over optimization should be 50% of the holy grail

first thing you read when search how to build a mechanical system, is that you should avoid over optimization and have an idea before start the optimization.

for me the first thing you should do is optimize everything.

whether system like ONS and FS works, then a simple optimization of MA period and day of week should spot these systems. Do the same with intraday and a few billions of system should be found.

now optmize if all parameters (for example if it will be MA crossing a TSF or a Linear Regression, etc) and whether mechanical systems really works then you should find trilions of systems.

"With four parameters I can fit an elephant, and with five I can make him wiggle his trunk"
- John von Neumann as quoted by Freeman Dyson.