Fractals, ZZ & the Pissing Dog | Page 42 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

ECN && STP && Scalping && Hedging - Broker Review - Oct 2009 24 replies

Alert on Fractals (MT) 2 replies

Zigzag & Fractals 10 replies

EA wanted for sound alert depending on fractals condition 2 replies

Fractals Edge Fx 3 replies

Attachments: Fractals, ZZ & the Pissing Dog
Exit Attachments

Fractals, ZZ & the Pissing Dog

  • #821
  • Aug 9, 2014 3:32am Aug 9, 2014 3:32am
There probably is a correct algorithm to find optimized values but I am not much into R or mathlab.
I tossed the entire next bar direction sets away and tried to predict the last fractals instead. The results are reasonable good even though there are some big errors in between. I moved away from trying to predict the next bar and suddenly the next bar direction is predicted very accurately.

I will look more closely into it and after evaluation trying to trade binaries before I sort out the other major flaws in the algorithm. Does anyone have a broker suggestion? I really want to stay away from bucket shops. Are there any reputable binary exchanges, even if they required a certain starting capital?
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Chart1.png Size: 50 KB
Click to Enlarge Name: Chart2.png Size: 44 KB
Click to Enlarge Name: chart3.png Size: 37 KB
Click to Enlarge Name: Chart4.png Size: 43 KB
Click to Enlarge Name: Chart5.png Size: 37 KB
Click to Enlarge Name: Chart6.png Size: 37 KB
  • #822
  • Aug 9, 2014 4:11am Aug 9, 2014 4:11am
Quoting Kilian19
There probably is a correct algorithm to find optimized values but I am not much into R or mathlab. I tossed the entire next bar direction sets away and tried to predict the last fractals instead. The results are reasonable good even though there are some big errors in between. I moved away from trying to predict the next bar and suddenly the next bar direction is predicted very accurately. I will look more closely into it and after evaluation trying to trade binaries before I sort out the other major flaws in the algorithm. Does anyone have a broker...
Nice picture Killan19, you are near. Do you still use Step 1 & 2 with its sub-sequence calculation, but in Step 3 - instead of searching on "Next Bar in opposite direction" , you are looking for "Last Fractal in opposite of ZZ direction" ? Can you give some more information about this ?
  • #823
  • Aug 9, 2014 4:54am Aug 9, 2014 4:54am
Quoting Kilian19
There probably is a correct algorithm to find optimized values but I am not much into R or mathlab. I tossed the entire next bar direction sets away and tried to predict the last fractals instead. The results are reasonable good even though there are some big errors in between. I moved away from trying to predict the next bar and suddenly the next bar direction is predicted very accurately. I will look more closely into it and after evaluation trying to trade binaries before I sort out the other major flaws in the algorithm. Does anyone have a broker...
Hi Killian,

What do all the different colours mean on your chart? Eg what are the arrows and boxes?

For binary options remember that you have to predict the next bar direction and also make sure that the next bar covers the spread eg if you use EURUSD and predict the next bar up you need to make sure that the next bar is up and > 1 pips greater then the last close price to cover the spread. I use Dukascopy and have managed to automate their Web Platform but the $5K account minimum may be a bit too steep for a start. Anyway they have a free Binary Options contest using the same platform so you can use that to test your theory and win real cash to open a real account.
  • #824
  • Edited 5:36am Aug 9, 2014 5:21am | Edited 5:36am
Quoting beachloup
{quote} Hi Killian, What do all the different colours mean on your chart? Eg what are the arrows and boxes? For binary options remember that you have to predict the next bar direction and also make sure that the next bar covers the spread eg if you use EURUSD and predict the next bar up you need to make sure that the next bar is up and > 1 pips greater then the last close price to cover the spread. I use Dukascopy and have managed to automate their Web Platform but the $5K account minimum may be a bit too steep for a start. Anyway they have a free...

I actually have 4 colors on the chart.
The red arrow means that the signal will be down starting next bar and blue vice versa (drawn at the end of the bar).

I mark the actual fractal of every leg I try to predict (cyan box) which is 1 bar before or after the predicted point. If I am off I just shift the start of the next prediction to the box. This way I don't have to worry about me having the starting fractal in the next leg or missing a bar or 2 for the next signal.

I had a problem with the computation of consecutive signals once 1 signal got out of synch. Shifting the start of the leg back to the real start after the fractal appeared solved it partially.

