Jorge Luis Borges imagined an empire whose cartographers drew a map so detailed it was the same size as the empire itself. It was, of course, useless. A map’s value is in what it leaves out — the irrelevant streets, the unchanged fields — so that what remains becomes visible.
Every model you build is a map. A linear regression draws one straight line through a cloud of points and declares, “this is the relationship.” The cloud disagrees at every point. That disagreement — the residuals — is not a flaw; it is the honest price of simplification. The danger arrives when you forget the residuals exist, when you treat the line as the territory.
The best practitioners hold two truths at once: the model is useful and the model is wrong. The gap between the line and the dots is where humility lives.