If a model was fit to the whole population (in the statistical sense of every member of the community of interest), and a new member came into existence (maybe by being born or something), do you think that the model would be overfit or predict this new observation well