A topological definition of dynamic attractors leaves out metric information relevant to modeling real-world systems, particularly how far the attractor persists against perturbations and error. In this talk we consider features of an ODE attractor that eigenvalues and basin size do not capture. We introduce a quantity “intensity of attraction” that both indicates a basin of attraction’s capacity to retain solutions under time-varying perturbations and yields a lower bound on the distance the attractor continues in the space of autonomous vector fields. We contrast two numerical approaches to computing intensity, based on set-valued Euler method and optimal control.