When you say 'simulated trading' - how specifically did you simulate? The ninja built in simulator? ninja backtest/optimize/walk forward? market replay? Each of these are different
sorry about the blow up. i think it's like poker....the entire objective is to get you to fund an account. then make money off your commissions, and hopefully you don't blow up too soon. maybe sue them for having such a flawed sim platform, if it does indeed say things that are inaccurate. but i bet they covered their behind. no different than me telling you how to make millions, but with no guarantees. you give me 10,000, i teach you nothing, and you blow up. well, someone made money.......me. the game is rigged!!! imagine if you worked for a big firm and controlled billions....you would have the power. now you are just the sheep.
Won't NT's "default" setting on fills (as opposed to "liberal") give you the same behaviour? Aren't stop order fills governed by the "slippage" parameter? i.e. it's in the hands of the tester to determine what behaviour he/she wants ... Yes, that's a good point; IB sim account will be a good simulator of a live IB account. Perhaps the principle applies whatever broker you use, i.e. a MB Trading sim account is best simulator of a MB Trading live account, etc ... i.e. the point is that the NT algorithm can not realistically represent all possible environments ... Yes, I have seen this too. There is no substitute for detailed observation of what's happening. Automated does not equal unattended.
Unfortunately not, Liberal or Default or whatever just applies to the backtest fill-type, Sim101 & MarketReplay are totally different and cannot be changed/fixed. For backtest, the Default fill-type is not even able to guarantee the slippage requested: this needs to be fixed by each trader (I have a "BetterThanDefault" fill-type for that very reason). Sim101 & MarketReplay are doing their own independent & wrong fills - of course in a slow market, their fills are correct, but not so in fast-market conditions (any news release, even a simple stop-gunning can lead to fantasy fills). All that being said, I trade live automated systems on Ninja and get pretty much exact replication of their live performance when backtesting them (only slippage does vary). I am not bashing Ninja, just reporting my own experience with it and what it takes to work around its issues.
Yes, I am concerned that the game is rigged! I don't mind a fair playing field, I can compete in that. But if the "game is rigged" and my trades are being falsely represented (or rigged) I do have a problem with that.