Fractals, ZZ & the Pissing Dog | Page 9 | 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

  • #161
  • Jan 10, 2014 6:54pm Jan 10, 2014 6:54pm
Quoting burnssss
@ gg53 I must be doing something wrong. capture introduce and comment: {image} 1 - start to calculate from the red vertical line 2 - introduce fractal pattern [1] and according to probability, this pattern is correct 3 - in a length L = 5 bars gives me high probability in length with fractal pattern [1] 4 - step to determine "next bar" and "Current bar" and both give me 100% bearish, so I put bearish signal. 5 - Loop calculating from this signal, but this time looking bullish signal. 6 - In conclusion, where the second red arrow is where I appear...
burnssss,

The process, until you get a matching probabilities, is as follows (refering to your chart in post #126 (http://www.forexfactory.com/showthread.php?p=7201349#post7201349):
Fractals Length
--------- --------
[] --------- 6
[] --------- 7
[1] -------- 8
[1] -------- 9
[1] -------- 10
[10] ------- 11
[10] ------- 12
....
....
....
The above should be checked against a database that you know(!) that have a pattern of [.......1] in it.

You keep this process until you get a probabilities confirmation (both pattern and length).
When you get also the "next bar" in opposite direction prediction - mark your signal and continue from that signal (not from actual ZZ).
Exception to that will follow in next post (Adaptability issues).

G.
  • #162
  • Jan 10, 2014 7:09pm Jan 10, 2014 7:09pm
Quoting gg53
{quote} burnssss, The process, until you get a matching probabilities, is as follows (refering to your chart in post #126 (http://www.forexfactory.com/showthread.php?p=7201349#post7201349): Fractals Length --------- -------- [] --------- 6 [] --------- 7 [1] -------- 8 [1] -------- 9 [1] -------- 10 [10] ------- 11 [10] ------- 12 .... .... .... The above should be checked against a database that you know(!) that have a pattern of [.......1] in it. You keep this process until you get a probabilities confirmation (both pattern and length)....
okei think this straight.
When brands [] of length 6 and 7, these likely are not studied because we do not have fractals.
Is that correct?

-If the first swing is not predicted by the algorithm, we start to calculate from zzconfirmado?

Thank you very much and sorry for asking so much.
I am also trying, through the function of the stochastic and complex mathematical iterate a number of times that a value function for convergence at a certain point, it should be the turning point.
  • #163
  • Jan 10, 2014 7:09pm Jan 10, 2014 7:09pm
Quoting smallcat
{quote} Hi tashi168, I do not quite understand what do you mean with "2". But if you are referring to the "candleUpdown" and/or "fractal" on csv file, based on this: Format: [Pair] [TF] [candleTime] [open] [high] [low] [close] [candleUpDown] [Fractal] [zzLegId] [zzUpdown] [lastZzEndPrc] Then it means that : 1) if we have a "candleUpdown" = 2, it means the candle has a doji with the same length of upper and lower leg of the doji. What i do is to look for the previous candle, and change this to [0] or [1] depend on the previous candle. If previous...
Thank you smallcat !!
  • #164
  • Jan 10, 2014 7:15pm Jan 10, 2014 7:15pm
Quoting crodzilla
{quote} No. It is tricky if there are no fractals... like after a signal. You have to "guess" if the first fractal is going to be a "1" or a "0" and start database search. You must have at least one fractal to search the database. We are looking for zigzag end-point. ALL zigzags end in at least ONE fractal. If previous zigzag was a bottoming zigzag, you could use the opposite "1" fractal as the first fractal. Carl
Thank you Carl !!

@GG & Carl,
Another question: We get 4-13 length from SET2 to filter SET1, which becomes SET3. If the fractal length > 13, then we just get 13 length from the right. As GG said, he gets all fractal pattern (does not throw away any pattern), so how about the length < 4 ? Shall we "add" one/some "0" or "1" to the fractal pattern to get the fractal length = 4 ? What is the rule for it ?

