Statistical fine-tuning of a conservative System

Hello Developers: 

I am looking for a very experienced quant trader that can help me optimize my trading system. I am usually looking for longer-term partnership. 

Please read this task carfully and approach me with ideas so we can start a dialogue.

I have been trading reversal-setups with certain conditions for a long time. I use a very conservative martingale-approach, when it comes to positions. I start with one very small position and increase up to 3 positions, leaving at least 1 ADR apart to space out entries. This keep my breakeven and take-profit levels close. This approach is low risk, because I start very small and I stop after 3 positions. 

Now the task that is costing me many sleepless nights:

For the very rare instances that my trades do not work out with 3 positions, I am trying to develop a system which keeps my profits and avoids big losses. I am asking the following questions: 

- Do I close all trades at a certain % drawdown?

- Is it better to hedge part of the trade after?

- Do I close position 1 and keep adding until we see the desired reversal?

I need help with how to figure out which approach works best with my system. So it should be an active discussion and in the end we need to run some backtests and look at what the numbers say. 

I dont have my rules in an EA yet. We would need to discuss step by step how to work through this problme. 

In the end I want to have staistical confirmation on which to base my drawdown-management for the future since I am trading a mix of automatic and manual. 

Looking forward to your answers.