The degree of a node $u \in V$, denoted $d(u)$, is the most straightforward node-level statistic: it simply counts the number of edges incident to that node. Despite its simplicity, node degree is often one of the most informative features used in traditional machine learning models applied to node-level tasks.