EARTHQUAKES AND MONEYBALL: EXPLAINING CASCADIA EARTHQUAKE PROBABILITIES TO STUDENTS AND THE PUBLIC USING BASEBALL ANALOGIES

Much recent media attention focuses on Cascadia's earthquake hazard. A widely cited magazine article starts "An earthquake will destroy a sizable portion of the coastal Northwest. The question is when." Stories include statements like "a massive earthquake is overdue", "in the next 50 years, there is a 1-in-10 chance a "really big one" will erupt," or "the odds of the big Cascadia earthquake happening in the next fifty years are roughly one in three." These lead students and the public to ask where the quoted probabilities come from and what they mean. These probability estimates involve two primary choices: what data are used to describe when past earthquakes happened and what models are used to forecast when future earthquakes will happen. Why different choices give very different estimates can be illustrated with simple analogies and examples, using people's familiarity with probabilities in sports. The data come from a 10,000-year record of large paleoearthquakes compiled from subsidence data on land and turbidites, offshore deposits recording submarine slope failure. The earthquakes seem to have happened in clusters of four or five events, separated by gaps. Earthquakes within a cluster occur more frequently and regularly than they do in the full record. Hence the next earthquake is much more likely if we assume that we are in the recent cluster, than if we assume that the cluster is over. A baseball analogy illustrates these ideas. The cluster issue is like deciding whether to assume that a baseball hitter's performance in the next game is better described by his lifetime record, or by the past few games, since he may be hitting unusually well or in a slump. Hence “the” probability that a batter will get a hit depends dramatically on assumptions made. The other big choice is whether to assume that the probability of an earthquake is constant with time, or is small immediately after one occurs and then grow with time. This is like whether to assume that a player's performance is the same from year to year, or changes over their career. Similarly, saying "the probability of an earthquake is N%" involves specifying the assumptions made. Different plausible assumptions yield a wide range of probability estimates. Hence as in sports, how to better predict future performance remains an important question.