The paper investigates the properties of massive subhaloes in the Milky Way's halo, using $\Lambda$CDM simulations. It finds that the majority of the most massive subhaloes are too dense to host any of the bright satellites (dwarf galaxies) of the Milky Way, with peak circular velocities at infall ranging from 30 to 70 km/s and infall masses between 0.2 and 4 × 10^10 M⊙. This suggests that galaxy formation becomes effectively stochastic at these masses, contrasting the well-established monotonic relation between galaxy luminosity and halo mass for more massive haloes. The study also shows that at least two (and typically four) of these massive dark subhaloes are expected to produce a larger dark matter annihilation flux than Draco, one of the most promising targets for indirect dark matter detection. The authors discuss potential explanations for this discrepancy, including the possibility that baryonic feedback or different dark matter physics could reduce the central densities of massive subhaloes. If massive dark subhaloes exist as predicted, they would be strong confirmation of the standard $\Lambda$CDM paradigm and could host ultra-faint galaxies. If they do not exist, it would require a reevaluation of the fundamental assumptions of abundance matching models and the physics governing structure formation.The paper investigates the properties of massive subhaloes in the Milky Way's halo, using $\Lambda$CDM simulations. It finds that the majority of the most massive subhaloes are too dense to host any of the bright satellites (dwarf galaxies) of the Milky Way, with peak circular velocities at infall ranging from 30 to 70 km/s and infall masses between 0.2 and 4 × 10^10 M⊙. This suggests that galaxy formation becomes effectively stochastic at these masses, contrasting the well-established monotonic relation between galaxy luminosity and halo mass for more massive haloes. The study also shows that at least two (and typically four) of these massive dark subhaloes are expected to produce a larger dark matter annihilation flux than Draco, one of the most promising targets for indirect dark matter detection. The authors discuss potential explanations for this discrepancy, including the possibility that baryonic feedback or different dark matter physics could reduce the central densities of massive subhaloes. If massive dark subhaloes exist as predicted, they would be strong confirmation of the standard $\Lambda$CDM paradigm and could host ultra-faint galaxies. If they do not exist, it would require a reevaluation of the fundamental assumptions of abundance matching models and the physics governing structure formation.