We present a novel approach in characterizing the optimal reliable multi-hop virtual multiple-input single-output (vMISO) routing in ad hoc networks. Under a high node density regime, we determine the optimal cardinality of the cooperation sets at each hop on a path minimizing the total energy cost per transmitted bit. Optimal cooperating set cardinality curves are derived, and they can be used to determine the optimal routing strategy based on the required reliability, transmission power, and path loss coefficient. We design a new greedy geographical routing algorithm suitable for vMISO transmissions, and demonstrate the applicability of our results for more general networks.