Mapping the Territory
Photo by Clément Falize on Unsplash
Over the past fifty years, researchers have catalogued hundreds of distinct cognitive biases. The sheer number can be overwhelming — a 2016 visual codex by writer Buster Benson and designer John Manoogian III organised approximately 188 named biases into a single diagram. Rather than memorising a list, it is more useful to understand the underlying architecture: why so many biases exist, and what cognitive problems they arise from.
Benson's framework proposes that all cognitive biases are responses to four fundamental problems the brain must solve under constraints of time, information, and memory.
Problem 1: Too Much Information
Photo by Brett Jordan on Unsplash
The world generates far more data than the brain can process. To cope, we filter aggressively — but the filters themselves introduce distortion. We notice things that confirm what we already believe (confirmation bias). We are disproportionately drawn to unusual, vivid, or emotionally charged information (availability heuristic). We pay more attention to changes and differences than to absolute values (change bias).
These filters are efficient but selective. What passes through shapes our model of reality; what is filtered out is simply missing.
Problem 2: Not Enough Meaning
Photo by Brett Jordan on Unsplash
Information rarely arrives with a complete context. We fill the gaps with stories, patterns, and assumptions. We find patterns even in noise (apophenia — finding connections that don't exist). We project our own knowledge and perspective onto others, assuming they see what we see (curse of knowledge). We construct coherent narratives from incomplete evidence and then treat those narratives as facts (narrative fallacy).
The brain's meaning-making machinery is extraordinarily creative — and correspondingly prone to generating plausible-seeming fictions.
Problem 3: Need to Act Fast
Photo by Enayet Raheem on Unsplash
Decisions often need to be made quickly, without complete information. We are biased toward action and completion, even when waiting would be wiser (action bias). We maintain prior beliefs and decisions longer than the evidence warrants (status quo bias, sunk cost fallacy). We prefer safe, certain outcomes over uncertain ones with higher expected value (risk aversion).
These tendencies keep us moving and protect us from the paralysis of infinite deliberation — at the cost of systematic errors in how we evaluate options.
Problem 4: What to Remember
Photo by Enayet Raheem on Unsplash
Memory is not a recording device. It is reconstructive — we rebuild memories from fragments each time we recall them, and the reconstruction process introduces errors. We simplify and generalise experiences (stereotyping). We edit memories in light of later information (hindsight bias). We remember the peak and end of an experience more than the whole (peak-end rule, studied by Kahneman and colleagues).
Memory biases matter in finance because we constantly learn from experience — and if our memories of past decisions are distorted, our learning is compromised.
Biases Cluster into Families
Photo by Nigel Hoare on Unsplash
Within these four categories, biases group into recognisable families:
- Information processing biases: confirmation bias, availability heuristic, representativeness heuristic
- Memory and narrative biases: hindsight bias, narrative fallacy, rosy retrospection
- Social and self-perception biases: Dunning-Kruger effect, fundamental attribution error, in-group bias
- Decision and risk biases: loss aversion, anchoring, framing effects, present bias
The families overlap and interact. Understanding the architecture — rather than memorising individual biases — gives you a lens that applies across new situations you haven't explicitly studied.