Fractals, ZZ & the Pissing Dog | Page 5 | 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

  • #81
  • Jan 7, 2014 6:10pm Jan 7, 2014 6:10pm
I spit my soda out my nose laughing when I read the indicator comment. LOL

Quoting gg53
{quote} nondisclosure00, That's the spirit.! I wish everyone leave planet Earth and come to our planet, Utopia. Maybe we should create an indicator ($49.97) to show them the right way? ;-} G.
  • #82
  • Jan 7, 2014 8:20pm Jan 7, 2014 8:20pm
Quoting gg53
{quote} tashi168, At the very begining, when we have no fractals, we check SET2 only for fractals pattern - ithe probabilities of it being longer. No need to check total length until this condition is met. Probabilities calculations description is on its way... and will be posted soon. G.
Quoting gg53
How-to: Calculating Probabilities Added to post #1 Probabilities (DOC & PDF formats) - by crodzilla - http://www.forexfactory.com/crodzilla Thanks, carl ! G.
Thank you G & Carl.
  • #83
  • Jan 7, 2014 8:24pm Jan 7, 2014 8:24pm
It looks like we are popular on other places too:
"Prediction of turning points based on ZZ on fractal calc (http://www.stevehopwoodforex.com/phpBB3/viewtopic.php?f=27&t=3210&start=50&hilit=gg53)"
In SteveHopwoodForex forums... ;-}

G.
  • #84
  • Jan 8, 2014 12:46am Jan 8, 2014 12:46am
Hi carl,

Thank you very much for your compact description regarding probability.
I have one question.

I was comparing your fractal pattern occurrences against my own ones.
I noticed some are equal or very close to equal, other differ completely.
Anyway I have to check my sets again.

But I also see in your examples the following
Example 1, page one. For the fractal pattern <101> you find 509 exact matches.
Example 2, page two. For the fractal pattern <101> you find 599 pattern matches.

Same pattern but different occurrences.
There is only a difference in description 509 exact pattern versus 599 pattern matches.

Are this the same queries on the same database, or do I miss something.
Some clarification is very much appreciated.

Best regards,

Huub
It is not important how often you fall down, but how often you get up again
  • #85
  • Jan 8, 2014 1:18am Jan 8, 2014 1:18am
Database & MQ4


Database usage.

The SQLite database and tables are accessible via standard SQL statements from within MT4 using the sqlite3_wrapper.dll see the site http://github.com/Shmuma/sqlite3-mt4-wrapper (http://github.com/Shmuma/sqlite3-mt4-wrapper) for more details as I will not help with this part.

Other tools.

The free version of this SQLite tool allows data visualisation and testing of sql statements http://www.sqliteexpert.com/ (http://www.sqliteexpert.com/) Note: images above taken from screen grabs of this tool.

Thanks, go4it! (http://www.forexfactory.com/go4it)

G.
  • #86
  • Jan 8, 2014 2:37am Jan 8, 2014 2:37am
Quoting Magix
{quote} Where GG has started to compile a list of possible trade outcomes, I have taken a single 5 candle trade pattern, which is based on the Fractal Indicator in Meta4. I have used stochastics to define ranges of overbought and over sold and used the Fractal 5 candle trade pattern to trigger the trades. The base level of the script is a cost averaging, add on increment EA, that is trigger on grid based. Meaning, that in order for a subsequent trade to be placed, a minimum grid requirement and the trigger of OBOS on stochs plus the Fractal pattern...
Yes, very good idea!
  • #87
  • Jan 8, 2014 3:41am Jan 8, 2014 3:41am
Gg53 Crodzilla thank you very much for the file and probability calculation. I'll play a little.
thank you very much
  • #88
  • Jan 8, 2014 4:12am Jan 8, 2014 4:12am
Hi all,
We (our team) have prepared the tables set1, set2 and set3 our internet server mysql. We have two sets of tables, the first in 2011-13 as indicated GG53 and one with a greater historical thinking in larger time frames.

just a sample: http://my.jetscreenshot.com/10822/m_...-dmcg-43kb.jpg

We are thinking that we may be able to share by internet access to this database with you, we will take a decision shortly.

Crodzilla Thanks, good job. I miss the calculation of conditional probabilities about what happened in the past. Maybe are implicit in your calculations.

I continue to study the matter and waiting for the probability model that incorporates the concept of conditional probability to be found in the dilemma of the 3 doors.

thanks again.
F.
  • #89
  • Jan 8, 2014 4:47am Jan 8, 2014 4:47am
Quoting smallcat
{quote} {quote} i check it again, and find that some times the end of ZZ is not the same as fractals. I still use the "Custom Zigzag" by Killian. Edit: For alone using of ZZ, this is good. But for using with GG's idea (together with fractals), i think it must be changed a little, like Killian changed the ZZ indicator to "Custom ZZ". Thanks -smallcat {image}

Please smallcat could you share the custom zigzag by killian you talked about. Want to try something with the indicator. Thanks in advance
  • #90
  • Edited 6:17am Jan 8, 2014 5:59am | Edited 6:17am
Quoting Fran_ms
Hi all, We (our team) have prepared the tables set1, set2 and set3 our internet server mysql. We have two sets of tables, the first in 2011-13 as indicated GG53 and one with a greater historical thinking in larger time frames. just a sample: {image} We are thinking that we may be able to share by internet access to this database with you, we will take a decision shortly. Crodzilla Thanks, good job. I miss the calculation of conditional probabilities about what happened in the past. Maybe are implicit in your calculations. I continue to study the...
Fran_ms,

My historical database contains data from January 1st, 2011 till 31st December, 2012 - leaving 2013 for "out-of-sample" testings/predictions.
Just to make sure you are doing the same, because your statement suggest otherwise.

