Mmh... no. This algo could also work decently on SPY and DIA, but it's optimized for QQQ/MNQ. Any different combination other than MNQ-only tests worse in terms of recovery factor, which is the metric that I personally value most.
And with increasing capital allocated to the system it is going to reach critical mass , take inefficiency out until its barely covering electricity bills.
I was kidding, of course. There's absolutely no chance to reverse engineer anything from entry/exit numbers only.
Can learn something about the strategy if decent amount of entry/exit data. I sometimes set algo up to to take very few but high probability trades and look on chart to find flaws in historical data that may have caused the "market inefficiencys" to occur. Then make stronger data filters. At some point no flaws due to data problems. When you can jump with single click from trade to trade. By watching where computer found opportunity i became more informed TA trader. Seeing some similarities in data before entries is enough(but for me it works only if algo takes similar trades). If algo uses too many strategies its hard to know and compare all to form understanding.
I was kidding too (although not sure up to what extent). Both positions closed yesterday at 10062.75, +89$ and +282$ respectively. New position opened at 10068.
You said earlier low volatility was important to your trend following strategy , I just wondered how important?