All models are wrong, but some are useful.

— George Box (Statistician), Robustness in the strategy of scientific model building, p. 201–236, 1978