A few months later, I've been teaching Anna and Luke and Will Ryan and others this rule as the "concrete-abstract pattern". Give a specific example with enough detail that the listener can visualize it as an image rather than as a proposition, and then describe it on the level of abstraction that explains what made it relevant. I.e., start with an application of Bayes's Theorem, then show the abstract equation that circumscribes what is or isn't an example of Bayes's Theorem.