In this chapter, the authors introduce break-down plots, which offer a possible solution. The plots can be used to present “variable attributions”, i.e., the decomposition of the model’s prediction into contributions that can be attributed to different explanatory variables. The authors introduce more formally the method of variable attribution. They first focus on linear models, because their simple and additive structure allows building intuition. BD plots offer a model-agnostic approach that can be applied to any predictive model that returns a single number for a single observation (instance). One consists of identifying the interactions that cause a difference in variable-importance measures for different orderings and focusing on those interactions. The other one consists of calculating an average value of the variance-importance measure across all possible orderings. Thus, the choice of the ordering of the explanatory variables that is used in the calculation of the variable-importance measures is important.