Cooperative transmission can exploit the broadcast nature of a wireless medium and leverage relay nodes to forward overheard data from the source to the destination. To optimize the performance gain, effective cooperation strategies are essential to identify the best relay(s) with a minimum overhead and enable forwarding with high success probability. In this paper, we focus on an opportunistic relaying scenario and develop two distributed cooperation strategies. Both adopt a backoff-based intergroup coordination, whereas the intragroup contention is based on either the forwarding probability or backoff timer. In particular, we employ stochastic geometry to address the impact of spatial distribution of relays. Considering a Poisson point process for random relays, we derive the probability distributions of the average received signal-to-noise ratio (SNR) and transmission success probability of potential relays. Making use of such statistics and location information, each relay can independently determine its contention parameters such as backoff time and/or a forwarding probability. We analytically evaluate the relaying performance and validate the accuracy with simulations. The results demonstrate the improvement over a pure probabilistic scheme and the gap to the upper bound of a centralized scheme with the preselected best relay.