This study empirically tests two foundation ecological theories: (1) pack hunting is a driver for the evolution of sociality; and (2) species have a finite energy potential, whereby increased maintenance costs result in decreased reproductive effort. Using activity and prey data from 22 packs of African wild dogs (Lycaon pictus), we parameterized a model detailing the energetic cost/benefit of cooperative hunting. Larger pack size increased foraging time, prey size, and capture probability while reducing chase distance, resulting in a rapidly increasing net rate of energy intake up to a pack size of five, which peaked at 10 individuals and then declined. With a streamlined body plan necessary for hypercursoriality limiting stomach capacity in smaller packs, it was demonstrated that the group hunting benefit will rather accrue to widely foraging predators than to "sit-and-wait" ones. Reproductive effort, measured by the number of pups born, revealed smaller litters with decreasing pack size, validated finite energy theory, and highlighted a "poverty trap" where smaller groups have lower foraging gains, smaller litters, and increased vulnerability to extirpation. Consequently, these results demonstrated a mechanistic example of pervasive selection for maximal body size (Cope's rule), leading to a macroevolutionary ratchet, where sociality linked to hypercursoriality is betrayed by an Achilles' heel.