In this lecture from the Yale Quantum Institute Course on Quantum Error Correction, held at Yale University in July 2025 in partnership with North Carolina A&T, Aleksander Kubica addresses the decoding problem in quantum error correction, the task of inferring which errors occurred from measured syndrome data. The lecture presents the mapping between this decoding problem and models from statistical mechanics, a connection that provides both theoretical insight and practical decoding algorithms. It builds on the course's earlier sessions covering stabilizer codes and specific noise models for hardware platforms like cat qubits and Rydberg atoms.