ok, well, i just learned the value of using the expert properties function of the tester, it turns out i was using .01 lots instead of .1, (even though i thought the coder i used only went as low as .1) thanks for the feedback!


now there's several strategies i may have found out how to make work on the tester.


& of thanks!

__________________
"One's ability to accomplish his or her goals is limited only by the ingenuity of how one uses what he or she already knows."- Eaglehawk
REMEMBER, "Genius is nothing but a greater aptitude for patience." –Benjamin Franklin

____________________________________

Have a simple ea you just can't figure out how to code??? Odds are you can make it here. Just remember to push complete when you're done, NOT SAVE!!!
http://sufx.core.t3-ism.net/ExpertAdvisorBuilder

I'm having some issues with the Modelling Quality.

I was getting around 50% all the time with the data the program pulls in itself. So I googled it and found this link:
http://www.metatrader.info/book/print/27#67
Which basically is the same info posted at the top of this thread with lots of pictures included.

So I imported the data for EURUSD and GPDUSD. And converted it to the other times and it worked great. I was getting 90% quality on 'every tick'. WOW, what a difference. It's easy to get inaccurate great results with poor quality.

Well, here's the problem. I closed to program and when I reopened it, "History Center" shows everything I had imported and converted truncated at 50,000.
If I go to "Offline Chart" it still says 600,000+ bars.

So I go to the history\MoneyTec-Demo folder where it stores the info and import the hst file the program uses from there, I'm back at 600,000+ entries I had in the "History Center"

But now whatever it says, whenever I model anything, it always only gives me 25% model accuracy, no matter what!! All the same test I did before where I got 50% then 90% are now only 25%.
What is wrong here? My "Max bars in history" is set to the really huge number...

-Adam

SBFXv4b193

I had this problem.
Go to "Tools", "Options" and "Charts" and set "max bar in history" and "max bars in charts" to 9999999999999999999999 for example. Just type it.
Then open Offline and probable you have two the same files (one file EURUSD M1 and the other one is the same EURUSD M1 for example with different number of bars). So import the "good" file once again (from your Metatrader to the same your Metatrader) in the same way as you did it to get 90%.

I got it working again.
I switched to MetaTrader instead of S B F X.
Backtesting I'm always getting 90% except M1.
On M1 I always get 25% and it's a pink bar.
Is there no way to get better than that with M1? Since there is no smaller data to interpolate or whatnot?

-Adam

Hi!

I want to test if this strategy is good or can be improved:

Take the GBPUSD-price of 9 am CET, add 15 pips to this price and this is the long-entry. Short-entry is 9 am price - 15 pips. Target is 30 pips, Stop is also 30 pips.

Can somebody tell me how backtesting works with the Metatrader?
Is it possible to test automatically different values for the Target? Maybe 27 pips profit has a better overall-result than 30 pips... That is what I want to test with it.

Thank you for some help!

Marcus

I can't believe that nobody is able to answer this little question??

You can becktest automatically the EA only.
So it should be created EA and after that you can backtest it.
How to backtest see Codersguru lessons (for example here).

But you can backtest it manually on historical data.

http://www.metatrader.info/node/67

here is a good article telling you how to get good test data. When i first started programming EAs i would always get like 1million $$ in just 1 year lol. Once i got the real data in it was more like 1k in one year

Go to options (Ctrl-O) select the "Chart" tab. Fill 99999999999 in the "Max bars in history" and "Max bars in chart" . This will get u to 88-90%