The “reconstruction” of traffic accidents has traditionally involved relatively simple momentum and energy based calculations using evidence of vehicle collision and post-impact travel. This evidence is used to estimate the “initial conditions” present at the beginning of the accident sequence. Those lacking training, knowledge, or ethical restraint may often employ faulty data in developing an unproven “model” or opinion regarding how the accident happened. The scientific method requires a logical application of the laws of physics to the careful examination of evidence, thus arriving at a hypothetical model of how the collision occurred. Yet the scientific method has not been fully satisfied until that model has been tested.
Actual testing of an accident model by the controlled crashing testing of vehicles is seldom practical. However, the reconstruction model of the accident can often be tested and refined through the use of computer simulations. Not every accident can be effectively modeled by computer simulation, due to both program and evidence limitations. However, if an accident reconstructionist has not tested his hypothetical model, he has not fully satisfied the requirements of the scientific method.