Recent changes in catastrophe models have been disruptive, changing measures of expected catastrophe losses by significant percentages. Should changes like this be expected on a regular basis? We will briefly focus on decision making in the context of the models and the science as they currently are, not as we want or expect them to be.