Rosling spent decades asking well-informed people basic questions about the state of the world and watching them get the answers wrong — not randomly, but systematically, in the same directions, for the same reasons. Knowing more, it turns out, is not the same as seeing more clearly.

There is a particular kind of wrongness that is worse than ignorance. It is the wrongness of people who are well-informed, follow the news, and have considered opinions — and are still systematically mistaken about the state of the world. Hans Rosling spent decades demonstrating this in lecture halls and conference rooms, asking audiences of doctors, journalists, and senior officials to answer basic questions about global health, poverty, and literacy, and watching them perform worse than chance. Not worse than experts. Worse than random. The instincts that were supposed to help them navigate complexity were actively misleading them.

Factfulness, written with his son Ola and daughter-in-law Anna Rosling Rönnlund, is his account of why that happens and what to do about it. Though the data is striking per se, Its most useful contribution is the taxonomy of misreading. Rosling identifies ten systematic instincts that distort our picture of the world, and three of them are worth remembering because they describe failure modes that are directly relevant to any analytical practice.

The gap instinct is the most structurally important. It is the tendency to divide the world into two categories — rich and poor, developed and developing, us and them — when the reality is a continuous distribution in which most of the world occupies the vast and varied middle. Survey categories, market segmentation frameworks, development economics indicators — these tend to produce the binaries they were designed to measure, and the binaries then shape what questions get asked and what answers are considered meaningful. Rosling's corrective is to look at the distribution rather than the poles.

The negativity instinct operates at a different layer, the affective one. It is the tendency to weight bad news more heavily than good, to register deterioration more readily than improvement, and to mistake the vividness of reported crises for an accurate picture of overall trends. This is a predictable consequence of how attention works and how news is selected. The practical implication for research is precise: what is salient in a given moment is not the same as what is structurally significant over time. The instinct toward the dramatic and the negative is a filter that distorts the signal, and it operates most powerfully on people who are most engaged with current information — which is to say, on exactly the people whose job is to understand what is happening.

The straight line instinct is the most technically precise of the three. It is the assumption that trends continue in the direction they are currently moving, at roughly the rate they are currently moving — that a rising line keeps rising, that a falling one keeps falling. Rosling demonstrates repeatedly that this is wrong: populations stabilise, diseases recede, literacy rates plateau. The shape of change is rarely linear, and the straight line projection — intuitive, easy to communicate, built into most forecasting — systematically misrepresents it. Read alongside Taleb's work on the unpredictability of rare events, this instinct describes the other half of the same problem: we misread not just the discontinuities but the continuities, projecting false linearity onto processes that are already changing shape.

What these three instincts share is that they are not corrected by more engagement with the present moment. Deeper immersion in current events, more interviews, more ethnographic attention — none of these tools can reveal that child mortality has fallen by half since 1990, or that the majority of the world's population now lives in middle-income countries, or that the straight line projection for population growth has been wrong for forty years. For those corrections, you need data. You need the long view that only systematic measurement over time can provide.

The harder question begins where the book's argument stops. Data reveals the errors our instincts produce — but data is not outside those instincts. It is produced by people and institutions operating within them. A poverty line drawn at one threshold rather than another doesn't just describe the gap instinct; it instantiates it, deciding by definition who falls on which side of the binary. A health indicator that measures what is easily measurable may systematically undercount what is not. The question of who collected the data, under what definitions, with what instruments, and in service of what purposes is not a supplementary concern. It is the prior question — the one that determines what the numbers are actually telling you, and what they are quietly leaving out.