If the stats disregard the last trade (if the last trade never occurred)
EA Studio (as well as FSB Pro) takes into account all trades. The programs close trades when the closing conditions are satisfied. It is MetaQuotes design decision to make MT closing the current position at the end of the backtest. I don't think this is correct. However it is very easy to me to make such behavior in EA Studio. Then we will have complete match between the stats. I'll add this very soon.
...we would have a SQN of 17.91 over 377 trades, during which we never lost more than one trade in a row?
The SQN formula is written by Dr. Van Tharp (http://www.vantharp.com/tharp-concepts/sqn.asp) and implemented in EA Studio and in FSB Pro. I'm not discussing the formula, it is what it is.
Please note that SQN 17.91 is not a big deal. We can go a wayyyyy better. It was easy to fix the TP to 5 pips,to disallow SL and EA Studio easily generated a strategy with SQN = 90. The result is confirmed in MT4. Consecutive wins 113, consecutive losses 1. Unfortunately MT4 doesn't show SQN.
Of course, this is a complete curve-fitting.
I guess it would have been helpful to me if I would have examined the underlying strategy of the system to figure out what was going on. The win/loss ratio is even, so I guess that the TP is extremely small compared to the SL, the SL is huge (or non-existent and there is no exit strategy when losing?), the strategy trades at break-even when it comes back from a loss, and a break-even amount counts as a loss? But that would still require wins and break-evens to alternate for 377 trades.
I think your misunderstanding comes from this the strategy trades at break-even. There is no break-even. There is a Take Profit of 5 pips.
EA Studio makes what you say to it. When you set that you want it to find the best strategy with TP = 5 pips, it makes it.
If it sounds weird to you now, please leave it for 2-3 days.
If you think there is a bug in the program, please explain what do you think it is wrong and what do you think must be. We will analyze the problem and will find a solution.
You can also export the complete journal to a csv and you can make the stats calculations in Excel. If you find a difference, please report.