Understanding how the Earth’s crust deforms under stress is important for assessing seismic hazards and plate boundary dynamics. One way scientists study this is by analyzing small earthquakes to estimate the direction and type of fault movement, known as focal mechanism solutions. However, these estimates rely on correctly identifying the first movement of seismic waves, which can be difficult, especially for small earthquakes. In this study, we developed a new method for determining fault movement called POSE, which uses a rule-based approach grounded in statistics and information theory. We tested this method against two artificial intelligence–based techniques in California and Tibet, two tectonically active regions. All three methods performed similarly in California, but POSE identified more usable data points and was more robust to noisy signals. In Southeastern Tibet, POSE produced significantly more fault solutions than the other methods and revealed detailed stress patterns that match independent GPS measurements. Our results show that POSE is a reliable and widely applicable tool that can improve how we map underground stress and fault behavior, especially using small earthquakes that are often overlooked.