SELECT COUNT(*) AS duplicates,
       field1,
       field2
FROM   some_table
GROUP BY field1, field2
HAVING duplicates > 2