I try to understand your thought process, but I don't.

Why would you develop a system, knowing on beforehand that it will clearly over-fit on your in-sample data?
Knowing how your system behaves on out-of-sample data should be key before going live.

Last edited by Grant; 02/05/22 00:36.