Database optimization often leans heavily on traditional techniques like indexing and query rewriting, but sometimes creativity can lead to better performance. This article explores unconventional optimization methods in PostgreSQL, highlighting how small changes can significantly improve query efficiency. For instance, a common oversight in querying with check constraints can lead to unnecessary full table scans. By enabling the `constraint_exclusion` parameter, PostgreSQL can skip these scans when queries include conditions that can’t possibly return results. This adjustment can save considerable time and resources, especially in reporting environments where users frequently make mistakes. Another example involves optimizing queries for a sales table by using a function-based index instead of a traditional B-Tree index. While adding a B-Tree index on a timestamp column reduced execution time from about 627 milliseconds to 187 milliseconds, it came with a hefty size of 214 MB. This highlights a trade-off between performance and storage costs. Instead of defaulting to conventional indexing, the article suggests considering lower cardinality indexes that can provide similar query performance with less overhead. By rethinking the approach to optimization, developers can achieve better efficiency without incurring unnecessary costs.