Ludic Fallacy

Also Known As: Model-World Fallacy Platonification Error

Statistical Error ID: ludic_fallacy

Definition

The Ludic Fallacy (from Latin 'ludus' = game) is the error of applying probability models derived from well-defined, closed systems (like games, casinos, or textbook problems) to messy, open-ended real-world situations. Coined by Nassim Nicholas Taleb in 'The Black Swan' (2007), the fallacy highlights how statistical models that work perfectly in controlled environments can be dangerously misleading when applied to domains with fat-tailed distributions, unknown unknowns, and irreducible uncertainty.

Examples

A bank's risk model, calibrated on 30 years of market data, classifies a financial meltdown as a 1-in-10,000-year event. The model treats the market like a casino with fixed probabilities. In reality, markets have feedback loops, correlated failures, and structural changes that make historical volatility a poor guide to tail risk.

An actuary uses mortality tables built on past centuries of data to price life insurance. The tables assume stationary demographics and stable disease patterns. A novel pandemic or a breakthrough in gene therapy falls entirely outside the model — not as an unlikely event, but as a structurally different scenario the model cannot accommodate.

A chess program is used to predict optimal strategies in diplomatic negotiations. Chess has perfect information, fixed rules, and no ambiguity about legal moves. Diplomacy has hidden information, changing rules, irrational actors, and outcomes shaped by factors (domestic politics, personal relationships) that no closed game-theoretic model captures.

Verification Steps

Verification Steps

Binary yes/no questions that an AI must answer to detect a reasoning pattern in a text.

Each of the 452 aspects has verification steps — simple yes/no questions designed to systematically detect whether a pattern appears in a text. For ad hominem: "Does the argument attack a person rather than their claim?" For false dichotomy: "Are only two options presented when more exist?" This ensures consistent, reproducible analysis.

View in glossary →

Binary (yes/no) questions an LLM must answer to identify this aspect: