Clustering Factor
An Oracle optimizer statistic that approximates how closely index-key order follows the physical placement of rows in the table.
Two indexes with similar selectivity can still have very different table-access costs. When index order tracks physical row placement, adjacent entries tend to touch fewer table blocks; poor clustering can turn ROWID lookups into scattered block reads.
The clustering factor is not a perfect physical-distance measurement. It is one input to the optimizer's cost model, and it explains why selectivity alone is insufficient when reasoning about the engineering meaning of an index.