The ratio of evidence between two competing models; quantifies how much the data favors one model over another.