The invention provides methods for optimizing a database management system process of a query. The methods are suitable for execution in a properly configured computer system. In an embodiment of the invention, a plurality of single column statistics are collected for a plurality of columns. This plurality of single column statistics provides an estimate of row counts and unique entry counts for a singe column operator. A preferred single column statistic is selected from the plurality of single column statistics according to a predetermined criteria. The preferred single column statistic is then stored. A selectivity estimate is then determined for predicates in the query using the preferred single column statistic. The selectivity estimate is used in optimizing processing of the query by the database management system.