The Atlas: A Taxonomy of Cognitive Bias, Logical Fallacy, Statistical Artifact & Systemic Failure

The Atlas is a structured reference — one of several tools built by Incognati — for the ways human reasoning, statistical inference, and institutional systems go wrong. Each concept has a definition, a mechanistic explanation, a semantic classification, and edges connecting it to related concepts.

The taxonomy spans individual cognitive biases (confirmation bias, anchoring), statistical artifacts (survivorship bias, p-hacking), rhetorical failures (ad hominem, straw man), and systemic distortions (regulatory capture, Goodhart's Law). Concepts are drawn from cognitive psychology, statistics, epistemology, decision science, information theory, and organizational behavior.

This is a reference tool, not a curated essay. If you are new here, start with the Curated Subset — 200 of the most recognisable concepts, with full definitions, examples, and countermoves. For deeper exploration, use the views below.

This atlas documents named and described phenomena — it does not include every possible way a system can fail, only those that have been identified, named, and described well enough to define. Concepts without definitions or without a meaningful mechanistic account are excluded.

The taxonomy is descriptive, not prescriptive. Including a concept does not imply endorsement or that the phenomenon is inevitable. Many concepts have mitigations; those are noted in individual concept records but are not the primary focus here.

57% of concepts have been formally studied in peer-reviewed literature. The remaining 43% are informally documented in professional, practitioner, or folk contexts — included because they appear regularly in reasoning about real systems, even where the empirical evidence is thin.