I have done some test with indices and the formula for calculating the value in the RDB$STATISTICS-column is really 1/count of unique values. So, is it correct to say, the less the value in the RDB$STATISTICS-Column is, the better is the selectivity of the index ?


Guido