Pattern robot

Pattern robot

I am not sure if it is allowed to link to Youtube. If not, let me know and I will remove it.

This is the strategy that I found interesting: (5) "The BEST Scalping Strategy For SMALL Forex Accounts! (75% Win Rate)" Proven 100 Trades - YouTube.

EDIT - I am editing post #1 frequently. I don't specify whenever I do it. So you have to reread it frequently.

So I started to make the robot. From version 6 it is showing pinbars only. And it is in general working better. I will explain about the wick sizes that can be adjusted later. For now just use those preset in the robot.

I will also make the suggested Engulfing some time in the coming future.

That´s really it! Very simple but a nice aid in the daily work as a manual trader.

I ad the code here. Everybody is allowed to take it and develop it further as long as they are publishing the MQ4 code here!!!

Please use this naming: Pattern_YourInitials_Version.

The trading rules are in the blue box. Picked up from the Youtube video. Attached is the robot as it is right now, the template I use as well as the indicators I use. They are all found at Forex Factory.

Screen will turn blue in case of a Long suggestion and Green for a Short. NOTE THAT YOU CANNOT JUST TRADE THOSE SUGGESTIONS. THEY ARE JUST HERE TO TELL YOU, THAT SOMETHING INTERESTING MIGHT BE GOING ON!!
You have to make your normal analyzing - allways!

A Pinbar is charactorised by having a long wick and a short wick. You can set those numbers in the settings. I have heard these figures (might not be correct):

WickLong must be minimum 33% of the size of the full bar from high to low.
WickShort not larger than 18% and not smaller than 3%.

For the Piunbar to work as a strategy, the must be no bars to the left that coses past the EMA50. How many?? I don't know, so I made it variable. It is called LeftBars in the settings.

There might still be minor faults in the code. Keep coming back to the trade to receive updates.

So this is the kind of signals that are the goal. Still need some programming til I get there.

And here is my screen. See how easy it is to identify interesting trades? Just go for the coloured charts.

Hi,

This system sounds interesting, but I am still having some doubts of entry points. Can you use some graph and explain the entry points for buy/sell. I maybe able to code something out of that.

Thanks!

Hi Fixer
Take a look at the video I link to right at the beginning of post #1. It is explained very fine.

When I come across some perfect setup, I will post them.

Be sure that you have version 7 of my robot - it is working well.

This chart I have very much been CHERRY PICKING, only to show the possible trades, when we trade a very trending chart. Note also 2 loosers.

Thanks! I saw the video and tried making an EA based on the same. It works with a small return on M15/M5, but it needs to be fine tuned. I will check your robot as well now and if mine works any better than what it is doing right now, I will share it.

Got this trade. If it brakes through the low of yesterday, then it will run. So I set break even when it gets down there and then I hope.

Note the Multi Time Frame bar in the right lower corner is nearly all red. I don't know how it meassures, but I trust it is right.

Super. Thank you.

My robot is not trading, only signaling. Look forward to see what you have made.

Thanks for your contribution @RiskFighter! I've been looking for strategies in smaller timeframes to complement my "portfolio". I will continue to follow your posts and perform tests asap. Good Work!

OK. I have stopped making my robots trade automatic, because they are way to unstable. Attached is a graph where I made this strategy trade with n*ATR. It is not useable. Result is the same on other pairs. This is TF M5. It gives way to many trades which again gives a large uncertainty.

I could get very fine results if I made it trade at chosen weekdays and time periodes. But then we are curvefitting and it does not work in real life.

BUT - the strategy is OK I think. It makes you trade the pull backs. Only does it also trade false pull backs. It will work if you trade it manually I think. And that is what I will do.

But it clearly needs som confluence indicator. I have allready attached the MTF at the chart. That helps, but it is not enough. So:

The tasks for the followers of this strategy is to try different ways of confluence to improve the manual trades of this strategy!! Don't sit on your hands. We can make this our best strategy if we work together!

I am able to get this result on EU on M5 from 2013-2020 with 0.1 Lot size. A lot of fine tuning needed, very unpredictable as of for now.

Some ideas:

1. Only trade Pinbars of a certain size. Because in flat market there are many small Pinbars that are fake.
2. Wait for a large confirmationbar after the Pinbar.
3. Tjek n bars to the left for not closing over EMA50. Because we should catch the touch of the EMA50, but not while the bars are closing many times in a flat market around EMA50.
4. Use n * ATR as SL. Because the instruments volatility is different.
5. Close the trade right away if a bar close over/under EMA50. This will reduce the losses.

I am not able to work on it before mid next week.

Thank you for your effort until now!

I would like the profit to be larger. One way is to eliminate as much fake signals as possible.

I tried fixing with things that I can manage, but seems like a dead end to me. I guess I am not in a position to optimise it further. Maybe will work on it again next week with a fresh set of mind. Till then, if you have any other ideas in mind then please share and I will try to code and see what I can do with that. I tried the other variant as well which was on other thread, but again, a dead end. I usually optimise with either a single trade/pair strategy and then try with with a martingale if not working with a single trade. It seem not working with either options. But then as I said, looking back at it next week with fresh set of mind may help. I am still working on the weekends as well for a few hours, so if you have any other ideas that can be coded, I am happy to be on the desk.

If you have tried the things that I mentioned in post 14 and it gave nothing, then I also believe that this is a dead end. And nobody else seems to have any ideas.

Introduce a divergence indicator for confluence

I think we gave up on this too early. I am busy on other systems right now, but I will take up this thread in the near future.