LLM Book Knowledgebase

Cognitive Biases and Heuristics

8 min read

Cognitive Biases and Heuristics

Cognitive biases are systematic errors in thinking that emerge from mental shortcuts (heuristics) which have evolutionary or adaptive roots but misfire in modern contexts. Daniel Kahneman’s Thinking, Fast and Slow offers the most comprehensive catalogue of these biases, grounded in decades of experimental research conducted with Amos Tversky. Understanding these biases is not primarily academic: each one has concrete implications for business decisions, leadership, investing, negotiation, and self-knowledge.

The Nature of Heuristics

A heuristic is a mental shortcut that produces approximate answers to difficult questions. The key mechanism: when System 1 faces a hard question it cannot answer quickly, it substitutes an easier related question and answers that instead, often without the substitution being noticed.

“This is the essence of intuitive heuristics: when faced with a difficult question, we often answer an easier one instead, usually without noticing the substitution.” — Daniel Kahneman

The practical implication: what looks like a direct judgment about a complex question (Is this CEO competent? Is this startup likely to succeed?) is often actually a judgment about something simpler and more immediately accessible (Does this person seem confident? Does their story feel coherent?). The easier question hijacks the harder one.

The Major Heuristics and Their Biases

Availability Heuristic

Judging the frequency or probability of an event by how easily examples come to mind. Events that are vivid, emotionally intense, or recently experienced feel more probable than they actually are.

“We defined the availability heuristic as the process of judging frequency by ‘the ease with which instances come to mind.‘”

Consequences:

  • Media amplification of risk: plane crashes are covered intensively; car crashes kill far more people but receive little coverage. People dramatically overestimate plane crash risk and underestimate car crash risk
  • Availability cascades: a small risk amplified by media attention becomes an “availability cascade” that distorts policy and public behavior completely out of proportion to actual danger
  • Success bias in learning: you learn about visible successes, not invisible failures. The companies that used a strategy and survived are available to you; the companies that used it and failed are not

Anchoring Effect

Numerical estimates are powerfully influenced by arbitrary starting numbers (anchors), even when the anchor is obviously uninformative.

“A key finding of anchoring research is that anchors that are obviously random can be just as effective as potentially informative anchors.”

Applications in negotiation, pricing, and estimation: whichever side introduces the first number exerts a gravitational pull on the final outcome. The antidote is to actively search for arguments against the anchor — “thinking the opposite” — rather than simply adjusting away from it (which typically produces insufficient adjustment).

Representativeness Heuristic

Judging probability by how well something matches a prototype or narrative, ignoring base rates.

The classic example: a “meek and tidy” person named Steve is described. Most people immediately judge him more likely to be a librarian than a farmer. But there are many more male farmers than librarians — the base rate dramatically favors “farmer.” The vivid personality description swamps the statistical reality.

“They made that big decision on the basis of a good report from one consultant. WYSIATI… They did not seem to realize how little information they had.”

The planning fallacy is a close cousin: people plan based on best-case scenarios (inside view) rather than base rates from comparable projects (outside view). The result is systematic underestimation of costs, time, and obstacles.

The Halo Effect

The tendency for a positive (or negative) impression in one domain to color judgments across unrelated domains. If someone is attractive, we tend to assume they are also intelligent, competent, and honest — even without evidence. In organizations, a CEO whose company thrives during a bull market gets credited with strategic genius; when the market turns, the same decisions look reckless.

“The halo effect is the tendency to like (or dislike) everything about a person — including things you have not observed.”

Loss Aversion and Prospect Theory

Perhaps the most economically significant cognitive bias: losses are felt approximately twice as powerfully as equivalent gains.

“For most people, the fear of losing 150. We concluded from many such observations that ‘losses loom larger than gains’ and that people are loss averse.”

The ratio is typically 1.5 to 2.5: it takes a potential gain of 250 to make someone emotionally willing to risk losing $100.

Consequences:

  • Status quo bias: people resist changes even when the changes are clearly in their interest, because the loss of the current state looms larger than equivalent gains
  • Sunk cost fallacy: continuing to invest in a failing project to “not lose” the prior investment
  • Negotiation asymmetry: your concessions feel more painful to you than they feel valuable to the other side, and vice versa — making agreements harder to reach than pure rationality would predict
  • Reference point dependence: satisfaction is not about absolute wealth but about movement relative to where one started

The key practical insight: the subjective experience of loss and gain are different from their objective magnitudes. This is not irrationality in a dismissive sense — it reflects a real asymmetry in evolutionary fitness, where losses were typically more dangerous than equivalent gains were beneficial.

The Narrative Fallacy and Hindsight Bias

Humans are compelled storytellers. We construct causal narratives to explain outcomes, even when the outcomes were substantially determined by luck. The narrative fallacy makes coherent stories feel like explanations — which is not the same thing.

“Narrative fallacies arise inevitably from our continuous attempt to make sense of the world. The explanatory stories that people find compelling are simple; are concrete rather than abstract; assign a larger role to talent, stupidity, and intentions than to luck; and focus on a few striking events that happened rather than on the countless events that failed to happen.”

Hindsight bias compounds this: after an outcome is known, it feels inevitable. We systematically underestimate how uncertain the future was when it was still the future, which leads to unfair attribution of blame and credit, and to overconfidence in our ability to predict future events.

The Preacher-Prosecutor-Politician Pattern

Adam Grant’s contribution to this territory is the identification of three motivated reasoning modes that systematically override genuine evaluation:

“We go into preacher mode when our sacred beliefs are in jeopardy: we deliver sermons to protect and promote our ideals. We enter prosecutor mode when we recognize flaws in other people’s reasoning: we marshal arguments to prove them wrong and win our case. We shift into politician mode when we’re seeking to win over an audience: we campaign and lobby for the approval of our constituents.”

These are not merely social roles but cognitive states — mental postures that determine what information gets processed and how. The antidote Grant proposes is the scientist mode: treating one’s current beliefs as hypotheses to be tested rather than convictions to be defended.

Bias in Groups vs. Individuals

A common fallacy is that group decisions are more rational than individual ones. In many cases the opposite is true: groupthink (the tendency of cohesive groups to suppress dissent and converge on consensus) can amplify rather than correct individual biases. The specific mechanism: whoever speaks first exerts an anchor effect on the discussion; high-status group members’ views are over-weighted; members with contrary opinions self-censor.

“A simple rule can help: before an issue is discussed, all members of the committee should be asked to write a very brief summary of their position. This procedure makes good use of the value of the diversity of knowledge and opinion in the group.”

Debiasing Strategies

No one can simply choose to stop being affected by cognitive biases — they are structural features of human cognition. But there are partial mitigation strategies:

  1. Outside view / reference class forecasting: Before estimating how long a project will take, ask what similar projects typically took
  2. Pre-mortem: Before committing to a major decision, imagine it is a year in the future and the decision produced a disaster — then explain why
  3. Structured decision processes: Evaluate candidates or options one attribute at a time rather than holistically, to reduce halo effects
  4. Deliberate “thinking the opposite”: When exposed to an anchor, actively generate reasons it might be wrong
  5. Intellectual humility practices: Grant’s prescriptions — seek disconfirming evidence, track your predictions and their outcomes, build challenge networks

Limits of debiasing

Kahneman is explicitly pessimistic about the ability of individuals to debias themselves through willpower or awareness. System 2 can sometimes override System 1, but it is slow, effortful, and easily fatigued. The most effective debiasing happens through environmental design and institutional procedures, not individual effort.

Backlinks