In the first picture you see the first blue arrow not having a box as there is no corresponding fractal close by.
In the first picture you see the second blue arrow not having a box but that's just a minor bug in the code which happens when predicted legs are only 1 bar long.
In the first picture at the highest high there is a box and actually a red arrow on top of the box which didn't make it into the picture

I also got code which will get rid of additional fractals at the beginning if the prediction is not correct but I disabled it at the moment. This would draw an orange box on the chart. I also created a dashboard which shows the 3 signals pattern match, length match and fractal prediction separately. I might post a picture later.


But if you look closely at the chart outright trading the signals would end up in a loss/be if you take spread into account, but the next bar direction seems to fit pretty well.

Regarding binaries: I have written 2 indicators who create a "volume" and pip histogram for every day separately. The movement every day of the week is slightly different and I will only pick out the good times. It comes close to a bell curve. I don't really trust binary brokers that much moving to 15 minute bars might help to sort out some issues. I don't have a problem depositing the 5k as long as I can trade small and withdraw the amount. I will definitely take a look at dukascopy.

Quote
Do you still use Step 1 & 2 with its sub-sequence calculation, but in Step 3 - instead of searching on "Next Bar in opposite direction" , you are looking for "Last Fractal in opposite of ZZ direction" ? Can you give some more information about this ?

Yes I still use step 1 and step 2. I can talk a little bit about fractals in the future.
  • #825
  • Aug 9, 2014 6:03am Aug 9, 2014 6:03am
Hi Everyone,

Sorry for “disappearing” for such a long time.

If you remember, from few of my last posts, I was quite pissed by MT4 600 build forcing us to change and modify a lot – without us having much option or saying on that matter.
So I decided to write my own trading platform, using configurable API’s for brokers that support open, free API. There are several of those. The platform supports MQ4 code import and C#.
Now we have 4 Beta testers, and awaiting their comments/bug lists.

No, it’s not for public distribution.

The reason for this post is this:
Suppose you have a GREAT indicator or system that gives you accurate entry signal.
…and you still can’t make consistent profit with it… Why?

Recent academic research on financial data shows that this is not enough!
An entry signal MUST be supported by Volatility, Momentum, Volume, Session time, etc. in order to be profitable and not turn around against you.

How is all that connected?
Well, the new platform can add a factor for each of those parameters to every indicator that exist on the trading platform - meaning that the entry signal STRENGTH can be increased or decreased based on overall market stat. This can be done in plain MT4, too.

More on that will be on a separate thread, within 2 weeks.

G.
  • #826
  • Edited 9:10am Aug 9, 2014 7:47am | Edited 9:10am
Quoting gg53
Hi Everyone, Sorry for “disappearing” for such a long time. If you remember, from few of my last posts, I was quite pissed by MT4 600 build forcing us to change and modify a lot – without us having much option or saying on that matter. So I decided to write my own trading platform, using configurable API’s for brokers that support open, free API. There are several of those. The platform supports MQ4 code import and C#. Now we have 4 Beta testers, and awaiting their comments/bug lists. No, it’s not for public distribution. The reason for this post...
Nice to see you again GG !!!
I have some questions :

a) So far i know, actually we can use the OLD MT4, may be the build 509 (http://www.mediafire.com/download/lfc6tbiwfm6fai2/Trend.zip) , which is still available for download. Yes, it is not belongs to our broker. But at least, we can still use it for getting our signal, and then trading like usual using our broker's MT4 build 6xx. Some other information on how to block Auto update (it will stay at build 509) & downgrade to build 509 is here: (note: i never try this downgrade for long time, just for testing it once) http://4xtrader.net/mt4-build-600-why-you-should-not-update-it/

b) May i know about "....using configurable API’s for brokers that support open, free API...." . If i understand you correctly, you try to create a "similar" of MT4 ? This is very interesting .. great idea ... Is it compatible for all of MT4 brokers account ? I mean, using this "Own MT4 platform", we can connect to all of our MT4 accounts at different brokers, with just changing our account number & password (and some other setting like URL / IP , etc ...) ???

edit: Ok, i will wait for 2 weeks ...

Thanks
  • #827
  • Aug 9, 2014 10:10am Aug 9, 2014 10:10am
Quoting giovanni4000
{quote} Nice to see you again GG !!! I have some questions : a) So far i know, actually we can use the OLD MT4, may be the build 509 (http://www.mediafire.com/download/lfc6tbiwfm6fai2/Trend.zip) , which is still available for download. Yes, it is not belongs to our broker. But at least, we can still use it for getting our signal, and then trading like usual using our broker's MT4 build 6xx. Some other information on how to block Auto update (it will stay at build 509) & downgrade to build 509 is here: (note: i never try this downgrade...
giovanni4000,

a) Yes, we can use older versions (not supported by many brokers) and we can prevent "AutoUpdate". But... Do you know for how long?
b) Some brokers provide an open API for their data stream and buy/sell commands. Our platform can connect to those brokers datafeed and trading commands. So we can connect to those brokers. MT4 datafeed is a prprietry datafeed and we are not interested connecting to those datafeeds.

