Galaxy clusters grow by accreting smaller groups, individual galaxies, and diffuse gas from the cosmic web, with Lambda-CDM predicting a relatively tight correlation between a cluster's current mass and its mass accretion rate based on hierarchical structure formation. However, observations reveal much larger scatter in the mass accretion rates than expected: clusters of similar mass at similar redshifts show wildly different accretion histories, with some growing rapidly and others appearing nearly quiescent, and the scatter does not decrease even when controlling for environment, concentration, or dynamical state (Fakhouri et al. 2010; Ludlow et al. 2013). Lambda-CDM struggles with this tension because the same cosmological parameters and initial power spectrum that set the cluster mass function should also tightly constrain the accretion rate distribution through the extended Press-Schechter formalism or N-body simulations. The observed excess scatter suggests either that cluster assembly is more stochastic than predicted, that the dark matter halos respond differently to accretion events depending on unmeasured properties, or that large-scale environmental effects or non-gravitational physics play a stronger role than Lambda-CDM allows.