The Three Cs of Statistical Thinking
Calm, context, and curiosity are the only defences against being manipulated by numbers.
Tim Harford distils his years of presenting BBC's More or Less and writing How to Make the World Add Up into three principles he calls the three Cs — Calm, Context, and Curiosity. The framework emerged from his observation that statistical errors are rarely caused by mathematical incompetence. People fail to think clearly about numbers not because they can't do arithmetic, but because they encounter data in emotionally loaded environments designed to bypass rational evaluation — newspaper front pages, social media feeds, investment pitches — and react before thinking.
Be Calm: Notice your emotional state before evaluating the claim. If the headline made you angry, afraid, excited, or greedy, you are not yet in a good position to evaluate the underlying number. Emotional arousal is not a signal that the claim is true or important — it is a signal that someone wants you to act before you think. The same analytical rigour applied in a calm state will frequently produce a very different verdict.
Get Context: A number without comparison is almost meaningless. Is this figure going up or down? What was it five years ago? How does the equivalent figure look in comparable countries? What does a different but related metric show? Harford argues that even expert statisticians cannot evaluate a number without context — the failure to provide it is often deliberate. Seek the denominator, the trend, the comparison.
Be Curious: Approach data as a starting point for exploration rather than as ammunition for a pre-existing argument. The question is not 'does this support my view?' but 'what's actually going on here?' Harford observes that data is most commonly weaponised in arguments, not explored — and the shift from fighting with data to wondering about data changes everything.
- Emotional arousal is a red flag that someone wants you to act before you think — notice it and pause.
- A statistic without comparison, trend, or denominator is almost always incomplete and often deliberately so.
- The purpose of data should be exploration of reality, not ammunition in an argument.
- The same information looks different to a calm mind than to a frightened or excited one — always evaluate from the former state.
- Simplicity of advice does not mean simplicity of the world — the three Cs work precisely because the world is complicated.
- Notice your emotional state before evaluating the claimWhen you encounter a statistic or claim — especially on social media, in a news headline, or in an investment pitch — pause and name how it made you feel. Angry? Excited? Afraid? Greedy? That emotional state is the first thing to acknowledge, not the claim itself. Emotional manipulation is the delivery mechanism; the number is the payload.Pro tipSearch for the feeling first, then the fact. 'This claim made me feel X' is the first sentence to write. The second sentence is the analysis.WarningDo not assume the claim is false because it provoked an emotion. It may be entirely accurate. The point is to defer judgment until you are calm.
- Seek context before forming a viewFor any number or claim, ask: What was this figure 5 years ago? Is it rising or falling? What is the comparable figure in similar contexts? What is the denominator? What does a related metric show? Context doesn't require expertise — it requires the habit of asking these questions before accepting a single data point as meaningful.Pro tipThe question 'compared to what?' is almost always the most valuable question you can ask about any statistic.WarningContext can be weaponised too — cherry-picked comparisons can make any figure look exceptional. Look for comparisons that weren't provided as well as the ones that were.
- Approach data with curiosity, not an argument to winBefore engaging with a statistical claim, ask whether you are trying to find out what is true or trying to find evidence for something you already believe. The former is curiosity; the latter is motivated reasoning. Curiosity means following the number wherever it leads, including to conclusions you don't like.Pro tipA useful test: could you steelman the opposite conclusion using the same data? If not, you may not understand the data well enough yet.
- Apply the framework to investment claims specificallyInvestment marketing is specifically designed to trigger the calm-bypassing emotional states the first C addresses. 'AI is the future', 'this market is different now', 'you'll miss out' — each maps to fear, excitement, or greed. Apply C1 (notice the emotion), C2 (what context is missing from this pitch), C3 (what is this actually telling me about risk-adjusted return).Pro tipHarford's example: 'if your mate down the pub is telling you to buy Tesla because electric cars are the future, the market already knows.' The narrative feels like insight; context reveals it is priced in.WarningThe three Cs don't guarantee a correct answer — they guarantee a better process. Some genuinely compelling opportunities also trigger emotional arousal.
Research by economist Stefanie Stantcheva found that zero-sum thinking — the belief that one person's gain must come at another's loss — is more prevalent in cities, which seems counterintuitive since cities are built on collaboration. Harford demonstrates C2 (context): understanding why city-dwellers rationally experience housing and competition as zero-sum explains the data without dismissing it.
Harold Pollack's 'finance on an index card' concept — that all essential financial advice fits on a 3x5 card — illustrates C3 (curiosity): the willingness to follow simple principles wherever they lead rather than seeking complexity to justify fees or expertise.
Irving Fisher's 1929 public statement that 'stocks have reached a permanently high plateau' is a case study in C1 failure — he was excited by his own thesis and had lost the emotional distance to evaluate evidence against it. He was also violating C2 — he had no context for what happened when previously 'transformative' technologies peaked (as with railways in the 1840s).
Harford developed these three principles across years of presenting More or Less — the BBC Radio 4 programme dedicated to examining statistical claims in the news — and consolidated them in How to Make the World Add Up. He notes he arrived at three rather than ten because he gave versions of the same underlying advice to 10-year-olds and to PhD statisticians at the Royal Statistical Society, and found the substance was identical. The reduction from technical complexity to three memorable principles reflects his belief that the barrier to good statistical thinking is psychological, not mathematical.