When running logic every tick - only looking at open trades and staying away from history trades - the backtesting suffers and takes longer.
Did you optimize the code in your EA ? are you doing things every tick that you only need to do once per bar ?
Read this thread: https://www.mql5.com/en/forum/144240
When using "control points" in backtesting - what happens?
What disadvantages creep in?
Are entry trades approximated or exit points or both?
What else gets approximated?
If on 15min and backtesting "every tick" - that is a lot of ticks and processing.
Possibly when using "control points" - code logic runs smoothed out to each 1 min bar - 15 times - open of 15x1 min bars within the 15min bar.
Or 3x5min bars within 15min bar.
Comparing results of both - the entries and exits are identical - the only thing that seems to change is the report. (but is trading the open only so not sure if is constant)
Using "control points" looks inviting but will i regret it?
Did you optimize the code in your EA ? are you doing things every tick that you only need to do once per bar ?
Read this thread: https://www.mql5.com/en/forum/144240
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
When running logic every tick - only looking at open trades and staying away from history trades - the backtesting suffers and takes longer.
If you move that logic to a DLL (probably c# or freepascal or c++ sample in mt4passing either each tick or an array[] - something efficient ) - well ............ will there be an improvement?
If there is an improvement - will there be a speed advantage if all the EA logic - bar the shell - is moved there?
Is anyone building their own DLL and will they share the speed advantages resulting when used in the optimiser?