Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Your and My Lessons Learned Losing and Gaining in Trading 5 replies

things you learned from forex and improved your life 12 replies

The most important lesson’s I’ve learned trading Forex, so far... 15 replies

Attachments: What I learned after losing $300,000 trading forex
Exit Attachments

What I learned after losing $300,000 trading forex

Quoting ryuryu
{quote} You absolutely right! Dividing on 5 because there are 5 steps. My mistake. So it doesn't matter how many days back to check. Taking data from last 5 or 10 or 222 days back is just to get more accurate average. But divide always on 5. {quote} Yes, sure. But I don't need the information how price was moving open/close... I need only 5 things and this approach works great for that purpose: 1. Avoid entries within -1 level - zero - +1 level because they are useless and just taking margin/balance 2. Avoid entries > 3 level and < 3 level because...
OK, now I see that you decided to divide by 5 simply because you wanted to have 5 equal intervals over that range.

But remember, the normal distribution curve is not linear, whereas you have decided to use linear divisions. If you want to create divisions based on the normal distribution curve, then one way is to use "probability paper". That is, the y-axis is scaled in units of %probability while the x-axis has a linear scale.

You can do normal probability plots in Excel if you have the Analysis Toolpak activated.
To free Gazans of Hamas, use whatever it takes.
Quoting DrDave
{quote} But remember, the normal distribution curve is not linear
Agree! This 5 zones are strange. Why 5? Why not 6 or 21? So I think this is like that: if price is moving in a rangle -1 - +1 then it is flat and we will do nothing. If it touch +1 and trend is up then X% that it will touch level 2, X% for 3 and so on.

I have found this: that formula (or just 0.25 step) is not good. There are no >85% success rate. In fact the formula around 0.25*1.6 give >x3 range. Because we expand it up from zero and down and also mulpitly on 1.6.
If I use multiplier 0.5 it gives exact 100% of X average. And here we have much better probalility. But there are some issues with breakthrous and fake shadows.
So if mult 5 then we have 10 zones (5up and 5 down) equal to X evarage back.

I'm trying to create a strategy for that. For example the rules are:
If at Tokio open 200/100/50 ema are in correct order then buy stop at 1st level and tp at second. Same for shorts but reverse. Also if there are previous day high or low between entry and TP then use prev day as TP or do not enter at all.
If someone can create simple mt5 EA or script to count the probability if will be perfect. I can provide detailed instructions.

P.S> there is new fixed version of indicator attached. Previous was buggy. This one use daily open as zero.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot2.png Size: 198 KB
Attached File(s)
File Type: ex5 WF_Zones_v2.0.ex5   41 KB | 381 downloads
Observer effect
Quoting ryuryu
{quote} This 5 zones are strange. Why 5? Why not 6 or 21? So I think this is like that: if price is moving in a rangle -1 - +1 then it is flat and we will do nothing. If it touch +1 and trend is up then X% that it will touch level 2, X% for 3 and so on. I have found this: that formula (or just 0.25 step) is not good. There are no >85% success rate. In fact the formula around 0.25 give x3 range. Becasue we expand it up from zero and down and also mulpitly on 1.6. If I use multiplier 0.5 it gives exact 100% of X average. And here we have much better...
I understand what your intention is. I simply suggested that rather than using a linear progression for your 5 steps (or whatever number of steps you decide to have), base the 5 steps on the normal distribution curve instead of linear.

So, if you plot your values on probability paper, then draw the linear regression line. if your data are approximately a normal distribution, it will make a very good linear fit. THEN from this plot, make your 5 equal divisions on the y-axis and look at what the % value is, and use THESE % values for your "interesting" price levels.
To free Gazans of Hamas, use whatever it takes.
Quoting ryuryu
{quote} This 5 zones are strange. Why 5? Why not 6 or 21? So I think this is like that: if price is moving in a rangle -1 - +1 then it is flat and we will do nothing. If it touch +1 and trend is up then X% that it will touch level 2, X% for 3 and so on. I have found this: that formula (or just 0.25 step) is not good. There are no >85% success rate. In fact the formula around 0.25 give x3 range. Becasue we expand it up from zero and down and also mulpitly on 1.6. If I use multiplier 0.5 it gives exact 100% of X average. And here we have much better...
Using 5 divisions is not some kind of voodoo. It just happens to be a division that you personally find useful for gaining some information.

