What type of EA

[seb 0]

Heyo, i've been working on that last project and i realize now it might be on the wrong side of EAs. I think it is intelligence learning type.
My project is an EA focusing on one indicator (MACD), taking as much information from it as possible(not really), instead of using many, possibly each one contradicting each other.
Theoretically there is almost an infinite amount of information given by an indicator. But the more you take in count to respect, the more you might miss opportunities when controlled by a bot, as, unlikely the brain it is not flexible at all. Here is my project, i thought you could tell me if there is a way to make it more flexible, or if i started on the wrong way on this one.
https://fxdreema.com/shared/hmnjAZjWd
Based on EUR/USD M15.

[AlphaOmega]

@seb-0 I have not looked at your EA yet but, here is my input from what I gathered in the 8 or so months I have been trying to create an EA... I used many different indicators and changed it to EA,s with a little added from what I was thinking the Indicators short.... and the more I tried to make the EA correct the less orders I got. The most perfect EA gave me like 9 orders a year.. Another EA gave me like 4000 orders in a year but the profit factor came out in the end the same than the EA,s that was good.......considering the risk factor
I think you are on to something.. NOT your EA... but the fact that you are sharing!!
If everyone in the forum put their minds together I think with the collective knowledge and luck we can create a profitable EA that everyone could benefit from... BUT...
Humans in general are selfish and find it very difficult to share... maybe they think there is not enough to go around?... my 2 cents
Thanks for your contribution and openess

[seb 0]

@alphaomega
I agree with you on the first part.
About sharing, event if it took me a bit of time, even if there are some blocks, and even if we collaborate, if the basics are bad this is not gonna give a profitable EA. I was more waiting for constructive critics because i think that, at this state my EA sucks ^^.

[AlphaOmega]

@seb-0 Well .. It sure looks very professional... I am not a programmer, far from it! Also I use mt4 so I cannot see what your EA is doing but looking at what you did I would not say you have a bad strategy...... Can you add this to your strategy and see if there is a positive difference?
https://fxdreema.com/shared/pLpzM1GUb

[roar]

My advice: dont try to find the perfect entry, or the perfect optimization parameters.

When I build an EA, I try to model the whole market so that the EA always has some opinion of the market state, just the confidence (and position target) is variable. This way you get the flexibility.

If you try to optimize single-indicator EA to buy/sell at fixed SL&TP a couple times every year, the metatrader optimization will pick up some favorable statistical noise for you. Useless. Always remember to test the strategy outside the optimization period, and try to get a good amount of trades because that provides some statistical certainty.

Here's an example of my thinking, in case someone wants to take a look. I'm not going to explain the whole system because frankly I can't remember lol. You will need netting-method mt5 though.
https://fxdreema.com/shared/iVW1zwFVc

[seb 0]

@alphaomega I don't think it fits sorry : (
@roar That actually is a lot of things to study, give me a bit of time !

[AlphaOmega]

@seb-0 I agree with Roar- cannot just use one indicator... you can incorporate BB and MA with mach my friend... they work good together

[seb 0]

So I've studied that a bit and I have 2 questions :

• in block 38, mid = (bbup+bbdn)/2
  mid isn't simply equal to "middle line"/"MA" ?
• in block 38, volgap = target + PositionGetDouble(POSITION_VOLUME)/lot1;
  lotB = lot1 + pctOfT/100volgaplot1
  Then whenever a buy position is gonna be opened, set the volume 3% HIGHER than initially if there is a sell position ; whenever an open position is gonna be opened, set the volume 3% LOWER than initially if there is already a buy position(correct me if i'm wrong).
  So you can have 2 opposite positions and you can have several positions in the same side if i'm right. Is that multiple positions strat, still netting method ?