The coming thread will be mostly on supporting market data which will enhance the "strength" of any indicator (standard or proprietry) that provide "entry signal".
G.
  • #828
  • Aug 9, 2014 10:14am Aug 9, 2014 10:14am
Quoting gg53
{quote} giovanni4000, a) Yes, we can use older versions (not supported by many brokers) and we can prevent "AutoUpdate". But... Do you know for how long? b) Some brokers provide an open API for their data stream and buy/sell commands. Our platform can connect to those brokers datafeed and trading commands. So we can connect to those brokers. MT4 datafeed is a prprietry datafeed and we are not interested connecting to those datafeeds. The coming thread will be mostly on supporting market data which will enhance the "strength" of any indicator (standard...
Great, thank you very much GG !!
  • #829
  • Aug 9, 2014 10:31am Aug 9, 2014 10:31am
Kilian,

For dukascopy, they have a starter pack at 100 $ minimum deposit.

http://www.dukascopy.com/europe/english/home/

I don't know for the conditions, if you must be european, etc.

Throw an eye !
  • #830
  • Aug 10, 2014 9:51am Aug 10, 2014 9:51am
Quoting Madmoney
{quote} {quote} You guys are right, it is not all in order. The golden nuggets are scattered all over the place, but I have read the old and new thread over 25+ times and I am confident of what needs to be done (thanks to gg53, Crodzilla, burnss, Kilian19 and many others) it's just my time constraints. I'm developing this algo to run in C/C++ via DLL to MQ4. The first most in important thing is to make sure the data you have is CLEAN and RELIABLE. This is the free tool I use to get my RAW data - http://www.strategyquant.com/tickdatadownloader/#programs...
Thanks Madmoney. Some of us have read the old & new (this) thread more than 10 times. And some has made more than 20 versions, but can not get good result. GG said that some of us have successfully built "right" system. Jimsterk is right, it seems easy, but in fact not. Can you give me some information about "creating DLL with C/C++" ??
  • #831
  • Aug 15, 2014 1:07pm Aug 15, 2014 1:07pm
Quoting beachloup
Thanks Carl. I'm still confused by a lot of things. The immediate issue I face now is that I am struggling to get the zz end point. I have the real time zz so far and the fractals it contains so far. I keep looking at my historic zzs and they tell me my leg should be longer (which is correct). After a few more fractals I reach the point where my fractal match is no longer in the historic collection of zzs. For example, my current fractals are 10110011. My next is 1. I have 215 historic zzs with a longer length zz and fractal length, and 29 with...
Did you read the section on error corrections?

Yea, it get complicated from here on in...

Carl
  • #832
  • Aug 19, 2014 3:45am Aug 19, 2014 3:45am
Quoting gg53
Hi Everyone, Sorry for “disappearing” for such a long time. If you remember, from few of my last posts, I was quite pissed by MT4 600 build forcing us to change and modify a lot – without us having much option or saying on that matter. So I decided to write my own trading platform, using configurable API’s for brokers that support open, free API. There are several of those. The platform supports MQ4 code import and C#. Now we have 4 Beta testers, and awaiting their comments/bug lists. No, it’s not for public distribution. The reason for this post...
Hi gg53
I am quite excited by this and keep checking all the time. Two weeks can not go fast enough
cheers
hannele
  • #833
  • Aug 27, 2014 7:36pm Aug 27, 2014 7:36pm
Did you guys bookmark the post# where gg53 said there's something about fractals we are not seeing? And he's not telling us. Can you point me to it?
  • #834
  • Aug 29, 2014 1:44pm Aug 29, 2014 1:44pm
Quoting summicron
Did you guys bookmark the post# where gg53 said there's something about fractals we are not seeing? And he's not telling us. Can you point me to it?

summicron,

Post #645 - Page 33 - http://www.forexfactory.com/showthread.php?p=7444223#post7444223

