A traffic light assistant on a smartphone is used in real and simulated driving in an experiment. The speeding behavior of the test subjects is assessed as metric of travelled distance with more than 5 km/h above the speed limit, subjective scores by NASA-RTLX, general fuel reduction and comparison of occurring accelerations. The speeding reduction showed numerically not the same values for real and simulated driving, but the reduction effect was present in both; in real with the smaller effect. According to subjective ratings the simulator driving is judged in tendency more demanding. A fuel reduction effect for the green wave coordination found in the simulator could also be found in real traffic, but also with lower effect. The experiments revealed no absolute validity for any metric, but relative validity for some metrics.