Revised Raff’s Method for Estimating Critical Gaps

The estimation of the critical gap was introduced in the 1970s to evaluate the capacity of vehicle and pedestrian movements at unsignalized intersections. The critical gap is the smallest gap that a driver is assumed to accept. The population of drivers, each with his or her own critical gap, will have a distribution of critical gaps; and it is this distribution, or its characteristics, that is the subject of this paper. These statistics for the critical gap are difficult to estimate; several methods have been developed and used. Raff’s method has been frequently used because of its simplicity. This paper examines the original method, an adaptation of the method that is frequently used, and a proposed adaptation of the method. Simplified traffic and driver behavior characterizations are used to demonstrate the ability of each process to predict the mean critical gap. When given exponential headways in the major stream and a uniform distribution of headways, unbiased estimates of the critical gap would be obtained by using a modified version of Raff’s method involving the accepted gap and the maximum rejected gap. The mean critical gaps estimated with a modified Raff’s method using the maximum rejected gap are slightly inferior to those from the maximum likelihood method. It is recommended that the maximum likelihood method be the preferred technique; it was used in measuring values for the Highway Capacity Manual. However, the modified Raff technique is an acceptable alternative.