Optimize with Every Tick or 1 Minute OHLC

 

I was testing this new EA with 1 Min OHLC and the results are really bad. Just tried to ran it with every tick and it became super profitable.

There is something I noticed. My other EAs are using M5 data and there is almost no difference using 1 Min OHLC and Every Tick

But this new EA is using H1 data. Why is there such an issue? 

 
doshur:

I was testing this new EA with 1 Min OHLC and the results are really bad. Just tried to ran it with every tick and it became super profitable.

There is something I noticed. My other EAs are using M5 data and there is almost no difference using 1 Min OHLC and Every Tick

But this new EA is using H1 data. Why is there such an issue? 

Because your EA depends on ticks. See this topic Strategy Tester mode (1 minute OHLC vs every tick)

All (not yet) about Strategy Tester, Optimization and Cloud
All (not yet) about Strategy Tester, Optimization and Cloud
  • www.mql5.com
How ticks are generated : The Algorithm of Ticks’ Generation.
 

I getting a little confused because all my EA use [0] meaning current bar

some are on m5 some are on h1

but most in fact all except one run 1 minute ohlc without much difference from every tick

only this particular EA needs to run on every tick mode. I'm happy that I try every tick mode, almost threw the EA out of the window.

The optimization time is killing me though 

 
doshur:

I getting a little confused because all my EA use [0] meaning current bar

some are on m5 some are on h1

but most in fact all except one run 1 minute ohlc without much difference from every tick

only this particular EA needs to run on every tick mode. I'm happy that I try every tick mode, almost threw the EA out of the window.

The optimization time is killing me though 

Doshur, don't you use small stoploss/takeprofit values for this EA or some different approach to define this values?

If so, in my opinion this could be another reason.
 
figurelli:
Doshur, don't you use small stoploss/takeprofit values for this EA or some different approach to define this values?

If so, in my opinion this could be another reason.
At least  more than 50 pips for tp and sl. I just found out one more EA of mine earn an extra 100% when using every tick mode.
 
doshur:

I getting a little confused because all my EA use [0] meaning current bar

some are on m5 some are on h1

but most in fact all except one run 1 minute ohlc without much difference from every tick

only this particular EA needs to run on every tick mode. I'm happy that I try every tick mode, almost threw the EA out of the window.

The optimization time is killing me though 

Every tick mode generates a lot more ticks on each bars. If your EA gives different results that can only means that it depends on ticks. We can't give you a more precise answer as we don't have the code.

 
angevoyageur:

Every tick mode generates a lot more ticks on each bars. If your EA gives different results that can only means that it depends on ticks. We can't give you a more precise answer as we don't have the code.

I understand, but is there any general guidelines?

80% of my EA seems to work ok in 1 min OHLC, the 20% coding style does not differ that much just that the conditions checking is different... 

 
doshur:

I understand, but is there any general guidelines?

80% of my EA seems to work ok in 1 min OHLC, the 20% coding style does not differ that much just that the conditions checking is different... 

I do agree with Angevoyageur, but a simple guideline would be code a 1 minute filter at OnTick() beginning, so all your code would be more aligned with OHLC information.

But as you stated before, if some of them are more profitable now, this would be a hard decision ;-)
 
doshur:

I understand, but is there any general guidelines?

80% of my EA seems to work ok in 1 min OHLC, the 20% coding style does not differ that much just that the conditions checking is different... 

If you can post a sample EA for each category (80/20) the we can try to identify the issue.
 
figurelli:
Doshur, don't you use small stoploss/takeprofit values for this EA or some different approach to define this values?

If so, in my opinion this could be another reason.
How to do the filter?
 
angevoyageur:
If you can post a sample EA for each category (80/20) the we can try to identify the issue.

I'm running more test on optimization

but the cloud is giving me problems... should the amount I spent in the cloud be refunded?