Flooding is the simplest mechanism for broadcasting in mobile ad hoc networks (MANETs), where each node retransmits a given broadcast packet exactly once. Despite its simplicity, flooding can result in high redundant retransmission, contention and collision in the network, a phenomenon referred to as broadcast storm problem. Several probabilistic approaches have been proposed to mitigate this inherent phenomenon. However, majority of these schemes uses fixed rebroadcast probability which is quite unlikely to be optimal in other simulation set up. In this paper, we propose an improved rebroadcast probability function which takes into account immediate neighbor information and some key simulation parameters (i.e. network topology size, transmission range and number of nodes) in determining the appropriate rebroadcast probability for a given mobile node. Simulation results reveal that this simple adaptation achieves superior performance in terms of saved-rebroadcast, number of retransmission node and end-to-end delay, without sacrificing reachability in dense network.