look this strange and impressive result!

New comment
 

I used demo account from FXOPEN on one of my computer. after a long time development, I got an good result like this picture above. it looks good. later I open another demo account (new download) and test it on another computer, I got totally differernt result like this below :

does anybody can explain this. how can i get a reliable or true test result so I am sure my EA is working or not? thanks in advance.

 
MQL_MM wrote >>

I used demo account from FXOPEN on one of my computer. after a long time development, I got an good result like this picture above. it looks good. later I open another demo account (new download) and test it on another computer, I got totally differernt result like this below :

does anybody can explain this. how can i get a reliable or true test result so I am sure my EA is working or not? thanks in advance.

Any mismatched chart errors on report? Also check journal tab.

 
i think the broker could have not open order when it was needed with to much request and so thinks, or he gives you other market datas and your indicators calculate wrong or your EA has some codeing problems and must be optimize.
 
Did you use the same timeframe? What about the spread? It might use different values, then it would make sense..
 

thanks all.

but I have no clue about that. I used the same broker same timeframe and same setting but different computer. that is totally different result.

I thought I had been close to get a good one before I got the second test result.

 

Hi

Yes been there and have the T shirt. Check out my previous post for more suggestions.

https://forum.mql4.com/22079

Stick to live accounts for back testing to minimise the bad history data problems it costs nothing. The brokers don't care about data integrity on demos but on live accounts it is crucial.

 
Ruptor wrote >>

Hi

Yes been there and have the T shirt. Check out my previous post for more suggestions.

https://forum.mql4.com/22079

Stick to live accounts for back testing to minimise the bad history data problems it costs nothing. The brokers don't care about data integrity on demos but on live accounts it is crucial.

This is an excellent suggestion.

 

thanks Ruptor,your suggestion is to open a live a account, to do backtest on it then a reliable testing result will be obtained, isn't it? so have you really got a reliable backtest in this way finally? if you don't mind which broker are you with and how long time for M1 in history data?

thanks a lot for all guys

New comment
Reason: