AlligatorEx.

New comment
 
I am posting an EA based on Alligatora, and test reports. I should tell right away that the code is made from scraps and I am a newbie myself. I am not afraid to use it, I will be glad if you will give me some tips for improvement. I would be very thankful for any suggestions and improvement. I will tell you right away that I have not mastered flashing as I am not yet familiar with the modes of optimization in the tester. I have tested it on Alpari. I have tested it with Alpari.
Files:
alligatorex.zip  92 kb
 
I would like to add a selection of TakeProfit, StopLoss, TrallingStop.
 
not much of a profit
 
I understand, I would like to know if such an EA has a place in life if you add a customizable TP and SL, I had TP written in the code as approximate.
 
I'm thinking of adding an osMA and/or Stohastic to it for a more accurate EA trade exit.
 

I'm going to give you some advice, because I wanted to try Williams' ideas, but I came to the conclusion that I had to remove a lot from his system in order to understand the effectiveness of the idea "in its pure form". But I have not got round to it.

I do not know why you set Take Profit and open the trend again when you reach it. It's better to try trailing, imho. By the way, why do you set the buy TP at 100 ppt and sell TP at 80 ppt?

2) Exits - better not a rollover but exit when a Close or two Closees have closed behind the middle line - red, you can also try green.

3 - placements - where you have takeprofits, you may want to scale in trend, this is option 1, with the size of the distance to scale in may be optimized. Variant 2 - we may fractionate by the breakdown of the last fractal. Variant 3 - we may refill by OsMA, stopchastics. Or by AC; they are approximately the same.

4. I would not change the entry and do not filter it.

5) Then you may try to put a stop loss at initial order placement.

This is how I planned to do it. Publish it if you do it. The main thing is not to try to implement Williams' system as it is described, this has already been done, for example

https://forum.mql4.com/ru/4951/page29

https://www.mql5.com/ru/code/10231

 
ZZZEROXXX:

I'm going to give you some advice, because I wanted to try Williams' ideas myself, but I came to the conclusion that I had to take a lot out of his system to understand the effectiveness of the idea "in its pure form". But I have not got round to it.

I do not know why you set Take Profit and open the trend again when you reach it. It's better to try trailing, imho. By the way, why do you set the buy TP at 100 ppt and sell TP at 80 ppt?

2) Exits - better not a rollover but exit when a Close or two Closees have closed behind the middle line - red, you can also try green.

3 - placements - there where you have takeprofits, you may want to scale in trend, this is option 1, with the size of the distance to scale in may be optimized. Variant 2 - we may fractionate by the breakdown of the last fractal. Variant 3 - we may refill by OsMA, stopchastics. Or by AC; they are approximately the same.

4. I would not change the entry and do not filter it.

5) Then you may try to put a stop loss at initial order placement.

This is how I planned to do it. Publish it if you do it. The main thing is not to try to implement Williams' system as it is described, this has already been done, for example

https://forum.mql4.com/ru/4951/page29

https://www.mql5.com/ru/code/10231


Thanks so much for the tips!!!! I also think that Williams needs a file to grind. Looking at the indicator, I came to a conclusion that the steps can be shifted back (3,0,-2) and build not by MedianPrice, but by Close. As a result the entry point comes earlier and lines are drawn right next to each other. Maybe I'm wrong and I cannot compete with powerful Williams, but as they say it is better to see on the chart than to read in the code.

As for your advices on updates of EA, I am very grateful to you and will try to implement them at weekend. I am writing EAs relatively recently and I did not manage to learn some of tricks to make my EAs work better.

After I finish the work I will surely post my Expert Advisor and reports for your judgment.

 
I would recommend that you start with the default alligator parameters to see the pure version, so to speak, so as to have something to compare it with, and then, after all these improvements, try to use the alligator parameters. And an earlier entry does not mean better when it comes to mash-ups.
 
ZZZEROXXX:
First, I would recommend to leave the alligator with default parameters to see the algorithm in its pure form so to speak and then, after all these improvements, try to use the alligator's parameters. And an earlier entry does not mean better when it comes to mash-ups.


On a slightly modified EA the profit margin is much lower with default settings, it is really risky to experiment with the exit, but imho no one has found the perfect point yet experimenting is the best way to find it.

At the moment I added adjustable parameters Alligatora and TakeProfit and StopLoss. I hope to add Trawl by the end of the weekend and fix the bug with early order closing as the trend continues.

 
Dizet_02:


On a slightly modified EA, the profit margin is much lower at the default setting


What did you rework in it that leads to a worse result? The TP, the SL, or anything else?
 
ZZZEROXXX:

What has been tweaked in it that leads to worse results? TP, SL, or something else?

As I said before with default Alligator settings entry and exit are slightly delayed, which is sensible for small TFs, I added an option for Alligator's parameters and in comparison the result was not in favour of the default settings. By moving Alligatora a bit back i.e. reducing the draw step, entry and exit are at a more optimal moment.
12345678...10
New comment
Reason: