Hey pple...

I recently asked Tross to test CT 1.93b with 99% modeling quality and the results were CRAP: http://www.forex-tsd.com/76429-post23.html. It dumped the deposit.

Another guy asked the same about CT 1_9 R2.2 and got more CRAP:http://www.forex-tsd.com/76459-post27.html. Deposit was dumped too.

Lets note that Tross' datafeed isn't from any broker who use MT4, so the reliability isn't very high, but it was a 99% backtest anyways. A 90% backtest is theoricaly less reliable than one of those at 99%.

Even so, Aragorn got disappointed with his results on his live at IBFX. It seems that it ran there as CT did on Tross' 99% backtest. I also got a bad result on my demo account while Aragorn was having his bad one. But I blamed (and still believe on that) CT's bad performance to the year's end.

Despite all of this, a 99% backtest giving us consistent Crap results must be a warning to us. Perhaps 90% backtest wouldn't be reliable at all for CT.

My Demo test (only nov-dec at FXDD Server), Aragorn's Live test (only nov-dec IBFX Server** I might be wrong) and Tross' 99% backtest (Entire 2006 over Gain Capital Servers' data), SO FAR, has one thing in common: Bad Results.

Perhaps Aragorn's and my results are only a reflect of EUR.USD bad pattern for CT Logics and Tross' 99% backtest wouldn't be reliable for us, given it is from a different data feed. But the "Bad results" coincidence between them must be a warning for us on taking conclusions over 90% backtests.