Fractals, ZZ & the Pissing Dog | Page 32 | 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

  • #621
  • Apr 20, 2014 9:03am Apr 20, 2014 9:03am
So the dog still at no 7 , now the zz start search for tech analize algo , great thank for the update . keep it well !
  • #622
  • Apr 20, 2014 10:15am Apr 20, 2014 10:15am
Quoting gg53
{quote} Etecy, 1. No, because there was no confirmed ZZ near by, so the algo continues as usual. Once second signal is activated, there will be no "error" checking - because it will cause repainting. 2. We always count from the arrow - except in cases of "error" and the leg is not finished yet by another signal. G.
@ gg53

You have stated somewhere:
"If, before finding the next probable swing point, I get an "error" prediction - I'm starting over the calculation from a confirmed MT4 ZZ."

Can you please say when you check for an "error".

Thanks so far.
  • #623
  • Apr 20, 2014 1:29pm Apr 20, 2014 1:29pm
Quoting Etecy
{quote} @ gg53 You have stated somewhere: "If, before finding the next probable swing point, I get an "error" prediction - I'm starting over the calculation from a confirmed MT4 ZZ." Can you please say when you check for an "error". Thanks so far.
Etecy,

I check for error if current leg is still in process and I'm getting a confirmed ZZ end-point more than 3-4 bars from my starting point.

G.
  • #624
  • Apr 20, 2014 1:32pm Apr 20, 2014 1:32pm
Quoting Mr_You
{quote} My wording was wrong. If I have a fractal pattern match and the length falls "true" within the logic is there a number of bars previous and after (the fractal pattern match most recent bar) that I should be looking for a nextbar/currentbar pattern SET3 match? In other words you stated nextbar/currentbar pattern matching happens around the fractal/length requirement. I was wondering if you have any guidelines or restrictions for where this pattern occurs around the fractal/length requirement. Thanks!
Mr_You,

If fractal pattern match and bars count match - start looking for "next bar in opposite direction" probability.

G.
  • #625
  • Edited Apr 21, 2014 5:49am Apr 20, 2014 4:57pm | Edited Apr 21, 2014 5:49am
Quoting K40
So the dog still at no 7 , now the zz start search for tech analize algo , great thank for the update . keep it well !
K40,

are you using 64 bit or 128 bit encryption in your posts?

G.
  • #626
  • Apr 20, 2014 7:19pm Apr 20, 2014 7:19pm
nor ... hehehe
  • #627
  • Apr 21, 2014 5:03am Apr 21, 2014 5:03am
@gg53

Arrows on screenshot show ZZ end points that have not been repainted.
Is the marked "2" last confirmed ZZ end point or "3" ?
Attached Image

Thanks a lot.
  • #628
  • Edited 6:14am Apr 21, 2014 6:03am | Edited 6:14am
Quoting Etecy
@gg53 Arrows on screenshot show ZZ end points that have not been repainted. Is the marked "2" last confirmed ZZ end point or "3" ? {image} Thanks a lot.
Etecy,

For me, "confirmed" ZZ end-point is when we got a fractal in opposite direction.
Reffering to your screenshot - it's 2 bars after the first opposite fractal mark, since fractals are marked only after 2 bars.
So, in this case, Point 3 ZZ is confirmed after 4 bars from it's marking,.
Attached Image


G.
  • #629
  • Apr 21, 2014 9:22am Apr 21, 2014 9:22am
@ gg53

Thanks so far for the explanations.
Would you please be so kind and show a screenshot from your indi where an "error" occured.
And maybe mark the latest zz end from which your counting continued.
Ignore if you asked too much. Thanks!
  • #630
  • Apr 21, 2014 9:40am Apr 21, 2014 9:40am
Quoting Etecy
@ gg53 Thanks so far for the explanations. Would you please be so kind and show a screenshot from your indi where an "error" occured. And maybe mark the latest zz end from which your counting continued. Ignore if you asked too much. Thanks!
Etecy,

My indie doesn't show that visually - that work is done "behind the scene" by the algo updating the databse.
It also doesn't show the ZZ legs and fractals.
It will require digging into the database for those updates and find them on the chart.
Sorry...

G.
  • #631
  • Apr 21, 2014 8:20pm Apr 21, 2014 8:20pm
Quoting gg53
Fractals, ZZ and the Renko bars (and don't forget the Pissing Dog...) The Renko version of the ForexGT_Arrow EA is finished (incl. Currency-Strength scanner and other supporting filters) . A major problem was the Backtesting of the Renko chart - which is always a problematic issue on almost all trading platforms. I think that we solved that issue too, and soon I'll post backtests results with at least 90% accuracy. Solving the Renko backtesting issue is a major cause of happiness here. Currently, the results of ForexGT_Arrow (Renko Ver.), EURUSD...
Hi G,

Congratulation ... great result ...
Can you explain how to use Renko to predict turning point ? So you do not use fractals and Zigzag for this ?
And how to test it ?