Thanks
  • #165
  • Jan 10, 2014 7:51pm Jan 10, 2014 7:51pm
Quoting Magix
{quote} I totally understand the concept and logic...was just a simple question. If anybody had taken and used the ZZ or Fractal pattern and only traded them, their success would be limited if they experienced any at all with reasonable Money Management in mind. I was asking a simple question based on what appeared to be a fairly generic pattern. Having worked with several strategies and indicators, the ones that seem to pick absolute tops and absolute bottoms typically repaint and going by observation alone is why I asked my question. Look at semafor...
Magix,

1. The algo is trying to predict SWING points, not ZZ end points. True, For that purpose we are using previous ZZ internal pattern "knowledge".
2. Our produced signals NEVER repaint. Look at the algo instructions.
3. We are not competing with either ZZ, Semaphor or fractals indicators.
4. If you look at past results (actual real time charts) of the predictions - you will notice profitable results even with poor money management...
5. ALL posted charts are real-time and not historical cherry-picking ones.
6. Most currently active participants had their doubt and yet they joined us here from previous thread because of live proof - which you aparently missed or diden't know where to look for it.

There might be other ways of doing it, even better ones or simpler ones, or more profitable ones - and we are more than happy to learn them.

G.
  • #166
  • Jan 10, 2014 7:55pm Jan 10, 2014 7:55pm
Starry & Magix,

Please... stop this.
Relate to the subject and not to the person.

Thanks

G.
  • #167
  • Edited 10:04pm Jan 10, 2014 8:12pm | Edited 10:04pm
Quoting gg53
{quote} burnssss, The process, until you get a matching probabilities, is as follows (refering to your chart in post #126 (http://www.forexfactory.com/showthread.php?p=7201349#post7201349): Fractals Length --------- -------- [] --------- 6 [] --------- 7 [1] -------- 8 [1] -------- 9 [1] -------- 10 [10] ------- 11 [10] ------- 12 .... .... .... The above should be checked against a database that you know(!) that have a pattern of [.......1] in it. You keep this process until you get a probabilities confirmation (both pattern and length)....
Wow great, thanks a lot G ....
Now i know where my logic was wrong .... i did not wait for next 2 candles to define the fractal in my database. Yes, the fractal appeared 2 candles late .... i did not include it in my algo ....

Edit: One more question G. How about the last fractal [1] at that picture ? It is appear at candle #14, but this candle is belongs to next zz leg.
How we calculate fractals pattern [101] ? Is it with zz length = 14 , although the candle #14 belongs to next zz leg ?

Edit again: I mean like this G. After we get fractal pattern [10] at candle #12, then if the probability calculation tells us that we must wait, at candle #13
we are trying to compare fractal pattern [100] and [101] to database (with zz length=13). If the result is also tells us to wait, then at candle #14
we try again to compare fractal pattern [100] and [101] with zz length = 14.
So, we don't care that the candle #13 & #14 is actually belongs to the next zz leg. Is that right ?

Thanks,
-smallcat
  • #168
  • Jan 10, 2014 9:53pm Jan 10, 2014 9:53pm
Quoting crodzilla
So, I accidentally had my algo indicator attached to the Mini-Dow yesterday and today. To my surprise, great signals!!! Please note, that the algo history/database is based on EURUSD 5-min bars. Yet the zigzag (real-time), fractals, and bar information are coming from this 1-Hour chart of the DOW!! Kind of shows you that all markets may be similar in fractal construction, no matter the timeframe. @GG53, I remember you talked about a universal indicator... I think this confirms that it is possible. Carl {image}
Thank you Carl ...
  • #169
  • Jan 11, 2014 12:50am Jan 11, 2014 12:50am
Hi GG53,

I'm testing my algo based on some charts you published begin of november, f.i. 5 november. Are these charts still valid w.r.t. to your latest algo.
If not could you publish some recent charts for comparision.

BR
Huub
It is not important how often you fall down, but how often you get up again
  • #170
  • Jan 11, 2014 12:56am Jan 11, 2014 12:56am
Happy New Year folks. I hope its an immensely pleasant and fulfilling 2014 for everyone.

GG53, its wonderful to see this thread. Thank you for starting it.

It seems there are more recent participants joining this discussion on pattern matching and GG's algo...I thought to give some sort of 'synopsis' and an idea of what, I think, the older readers (which came from a previous thread) are looking to re-create. I wont bore anyone with words though, take a look at the images below. The pink and blue arrows depict real time signals of GG's indicator. (oh, and the algo doesn't repaint).

Regards.

----------------------------------------------------
Attached Images
  • #171
  • Jan 11, 2014 1:02am Jan 11, 2014 1:02am
And these..
Attached Images
  • #172
  • Jan 11, 2014 2:39am Jan 11, 2014 2:39am
Quoting emonts
Hi GG53, I'm testing my algo based on some charts you published begin of november, f.i. 5 november. Are these charts still valid w.r.t. to your latest algo. If not could you publish some recent charts for comparision. BR Huub
emonts,

Yes, they are still valid.

