Interesting. btw, I added that 'NextLevelSL_On' switch at the time, in the hope that it would give a rough indication that the filtering I was attempting was indeed providing an 'edge' of sorts - with it enabled, a blown account can't happen, but you can be 'nibbled' into bankruptcy!

The main problem with this is that while you can severely filter your trades, I could see no real 'edge' to doing so - instead of this filtering, try selecting the trades at random, opening a similar number for the period in question - you will get a similar result as probably 70-80% of all trades would have been profitable anyway, and within the max. levels set. (whereas probably 95% would 'eventually' have proved profitable after possibly severe drawdowns) Also, the PF is unacceptedly low, which is why I gave up on it...

One final point, though I do not know exactly how MT deals with timeframes during backtesting, I'm pretty sure that unless an EA specifically trades EOB, (10point3 doesn't) selecting higher timeframes for backtesting is less accurate as you are dealing with interpolated prices...