We prescribe the minimal set of experimental data and parameters that should be reported for random-laser experiments and models. This prescript allows for a quantitative comparison between different experiments, and for a criterion whether a model predicts the outcome of an experiment correctly. In none of more than 150 papers on random lasers that we found these requirements were fulfilled. We have nevertheless been able to analyze a number of published experimental results and recent experiments of our own. Using our method we determined that the most intriguing property of the random laser (spikes) is in fact remarkably similar for different random lasers.