New Trading System. To Heuristic or not to Heuristic

Greetings.

I have been working on two algo based trading strategies for a few months now; the end goal would be to develop an algo with definable edge for each of these market types; strong trending, sideways trending, range bound…both long and short etc…

One of the most joyous and simultaneously painfully annoying parts of the process is the sheer number of ways you could approach the problem; to achieve EV, to maximise EV/risk, should one go for a higher strike rate with a low signal count (better EV/Risk) or should one maximise overall profit by going for a lower strike rate with higher volume of trades.

if I run the algo across multiple currency pairs, Indices, Commodities, how to identify the most efficient set of instruments to apply at any given time, is compounding a net positive, if so should one compound at the instrument level or the overall account level.

How to maximise EV, Strike Rate, Signal Count in system fundamentals. How best to maximise profit with the most appropriate risk management tools.

So many possible lines of enquiry.

Maybe it is better to find a solution that works well but is not perfectly, 360 degrees optimal.

How long do you spend back testing, every combination, iteration of the variables that affect the fundamental Edge and risk management of your system; it would take at least 6 months to do it properly no.