We analyze the effects of dynamic packet traffic on jamming attacks in wireless networks. For random access over collision channels, the jamming problem is formulated as a non-cooperative game in which nodes choose their transmission probabilities under energy and delay constraints. We relax the standard assumption of backlogged nodes and evaluate the Nash equilibrium strategies for random arrivals, which introduces the possibility that jamming attacks fail due to empty packet queues at the transmitters. The maximum feasible throughput is derived depending on whether jammers have the queue state knowledge, or not. We also model the effects of erroneous queue state inference due to random packet traffic and incorporate the channel sensing capability before jamming. The analysis extends from one transmitter-jammer pair transmitting over a single channel at a single access point to multiple transmitters and jammers, and then to an arbitrary number of subchannels at multiple channel access points. In the resulting jamming games, we show that jammers cannot effectively increase the average energy cost and cannot decrease the feasible throughput for transmitters, if they face uncertainty on transmitter queue states. Therefore, medium access is less vulnerable to jamming attacks under increasing traffic uncertainty. This motivates the use of traffic dynamics as a defense mechanism to mitigate denial of service attacks in wireless access.