Statistical Word of the Week
Catastrophe Modeling Assistant
I saw this job posting a while ago, and, in my next life,
that's what I want to come back as - "catastrophe modeling assistant." In some jobs (network system administrator, travel and logistics manager, scheduler), when everything is going well you are unnoticed. Any departure from this state can only be bad news for you. And in the work world, unlike in Hollywood, bad publicity is, indeed, worse than no publicity. But, for a catastrophe modeler, as for the news media, disasters constitute intrinsic validation of your work. And potential disaster, rather than being an underlying source of dread and foreboding, is your very reason for being. Moreover, being an assistant to the catastrophe modeler would further insulate you from any direct responsibility without reducing your proximity to the excitement.
In some respects, catastrophe modeling is, of course, simply a "long tail" extension of the work done in risk simulation and modeling (in which we have several courses). It simply requires more data, or proxy data, or more guesswork. Fancier mathematical models of catastrophes have been around a long time, and one of their interesting aspects is the "stressed dog" scenario. This is the phrase used to describe models that have "cusps" where, in a given set of circumstances, the system being described can head off in radically different directions based on minor changes in those circumstances. A "stressed dog" might cower and whimper, or it might charge and attack.