how come a line of my backtest has a profit x and a dd y....and when i run the single test the profit and dd is way different from the backtest line result? - page 2

New comment
 

Does your EA use indicators? Because almost every time I've seen those errors they come from indicators (and most times when testing with "open prices only"). Putting this line above in each indicator should solve the issue if that is the reason

#property tester_everytick_calculate

If that doesn't solve it, or you don't have access to the indicators, then try backtesting with all ticks. You can compare non visual and visual single backtests to verify if that works (a non visual backtest should give the same result as an optimization pass so you don't need to do a full optimization to test it. The correct result is most probably the visual one, however).

And if indicators aren't the reason, then the EA's code really has to be messed up to be able to give different results...

 
@Manuel Alejandro Cercos Perez #: Does your EA use indicators? Because almost every time I've seen those errors they come from indicators (and most times when testing with "open prices only"). Putting this line above in each indicator should solve the issue if that is the reason. If that doesn't solve it, or you don't have access to the indicators, then try backtesting with all ticks. You can compare non visual and visual single backtests to verify if that works (a non visual backtest should give the same result as an optimization pass so you don't need to do a full optimization to test it. The correct result is most probably the visual one, however). And if indicators aren't the reason, then the EA's code really has to be messed up to be able to give different results...

OP does not have access to the code ... post #4

 
Manuel Alejandro Cercos Perez #:

Does your EA use indicators? Because almost every time I've seen those errors they come from indicators (and most times when testing with "open prices only"). Putting this line above in each indicator should solve the issue if that is the reason

If that doesn't solve it, or you don't have access to the indicators, then try backtesting with all ticks. You can compare non visual and visual single backtests to verify if that works (a non visual backtest should give the same result as an optimization pass so you don't need to do a full optimization to test it. The correct result is most probably the visual one, however).

And if indicators aren't the reason, then the EA's code really has to be messed up to be able to give different results...

I don't see how this directive "tester_everytick_calculate" could explain the difference between an optimization pass and a single backtest without visual mode (as it's done when started from the optimization window). Do you have any reproducible example ?
 

what i observed is when the sl and tp are both higher than 3k points on the sl and tp (from 4k to 18k 22k )... the profit and dd of the results and single backtest doesnt match

and when the values are below 2k the values match....

im backtesting nas100 10 min and 5 min and 1 min

you guys think that is the issue? lots of points?

this is not weird for u guys too?


there is 3.7 k points on 1 average candle on nas100 10 min

is this even correct?



The most banal trading strategy - No loss coverage and no loss coverage in the market: Igor Zakharov (Part 2)
The most banal trading strategy - No loss coverage and no loss coverage in the market: Igor Zakharov (Part 2)
  • 2019.03.10
  • www.mql5.com
Not us, but you)) the strategy does not lead to large losses, for a simple reason. Without loss coverage (like martingale) the drawdown lasts for years. Just clarify, please, "no loss coverage" and "covered"
 

how come on the mt5 im testing nas100 10m... and i put ATR for the last bar its says like 12.... and o the bar measure is 1.200 points?


why its that huge difference?



 
tadeu23 #: how come on the mt5 im testing nas100 10m... and i put ATR for the last bar its says like 12.... and o the bar measure is 1.200 points? why its that huge difference?
Because ATR is in price range, not in points.
 
Fernando Carreiro #:
Because ATR is in price range, not in points.

ok but its accurate this?

nas100 M10 one average candle have 1.200 points?

look at this other example.... 2,37% = 35.300 points?

how can we backtest with thatkind of unit? its a huge interval of points



 
tadeu23 #:  how can we backtest with thatkind of unit? its a huge interval of points

There is nothing abnormal about that. There are many symbols with even larger ranges.

I think you need to do some research about basic trading principals, so that you can understand the diference between price, price range, point (size) and tick (size).

EDIT: And when including screenshots, include the price legend (scale on the right), so we can see what you are referring too.

 
Fernando Carreiro #:

There is nothing abnormal about that. There are many symbols with even larger ranges.

I think you need to do some research about basic trading principals, so that you can understand the diference between price, price range, point (size) and tick (size).

EDIT: And when including screenshots, include the price legend (scale on the right), so we can see what you are referring too.

here is a video that i did showing the problem

i hope this can be more clear than my explanations

https://www.dropbox.com/sh/dtr3c009yjsnny5/AABEYRgittOmg-_-ENdkCXxOa?dl=0

Brandon
Brandon
  • www.dropbox.com
Compartilhado com o Dropbox
 
tadeu23 #: here is a video that i did showing the problem. i hope this can be more clear than my explanations

There is no use in showing us videos or to continue this discussion.

Without source code we cannot analyse how the EA is working in order to explain the discrepancies.

99% percent of the time, the problem is the EA code, so contact the EA's author, and resolve the issue with them.

123
New comment