mathematical.

We love straight lines because they are simple. Draw two points, connect them, and we have a relationship. That’s our default mental model of cause and effect: if I work twice as hard, I’ll earn twice as much; if the car goes twice as fast, it must consume twice the fuel. But nature laughs at our lines. Most relationships are nonlinear: they bend, flatten out, or explode beyond certain limits. Linear thinking seduces us because it’s easy to compute, yet it hides the truth by ignoring curvature.

I explain this with a story about air resistance. If speed and drag were linearly related, airplanes could fly with predictable efficiency. But as speed doubles, drag quadruples. The relationship isn’t proportional; it’s exponential. Recognizing that many systems behave this way—whether in economics, population growth, or disease spread—changes how we make decisions. The COVID-19 pandemic made us all witnesses: infection rates grow nonlinearly, and small changes in behavior make enormous differences in outcomes.

Mathematical thinking means always asking whether the line we draw through experience truly matches the world’s curve. Linear approximations can be useful, but only if we keep in mind what they ignore. Once you see the world as nonlinear, your mind stops trying to force symmetry where there is complexity. That’s when mathematics becomes not constraint but liberation.