G.
  • #91
  • Jan 8, 2014 6:09am Jan 8, 2014 6:09am
Quoting Samuelkanu
{quote} Please smallcat could you share the custom zigzag by killian you talked about. Want to try something with the indicator. Thanks in advance
Samuelkanu, smallcat, killian,

My algo is using ZZ as a "unit" to analyze its internal components (fractals, total length).
The ZZ end-point is NOT my target. My target is the most probable trading SWING point.
It's true that many times they are same but often it's not - because ZZ end-point is the HIGH or LOW of the bar and not its CLOSE.

So looking for the "next bar in opposite direction" will give you different result than ZZ end-point - and it's not a "miss" or "error" as long as the trading entry point is better or equal or at least very near the ZZ end-point.

G.
  • #92
  • Jan 8, 2014 6:11am Jan 8, 2014 6:11am
Right GG53, 2011-2012, little mistake. today implementing database on gpbusd too. Thanks.

F. (from Utopy) lol
  • #93
  • Jan 8, 2014 6:14am Jan 8, 2014 6:14am
Quoting gg53
{quote} emonts, Let us bring everyone to that level by explaining the methods and logic done so far. After everything is available to everyone - we'll start answering specific questions. I think that otherwise this thread will become impossible to follow. G.

am here gg and happy new
  • #94
  • Jan 8, 2014 7:42am Jan 8, 2014 7:42am
Quoting emonts
Hi carl, Thank you very much for your compact description regarding probability. I have one question. I was comparing your fractal pattern occurrences against my own ones. I noticed some are equal or very close to equal, other differ completely. Anyway I have to check my sets again. But I also see in your examples the following Example 1, page one. For the fractal pattern <101> you find 509 exact matches. Example 2, page two. For the fractal pattern <101> you find 599 pattern matches. Same pattern but different occurrences. There is only a difference...
emonts,

I have been continuing to tweek my algo. It is almost an hourly occasion that I find myself testing, coding, and de-bugging this algo.
I am now have over 400 versions of the program. Recently, I have included fractal and zigzag identification into my program code, so that I have more control over it, removing that function from MT4. Now MT4 only sends recent bar data.

Whenever I tweek that code (fractal/zigzag), the whole database changes. That is what you are seeing in my screenshots of the document.

For that reason, do NOT take my screenshots as gospel, as far as database contents and statistics.

Carl
  • #95
  • Jan 8, 2014 9:56am Jan 8, 2014 9:56am
Hi GG,
In the picture, supose your indicator give's you the signal S2. After that what are your fractals sequence to check the pattern match?
You start at the beggining of ZZ - sequence 11011 ?
Or you start at signal? sequence 11 ?
Thanks
Attached Image
  • #96
  • Jan 8, 2014 10:36am Jan 8, 2014 10:36am
Quoting gg53
How-to: Calculating Probabilities Added to post #1 Probabilities (DOC & PDF formats) - by crodzilla - http://www.forexfactory.com/crodzilla Thanks, carl ! G.
Hi,

thanks Carl for the great doc it is well explained.

GG, are you calculating the probabilities in the same way as Carl (you made together the doc) or your way is a bit different?
thank you!

regards
  • #97
  • Jan 8, 2014 1:43pm Jan 8, 2014 1:43pm
Quoting killerno
{quote} Hi, thanks Carl for the great doc it is well explained. GG, are you calculating the probabilities in the same way as Carl (you made together the doc) or your way is a bit different? thank you! regards
Thanks, killerno.
I asked G that as well.
I have not heard anything different.

@MBEng, G has stated that his algo starts a new fractal count after each signal.

Carl
  • #98
  • Jan 8, 2014 1:58pm Jan 8, 2014 1:58pm
Quoting crodzilla
{quote} Thanks, killerno. I asked G that as well. I have not heard anything different. @MBEng, G has stated that his algo starts a new fractal count after each signal. Carl
Hi Carl,

But what happened to the 3 doors theory.
Fully disappeared in the probability calculations or was it never in ?

BR
Huub
It is not important how often you fall down, but how often you get up again
  • #99
  • Jan 8, 2014 3:38pm Jan 8, 2014 3:38pm
@ Crodzilla

In "bars in length until" you include the length in real time?
if I have length = 5

and my database I have:

1 1 2 3 4 4 4 5

in length would be bar until 7 without 5 or 8 with 5 ?

Many thanks
  • #100
  • Jan 8, 2014 4:13pm Jan 8, 2014 4:13pm
Quoting MBEng
Hi GG, In the picture, supose your indicator give's you the signal S2. After that what are your fractals sequence to check the pattern match? You start at the beggining of ZZ - sequence 11011 ? Or you start at signal? sequence 11 ? Thanks {image}
MBEng,

sequence 11 .

G.
1 trader viewing now