Universal Principles of Design, Updated and Expanded Third Edition

018 Box’s Law

All models are wrong, but some are useful.

Box’s law, proposed by the statistician George Box, states that models are simplified representations of real-world systems and as such are always “wrong” in the sense that they do not capture the full complexity of those systems. Despite this, models can be right enough to be useful. For example, maps are a type of model; they are wrong in that they are scaled-down simplifications but right enough to be useful.1

Models are useful when they have explanatory or predictive power. For example, models in evolutionary biology tend to have high explanatory power but low predictive power — i.e., they can explain mechanisms of change in great detail but are limited in the kinds of practical predictions ...