Multicollinearity: when two predictors tell the same story

Correlated predictors do not hurt your predictions. They make your coefficients unstable, your standard errors huge, and your model impossible to explain.

Linear Regression, Assumptions