“How do you fit a model to data it has never seen before?” asks Northern California Chapter ARCS Scholar Saahil Shenoy. Model accuracy is key to predicting near future events of climate, power use, and economics, thereby directing major decisions from disaster preparedness to interest rates. Supported by his ARCS Award, Shenoy pursues the timely field of statistics of extreme events within the Ph.D. program, Department of Physics, Stanford University. Many statistical models have relied upon familiar bell-shaped curves, but such models struggle to incorporate data points falling far outside what the normal distribution predicts. In a time when rare events might not be so rare, Shenoy uses exponential distributions which rise on one side toward infinity to include large but rare data points, making forecasts more useful. Shenoy recently built a model telling electricity providers how much power they can expect to need 24 hours in advance; it’s particularly relevant as electricity is sold on open markets one day ahead of time. As data science drives the accuracy of countless branches of study, the applications of Shenoy’s work to the field of research alone may strengthen the work of all of his fellow ARCS Scholars as well!