We discuss the accuracy of a mechanism, called Oracle, to assess position and transmission ranges in VANET. We prove that, in the best case, the assessments of the oracle are exact. Moreover, we prove that the average case of the Oracle coincides with its best case. Therefore the Oracle is optimal. The errors that afflict the assessments of both the positions and the transmission ranges depend on the relative speeds between the vehicles. We prove that the distribution of those relative speeds is an unimodal distribution peaked in zero. We show and briefly discuss the results of several experiments we made, that fully confirmed our theoretical results.