Using only script-kiddie skills, it may be possible to handicap the outcome of decisions of national importance. This talk presents a walk-though of a project to make more accurate predictions of US Supreme Court case outcomes. That could be a useful thing, if you had something at stake. Conventional techniques for predicting outcomes rely on legal expertise and knowledge of the policy issues at stake in a case and the justices' voting records. Forget all that: we're going to see what we can do with perl and XML transcripts of oral arguments. It's only 20 minutes of your life, but it might equip you to astound your lawyer friends, or make some canny investments.