G.
  • #835
  • Aug 29, 2014 9:57pm Aug 29, 2014 9:57pm
Lots of information here.... I admit I am kinda lost, but I don't mind running through all the 42 pages (goodness only 42) and try to catch on what u guys are really doing before I comment. Any key posts I should look for if anyone can help me out? (Other than the 1st page). I have my own ZZ work, so I think I would catch on fairly easily soon.
  • #836
  • Aug 29, 2014 10:13pm Aug 29, 2014 10:13pm
Quoting Relativity
Lots of information here.... I admit I am kinda lost, but I don't mind running through all the 42 pages (goodness only 42) and try to catch on what u guys are really doing before I comment. Any key posts I should look for if anyone can help me out? (Other than the 1st page). I have my own ZZ work, so I think I would catch on fairly easily soon.
Rel, I think gg53 is going to start a new thread "soon" (see post 825)... may be worth waiting for that?
Otherwise, the key posts to start with are those made by gg53.
I'm sure you'll do some good stuff with this Rel
  • #837
  • Aug 29, 2014 10:22pm Aug 29, 2014 10:22pm
Quoting 4xbones
{quote} Rel, I think gg53 is going to start a new thread "soon" (see post 825)... may be worth waiting for that? Otherwise, the key posts to start with are those made by gg53. I'm sure you'll do some good stuff with this Rel
Thank you. I will see what I can find. I have a personal problem with fractals thou; as much as I like MTF analysis, I would love to see a consistent and simple proof that fractals exist. Currently I have a nonstandard approach towards MTF analysis that is in the works, but I do not mind giving the concept of fractals one last try. I ponder that perhaps I misunderstood the concept?
  • #838
  • Aug 30, 2014 12:40am Aug 30, 2014 12:40am
Quoting 4xbones
{quote} Rel, I think gg53 is going to start a new thread "soon" (see post 825)... may be worth waiting for that? Otherwise, the key posts to start with are those made by gg53. I'm sure you'll do some good stuff with this Rel
4xbones,

The new thread will be on a different subject.

G.
  • #839
  • Aug 30, 2014 1:11am Aug 30, 2014 1:11am
Quoting gg53
{quote} 4xbones, The new thread will be on a different subject. G.
Gotcha... looking forward to it.
  • #840
  • Aug 30, 2014 7:02am Aug 30, 2014 7:02am
This is how I interpret this thread. Just my honest comments and I hope I am not derailing this thread in any manner.

The foundation of the system lies in the chart data and the indicators used. This comes before everything else, as later the chart data and the indicators will provide more 'metadata' (in the form of SET1, SET2 and SET3). Later, logic via a model is applied to the metadata to seek out edges/probalilities.

The crux of the matter as I see it is that the foundation requires some examination. The justification is that the better the foundation, the better the metadata and hence accurate edges/probabilities.

Chart data;
I've noticed that time based candles are used and there was an attempt to use Renko as chart data.

The Fractals indicator;
http://ta.mql4.com/indicators/bills/fractal
http://www.tradingfives.com/articles/elliott-wave-fractals.html

The ZZ indicator;
http://ta.mql4.com/indicators/trends/zigzag
http://www.mql5.com/en/code/7796

My question; why these 2 indicators? My experience with these chart data and indicators is that they have problems reflecting what price is really doing. I am not questioning the intent of use of these indicators; on an objective level, we all want to look for extreme points of price to enter + exit. I am questioning the formula of these indicators and how was the decision made to accept them as part of the overall foundation. Personally, I would not dare to use something I do not understand sufficiently.

However, I like the logic via a model. It is very well organized and such effort makes things easier much later. I do it myself as well so I know where all this is coming from. I was only surprised that no one seemed to question the foundation and went on straight to assume the foundation is all right.

A side note; Bill Williams Fractal's, as good as it sounds as it is, IMO is not a good reflection of Benoit Mandelbrot's ideal of Fractals. It took Bill Williams a long time to concede that his system is not suitable for trading forex (http://masterforex-v.su/forex001002002.htm). My personal experience is that it is just another indicator that relies on chart data; the chart data is assumed to be time based candles. If the chart data has 'problems', the indicator will have 'problems'. I avoided the indicator because the justification of the formula is weak; why 2 bars? Why not 3? Or even 5? That led me to use Barros swings for my ZZs instead, as its formula is more justified.

Moreever, I cannot find simple statistics to prove that the market is of fractal nature; I had to do it myself. (http://www.forexfactory.com/showthread.php?p=7700911#post7700911; the statistics for the 'higher timeframe' matches the 'lower priceframe'). I find its better to do things this way, as I will be aware that I did not leave any stone unturned. In trading, to do that, IMO is to open up risk that could had been prevented and calculated in advanced.

Hope I wasn't too pessimistic, I just wanted to understand in depth the decisions made in this thread.
1 trader viewing now