Cardinality estimation in relational databases is a well-studied problem, but graph databases rarely follow the uniformity and independence assumptions that these estimators make, and therefore lose precision.