Traffic itself is simple-just analyzing it is not

Traffic itself is simple-just analyzing it is not. So why should we make our lives difficult by trying to identify the underlying determinism of traffic by analyzing given time series with the usage of stochastic tests like regression or spectral analysis or with the usage of strange attractors as they are utilized in nonlinear dynamics? And, moreover when applying such models for analysis we should be aware that we usually filter out all heterogeneous dissimilarities which can cause completely different explanations of our observation. If the above hypotheses are correct as discussed in the paper, it is much more easy to produce traffic in our computer terrarium by generating artificial agents that request mobility in TrafficLand and employ simple rules, such as preferences for or against choosing a specific freeway, an individual travel speed, etc., that interact with each other under the given environmental conditions; then we just examine the behavior of the whole population from a macroscopic point of view in order to compare it with real world observations. It is illustrated that by doing so, we discover simple fundamental individual mechanisms that generate the highly complex and dynamic collective behavior of traffic.