G.
  • #173
  • Edited 9:02am Jan 11, 2014 2:47am | Edited 9:02am
Quoting Thoughts
And these.. {image} {image} {image} {image} {image} {image} {image} {image} {image} {image}
Thoughts,

There was about 6 screen shots taken on Friday close predicting next week open (6/6).
Can you post those separately with next week results?

Thanks.

G.
  • #174
  • Jan 11, 2014 2:48am Jan 11, 2014 2:48am
Quoting Thoughts
And these.. {image} {image} {image} {image} {image} {image} {image} {image} {image} {image}
pls share this indicator?
  • #175
  • Jan 11, 2014 2:49am Jan 11, 2014 2:49am
@ GG53

Don't get me wrong on this one.

Why don't you share what you got so far? Isn't it taking a lot more time explaining and answering every question again? If you share what you got we can all work from there and we know we all got the same.

You shared what you got in "gg53 killing zone" as well so I don't quite understand why you won't do it here?

Regards,
Dennis
  • #176
  • Jan 11, 2014 2:53am Jan 11, 2014 2:53am
Quoting Dennis189
@ GG53 Don't get me wrong on this one. Why don't you share what you got so far? Isn't it taking a lot more time explaining and answering every question again? If you share what you got we can all work from there and we know we all got the same. You shared what you got in "gg53 killing zone" as well so I don't quite understand why you won't do it here? Regards, Dennis

mr gg is a good man and i know one day he will share even if is not the code but indicator.am doing manual calculation and i dont think is good to be given an indicator blindly without knowing what it does.be patient but in the mean time is good to learn what is done by the indicator
  • #177
  • Edited 5:20am Jan 11, 2014 3:08am | Edited 5:20am
Quoting Dennis189
@ GG53 Don't get me wrong on this one. Why don't you share what you got so far? Isn't it taking a lot more time explaining and answering every question again? If you share what you got we can all work from there and we know we all got the same. You shared what you got in "gg53 killing zone" as well so I don't quite understand why you won't do it here? Regards, Dennis
I was writing about that GG. Share everything without any deeper info what the indi does. give me everything right now.. lol.

Dennis,

I think GG already shared almost everything with us, only 1 step explanation missing. And I did not know/or only few other people in FF forum like GG, who always answering all questions, even if we asked tousend times.

regards
  • #178
  • Jan 11, 2014 3:40am Jan 11, 2014 3:40am
Well, "Adaptability & Prediction errors" is on it's way according to gg. So let's be patient.
  • #179
  • Jan 11, 2014 4:11am Jan 11, 2014 4:11am
Quoting smallcat
{quote} Wow great, thanks a lot G .... Now i know where my logic was wrong .... i did not wait for next 2 candles to define the fractal in my database. Yes, the fractal appeared 2 candles late .... i did not include it in my algo .... Edit: One more question G. How about the last fractal [1] at that picture ? It is appear at candle #14, but this candle is belongs to next zz leg. How we calculate fractals pattern [101] ? Is it with zz length = 14 , although the candle #14 belongs to next zz leg ? Edit again: I mean like this G. After we...
That's what I understood, but applying "next bar" on the sail 14, Predicted no swing.
So my question is:
1 - we must do if the swing is not predicted?
2 - referring to post 126. If not predict this swing and start from the assumption confirmed last zz, zz the following has a fractal in the end, that applying "next bar" not predicted. Is this normal?. This is where adaptability come?.

thank you very much

EDIT: who wants indicator has all the instructions in post 1
  • #180
  • Edited 4:27am Jan 11, 2014 4:15am | Edited 4:27am
Quoting Dennis189
@ GG53 Don't get me wrong on this one. Why don't you share what you got so far? Isn't it taking a lot more time explaining and answering every question again? ..
Quoting trexy123
{quote} pls share this indicator?
There is a note in Post 1: "**Note: This is a "work in progress" - so don't ask for finished, working indicator/EA."
Quoting gg53
.. Let us bring everyone to that level by explaining the methods and logic done so far. ... G.
You have this work in progress shared here, in an very organized manner, and also you have the availability of gg53 to respond to pertinent questions.
But you want money handed on a platter, under various arguments, and you are disregarding his request...
Maybe you should make some efforts, read the method, do some work, and ask questions only to gain further understanding - do some minimum effort yourself

There is a nice melody titled "My way" played by Julio Iglesias. Since this is gg53's way of sharing his work then we all should respect that.
0 traders viewing now