Thanks
-Gio
  • #632
  • Apr 22, 2014 10:59am Apr 22, 2014 10:59am
Quoting giovanni4000
{quote} Hi G, Congratulation ... great result ... Can you explain how to use Renko to predict turning point ? So you do not use fractals and Zigzag for this ? And how to test it ? Thanks -Gio
giovanni4000,

Doing it on Renko requires additional steps.
The logic is simple but the work is hard...
Simply divide your Renko into smaller Renko's blocks.

BackTesting Renko bars is a bitch...
First D/L reliable Tickdata (try TickStory and data from Dukascopy).
Erase offline MT4 .hst file and close MT4.
Export to MT4 from TickStory.
Start MT4 via TDS (TickDataSuite, Brit's creation).
Create Renko offline chart using M1 and Renko EA (there are several EA's to choose from).
Open Renko offline chart.

Backtest now should show 99%+ accuracy.

I don't want to go into much details on Renko subject in this thread.

G.
  • #633
  • Apr 25, 2014 5:39am Apr 25, 2014 5:39am
Quoting gg53
{quote} giovanni4000, Doing it on Renko requires additional steps. The logic is simple but the work is hard... Simply divide your Renko into smaller Renko's blocks. BackTesting Renko bars is a bitch... First D/L reliable Tickdata (try TickStory and data from Dukascopy). Erase offline MT4 .hst file and close MT4. Export to MT4 from TickStory. Start MT4 via TDS (TickDataSuite, Brit's creation). Create Renko offline chart using M1 and Renko EA (there are several EA's to choose from). Open Renko offline chart. Backtest now should show 99%+ accuracy....
Thank you very much G ...

-Gio
  • #634
  • Apr 27, 2014 4:00am Apr 27, 2014 4:00am
Quoting gg53
... Length (bar count) - yes, it's the least important, BUT... You have to take into account large bars (pips wise) and not just the number of them, they are equivalent to two or three "regular" bars in other ZZ's. G.
gg and guys,

Do we go back all the way to SET 1 and recount everything? Those two years worth of historic bars.

The way I did SET 1, months ago, was I didn't treat long bars this way. No matter the length/pips, I just gave it 1 or 0 DNA.
  • #635
  • Apr 27, 2014 1:31pm Apr 27, 2014 1:31pm
Quoting summicron
{quote} gg and guys, Do we go back all the way to SET 1 and recount everything? Those two years worth of historic bars. The way I did SET 1, months ago, was I didn't treat long bars this way. No matter the length/pips, I just gave it 1 or 0 DNA.
When determining the ZZ length you check the size (in pips) of each bar and follow the classification given in this post (http://www.forexfactory.com/showthread.php?p=7395099#post7395099). Which basically says "5 pips equals 1 bar for ZZ length purposes".
  • #636
  • Apr 28, 2014 1:39pm Apr 28, 2014 1:39pm
Quoting summicron
{quote} gg and guys, Do we go back all the way to SET 1 and recount everything? Those two years worth of historic bars. The way I did SET 1, months ago, was I didn't treat long bars this way. No matter the length/pips, I just gave it 1 or 0 DNA.
Yep, another change to the algo. That's ok. Getting closer still...

Carl
  • #637
  • Edited 2:58pm Apr 29, 2014 2:35pm | Edited 2:58pm
Hello! G,
Is this the same thing which Forex GT_Arrow calculates from the the history of pairs for swing points? I want to ask this for simplify my concept.

Thank you.

Regards,
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Arrow.PNG Size: 31 KB
Click to Enlarge Name: Arrow_.PNG Size: 32 KB
  • #638
  • May 1, 2014 4:45am May 1, 2014 4:45am
Quoting crodzilla
{quote} Yep, another change to the algo. That's ok. Getting closer still... Carl
SET1 (the long bars) needs to be recalculated too? I just need to be sure.

I haven't the time yet to check this out but I'm anxious to know, anyone did the algo change yet and seeing improvements?
  • #639
  • May 1, 2014 7:30am May 1, 2014 7:30am
Quoting summicron
{quote} SET1 (the long bars) needs to be recalculated too? I just need to be sure. I haven't the time yet to check this out but I'm anxious to know, anyone did the algo change yet and seeing improvements?
It seems that bar lengths that exceed some pre-determined lengths need to be counted as 2, 3 or more bars, when storing the lengths of zigzags. This oversight could be why so many of us see lots of signals, when we really should not. More lengths will be pushed into a longer space.

So, SET1 needs to include the increased bar lengths. So, that when these lengths happen in the future, the database has them.

Carl
  • #640
  • May 2, 2014 6:26am May 2, 2014 6:26am
Quoting QueFx
Hello! G, Is this the same thing which Forex GT_Arrow calculates from the the history of pairs for swing points? I want to ask this for simplify my concept. Thank you. Regards, {image} {image}
QueFx,

That's ForexGT_Arrow results for same time period as yours on EURUSD M15.
Attached Image (click to enlarge)
Click to Enlarge Name: eurusdm15.png Size: 38 KB


G.
0 traders viewing now