Others may prefer 8, 9, 20, 21, 100, whatever, and have perfectly logical explanations for the choice. Even if someone made the choice by picking a random number within some range, there is a logical explanation for how he arrived at that number.
To free Gazans of Hamas, use whatever it takes.
Quoting ryuryu
{quote} This 5 zones are strange. Why 5? Why not 6 or 21? So I think this is like that: if price is moving in a rangle -1 - +1 then it is flat and we will do nothing. If it touch +1 and trend is up then X% that it will touch level 2, X% for 3 and so on. I have found this: that formula (or just 0.25 step) is not good. There are no >85% success rate. In fact the formula around 0.25 give x3 range. Becasue we expand it up from zero and down and also mulpitly on 1.6. If I use multiplier 0.5 it gives exact 100% of X average. And here we have much better...
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 193 KB

Your chart has some other interesting information.

Start from the left with the green arrow.
What was the percent change from the open (your white zero line) to the open of the following day?
Was this open higher of lower?
If you simply place an order shortly after the new order in the direction based on if it was higher or lower than the prior day's close, how many % levels are reached? Is it VERY likely to reach the first level (red line)? If yes, then could you make a strategy of entering in that direction, putting in an appropriate SL, then when that first level is reached do something such as take profit, or move SL up to some level such as breakeven or 50% of your first level, etc.?

When deciding on an initial SL, look at how price moves to reach your first "interesting" level, and how many levels that it retraces on its way to the first level in your trade direction.

I think that every for day on your chart except the last 2 days that your target level (red line) is reached before the the red line on the opposite side is reached.

You have only a small sample here, so is this typical behavior or unusual behavior? You need a lot more data to determine that.
To free Gazans of Hamas, use whatever it takes.
Quoting DrDave
{quote}Your chart has some other interesting information. Start from the left with the green arrow. What was the percent change from the open (your white zero line) to the open of the following day? Was this open higher of lower?
Dashed lines are previous day high/low. White is zero. So next day open was higher than previous one.

Quoting DrDave
{quote}If you simply place an order shortly after the new order in the direction based on if it was higher or lower than the prior day's close, how many % levels are reached? Is it VERY likely to reach the first level (red line)? If yes, then could you make a strategy of entering in that direction, putting in an appropriate SL, then when that first level is reached do something such as take profit, or move SL up to some level such as breakeven or 50% of your first level, etc.? When deciding on an initial SL, look at how price moves to reach your...
Very interesting! Yes, if next day open is higher/lower it has very high probability to hit first level (red) in that direction. To calculate how many levels will be reached I have to create EA to get some data. But quick look at the charts shows that first level will be touched not always. Check few days back on gbpusd - there were 3 days in a row then daily open was lower but first red was not touched.

Also the problem here I see is very high spreads in first hour of the day, so it is better to check for spread and once it become tight place the order.

If you check the last day (Friday) on eurusd you will see that day was opened higher than previous one. And first red level was touched. But it was very dangerous. But if check for MA order it was incorrect, it was not "long". It was "wrong angle sort". So in this case it is better not to trade at all.

BTW Monday will open lower and MA order will be for short.

So I prefer check the direction using MAs, place entry at first level and exit at second. That will me more accuarate.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot2.png Size: 324 KB
Observer effect
Double take when I saw this on the front page! I really want this to be true, 20k, 50 even 100k, but fucking over a quarter MM is fucking awesome. I really want this to be true cuz this is the kinda shitt that keeps me coming back to forexfactory...fucking A. I had to read twice, I saw Albert Einstein, kelly klarkson or whoever, money management...64point ratio and not one mention of a stop loss. I have to hand it to this guy for taking it like a champ. This guy needs to start a EUR forum of his own next. People giving the guy pats on the ass and then the forum naturally morphs into something about everyone's bs technicals...Damn I love this shitt!

OP, keep on rocking man...remember you're not profitable till you dig yourself out that hole... Damn, I hope this was real...
To sum up:
For example we are using 15m TF. Then first we need to place 3 EMA (50, 100, 200). Then check the direction at 22:00 previous day. Then check it at 02:00 current day. If they are in correct order both at check time then calculate the distance. For longs it must be higher and lower for shorts. Then we must check previous day zero, it must be lower than current for longs and higher for shorts. If all match then we can place stop limit order at 1 level and tp at 2 level. One more: if previous day HL is lower/higher than entry or TP then do not enter at all.
Multiplier is 0.5, bigger number = less probability to hit the target.
P.S. as CeeSo said it is time sensitive approach, so if at 17:00 there targets not hit then remove all pending orders and wait for next day setup.

