@kriordan By default, the indicators use ID 0 values - those change as the candle develops, and thus you cant always see the cross in the final value.
Its usually best to use candle ID 1 on all cross conditions. You can find the ID setting from the "More settings" button.
@trader-philipps Yeah it surely depends on the pattern.
Its just that diagram talking about some mystical "recovery zone" and both ways trading, it looks very much like there is no downside to that strategy. There is always a downside.
@trader-philipps just to illustrate where the "* 1 usd" will come into play, and that there you can also change the dollar amount to "OrdersTotal() * 2", for example.
@dragonfly617 Mql4 function OrdersTotal() will return the number of trades, so you can make a block like this:
Note that the function will also count the pending orders.
Just place a sell stop at the level of buy SL, and vice versa?
But realize that you will always buy the top and sell the bottom, not very profitable.
@thefaunt said in Tried and proven strategy - can it be EA’d?:
@roar thanks for you efforts roar - it’s really appreciated. I’m new to this whole EA creation, so I’m in the process of learning fxdreema to create what I want, and also how to test it properly on MT5.
How did you do the optimisation? As currently I’m creating a new EA for each variable I want to modify and running the tests each time; which seems a lot more time consuming than what you’re doing!
Thanks again
Its no problem, I like to practise building systems.
To start optimizing, just set up the strategy tester for a normal backtest, and finally select Optimization: genetic algorithm.
Then go to your inputs-tab to set the optimized parameters and their ranges, just like I show a couple posts above.
@kop said in just asking:
@roar sure . but when we lose we lose $1 and when we win we win $3 . and take small profits up and down ??? partial close
If you take bigger profits than losses from the random market, it's just the probability distribution that changes.
In the case of $1 loss and $3 profit, your trades will lose 75% of the time and reach TP 25% of the time.
Expected profit: 0.25 x 3 - 0.75 x 1 = 0
Surely its possible to make a win in a random market, or at a roulette table, or throwing dices. Every win will eventually have an equal amount of losses, though.
The expected return is 0.
Well, 0 minus all trading costs.
Genetic optimization backtest of 10496 combinations will take some time, I'll let it run on EURUSD 2018-2020.
Edit: It found one profitable combination. $5 net profit with $260 drawdown
Someone might wonder, what if I used ma9 instead of ma8?
What if I used different expiration for the order?
What if I used different percentage for my TP?
What if I used different lookback period than 5 candles?
Here's all those scenarios, and all possible combinations of those scenarios:
@thefaunt Hi!
Trading the trend with moving averages is literally the oldest trick in the book. Nobody profits from it, especially on small timeframes.
You must be more clever than that, and I can guarantee you that you will not find anything but total BS from youtube. Its just to get views.
When scalping, you dont want to follow the trend - small timeframes very rarely have a continous trend and most of the time your sl will get hit.
I don't mean to be overly harsh or unfriendly, its just the reality of this business 
Anyway I made the EA just for the giggles: https://fxdreema.com/shared/AUYaS0NWb
Its in mql5.
I used mas 8, 13, 21 (all fibo numbers)
Open buy order when last candle low is between ma 8 and 13
Open sell order when last candle high is between ma 8 and 13
Entry at the 5-bar extreme, stop-loss at the opposite 5-bar extreme, take-profit 200% of SL.
Result?
A straight line down the drain:
@akulus You can generate random numbers with MathRand mql4 function. Here's an example that buys 20% of the time and sells 20% of the time:
https://fxdreema.com/shared/zqkptwoqb
If you use MathRand, remember to first set up the pseudorandom generator with function MathSrand.
https://docs.mql4.com/math/mathrand
https://docs.mql4.com/math/mathsrand
@traderline The little extra window on the right is for quickly modifying variables, without needing the actual "modify variables" block.
I increase the variable value by 1 on every new block (new value equals old value +1)
The V1 is just a shortcut method to refer to some variable, I could write "counter+1" just as well.
Here's an example how to create such "time windows":
https://fxdreema.com/shared/Oyixkzrod
@traderline So the time between 2 condition passes must be less than 6 candles?
@traderline
You mean this:
There are 2 conditions,
1st condition is active on candles 1,2,3,4,5,6
2nd condition only comes active on candle 5, and this is too late?
((highOf10 - lowOf10)/price)pow(1.5)
in mql4:
MathPow(((highOf10 - lowOf10)/price),1.5)
https://docs.mql4.com/math/mathpow
@biztet said in How to skip 1st trade signal?:
Much Better!
So i understand here that the Counter Reset is not really necessary, right?
Well, the system has to know when the "trade streak" ends and new one begins. If you dont reset it, the EA will skip first macd cross and then trade every other cross for the next million years. I guess that's not what youre after here?
I have another question. Still thinking of catching the news candle.
within 1sec/2sec from open candle, if the pips away from open price let say more than 5pips,
then we open trade.
Buy if the price > open price
Sell if the price < open price
and close trade on new candle.
By using the second filter and the counter blocks, this should work:
https://fxdreema.com/shared/FyO3ZKGod
The EA code is run on every new tick, again and again. Once per bar restricts the flow to one tick per bar.
Now, if all its above conditions are true, the counter block increases its internal counter (0,1,2,3,4....)
In this case, the counter block counts only 1 tick (one bar and one macd cross because our conditions) and passes the next time, because it has reached the number 1.
The reset block is just for cleaning up unwanted effects, because the counter doesnt care if there's a million bars difference between the count increases.
I don't know if my explanation helps any bit, I'm too tired to think straight lol