In multi-hop ad hoc networks, selfish nodes may unduly acquire high quality of service (QoS) by assigning higher priority to source packets and lower priority to transit packets. Such traffic remapping attacks (TRAs) are cheap to launch, impossible to prevent, hard to detect, and harmful to non-selfish nodes. While studied mostly in single-hop wireless network settings, TRAs have resisted analysis in multi-hop settings. In his paper we offer a game-theoretic approach: we derive a formal model of opportunistic TRAs, define a TRA game with a heuristic rank-based payoff function, and propose a boundedly rational multistage attack strategy that both selfish and non-selfish nodes are free to use. Thus non- selfish nodes are allowed to respond in kind to selfish ones. We characterize the form of equilibrium that the multistage play reaches and verify via simulation that it often coincides with a Nash equilibrium in which harmful TRAs are curbed in the first place, whereas harmless ones need not be.