If anything, ecological research that employs models should be able to answer practical and critical questions, besides theoretical ones. More efforts should be put on the variable selection to improve the understanding of causality between actions and impacts. Big data should not be tortured, without preliminary hypothesis, until they confess something useful only to publish flawed papers. They should, instead, improve model elaboration to align them with the management decision-making, quantifying uncertainty, in order to have enough predictive performance. Similarly, advanced computing resources should not be wasted in purposeless data exploration. They should, instead, increase the sensitivity and validity of our analyses, to better understand the adequacy and importance of models in improving forecasting, solving management problems, and advancing theories.