For sequential inserts into a table or table instance with no secondary keys, INDEXCHK calculates the estimate assuming that all the tables pages (both data and indexes) are filled to their maximal value.
For random inserts, INDEXCHK calculates the estimate assuming that the top level of any index is filled to its maximal value and that all lower pages are sixty seven percent full. The result is only an estimate and may not correspond to the practical reality. Where a table is subject to random inserts and deletes, the estimate may be reduced further.
When secondary keys are defined, the estimate assumed that, even if the data is inserted in sequential order in regard to the primary key, the inserts are random in regard to any secondary keys. The secondary key providing the lowest number of rows is displayed.