A place for philosophy? The rise of modeling in statistical science

Abstract

Large statistical models seem to have reached epidemic proportions. I will document the harm they are currently causing and contrast it with some success stories of smaller modeling efforts. A case study involving a model for flipping coins shows how “adding a few bells and whistles” can lead to nonsense. Finally, an effort to list the “real uses” of models is offered as a start to making sense of current modeling efforts.