Adjusted plus-minus ratings, etc.

Thanks for the link. I'm trying to learn how to best apply Bayesian methods to basktball, and I fit a hierarchical model similar to the model in the "basketball statistics" post above (see http://www.basketballgeek.com/2009/09/07/a-basic-hierarchical-model-of-efficiency/).

This kind of model doesn't take into account how basketball actually works at the player level, so I'm pessimistic on the usefulness of this method. I think those checks you suggest would prove worthwhile to help answer this.

By doing that sort of stuff for more recent models I've built really helps to see if the model is actually modeling the underlying process well or not. It's really useful and, as you suggest in your book with Hill, isn't done often enough.