I have provided sreenshot. First entry is OK for short, because at check time MAs was in correct order with pretty big distance between. Previous zero was higher, previous low was lower than entry/TP.
Second entry was good in case of MAs and previous day zero, but previous day low was at the point of entry and higher than TP. So it is dangerous to enter here.
Third entry is totally incorrect, MAs was wrong, previous zero was lower (must he higher). And as you see SL will be hit first.

As an example: eurgbp will not be tradable this Monday because prev day low will be higher than possible TP.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 570 KB
Observer effect
I can`t say anything but a big thank-you mate... Studied your comments during August, then generalise your quite superb view in my own system, now have got very successful trades recent 3 weeks... far from perfect honestly! I m up 7% in my balance just for 3 trades during the time. Thank you so much again for sharing this wonderful thread here. awesome!
Thanks for your story. Negative experiences are also important, people learn better from mistakes. The main thing is to draw the right conclusions after.

I have created little EA, it will place entry at for example level 1 and -1, stop at zero and tp at 2 and -2.
This is EU stats for 2020.

1. 0.25% static levels
2,3,4,5 - on screen
Attached Images
Observer effect
Quoting ryuryu
I have created little EA, it will place entry at for example level 1 and -1, stop at zero and tp at 2 and -2. This is EU stats for 2020. 1. 0.25% static levels 2,3,4,5 - on screen {image} {image} {image} {image} {image}
Cool! I like the result 4...
Did you use all parameters as explained in your post
https://www.forexfactory.com/thread/...8#post13186358
Did you check for other pairs?
Congrats!
Attached File(s)
File Type: ex5 WF_Zones_EA_v1.1.ex5   168 KB | 296 downloads

Wait a little I'll attach the EA here. It has some issues I need to fix them. I have tested another pairs they are different. If try all 28 - loss with any settings.
About parameters: yes, but without MA checks. I'll add them later.

update: EA attached, try it yourself. i must be attache at daily TF (and strategy tester at daily TF too, I don't know why, this is not very good for testing, may be I'll change it later). All the settings are pretty simple. Feel free to ask if any questions.
Observer effect
If you have 25 years old. Why are you traiding? What is a reason? What analize do you use?
Thanks for sharing. I also have the same feeling.
I have found one thing about this strategy. It can hit level and reverse. GBPUSD for example act like this.
So it is better to use moving average to enter, not the price. I'll code my EA accordingly, so you can choose place pending orders or use market order using ma. I'll share it here to let you test the things.
Observer effect
Quoting Ah-Ong
Here is the indicator calculates the statistic of percentage gain for Open to High & Low. The result may slightly different for different brokers. Let me know if the calculation wrong. {image}New update version at https://www.forexfactory.com/thread/...1#post13139681
thanks for this indicator I just joined FF to ask you if you can write an indicator that calculates probability of the next day move based on previous day move.

for example if pday moves +0.5% the next day will have n% chance of moving +0.5% , m% chance of moving +0.25% , ...

to be more clear I mean just like your indicator it should finds days that move 0.5% and analyses the next bar.

thank you.
Many of you have asked several times about who my mentor is and I have decided to share it in this thread and then I am leaving FF factory for good. I started the thread to help people and it turned out that it only distracted me and hurt my trading. So this is my final contribution to FF and I wish you all the very best of trades
Inserted Video
Quoting CeeSo
So this is my final contribution to FF and I wish you all the very best of trades
Thank you CeeSo so much! Green days forever! Your simple idea was last brick in the wall
Observer effect
Quoting CeeSo
Many of you have asked several times about who my mentor is and I have decided to share it in this thread and then I am leaving FF factory for good. I started the thread to help people and it turned out that it only distracted me and hurt my trading. So this is my final contribution to FF and I wish you all the very best of trades https://youtu.be/IRPEgt5tXbU
Or maybe this is just you or you get a few bucks promoting him, trying to sell some coaching? Huh?
Anything above the line is an uptrend, below the line is a downtrend.
0 traders viewing now