Fractals, ZZ & the Pissing Dog | Page 21 | 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

  • #401
  • Feb 6, 2014 8:40am Feb 6, 2014 8:40am
Quoting gg53
<p>{quote}</p> <p> </p> emonts, That's a very good way to "backtest". Just remember that my database is evolving due to prediction "errors" and might be a little different than your currently. G.
Hi GG53,

I know that my DB is not up to date.
Can you tell us roughly how much your set 3 did grow due to error prediction over the last year.


BR
Huub
It is not important how often you fall down, but how often you get up again
  • #402
  • Feb 6, 2014 9:26am Feb 6, 2014 9:26am
Quoting emonts
{quote} Hi GG53, I know that my DB is not up to date. Can you tell us roughly how much your set 3 did grow due to error prediction over the last year. BR Huub

emonts,

About +22%.
But that shouldn't be a big factor since when prediction "error" occurs your algo should start from "true" ZZ leg end.
This will offset some of the differences in databases.

G.
  • #403
  • Feb 6, 2014 9:43am Feb 6, 2014 9:43am
(Continued...) On our way to "double within a month" (or not...)

So far I found 2 bugs that cost me dearly.
Found another issue with announcement - that also caused some heavy loss - huge spread opening far beyond normal.
This issue wasn't fixed yet - I just coded stop trading near news announcements.

So far the account is +16.2% in 8 trading days.
With this rate I don't think I can make that target.

G.
  • #404
  • Feb 6, 2014 9:49am Feb 6, 2014 9:49am
my attempt to predict just the very first ZZ end at the beginning of 2013 :-)
it's not correct but still good, i'm finding a way to implement "adaptability model" to correct this "error" before predicting the next one..

Attached Image (click to enlarge)
Click to Enlarge Name: 2014-02-06-eurusdm5.png Size: 36 KB


@gg: could you please tell something about your last 3 consecutive lost trades? Is that bug of your EA or prediction errors? Thank!
  • #405
  • Feb 6, 2014 12:10pm Feb 6, 2014 12:10pm
Quoting xixi
my attempt to predict just the very first ZZ end at the beginning of 2013 :-) it's not correct but still good, i'm finding a way to implement "adaptability model" to correct this "error" before predicting the next one.. {image} @gg: could you please tell something about your last 3 consecutive lost trades? Is that bug of your EA or prediction errors? Thank!

xixi,

It was an EA bugs.
One was BUY instead of BUY STOP, the other selecting wrong pair because of the 0 based arrays.
All 3 consecutive losses are the same currency pair.

G.
  • #406
  • Feb 6, 2014 6:50pm Feb 6, 2014 6:50pm
Quoting fxrambo
{quote} now about the pip distance is what my mind has as i deal with manual insertion of data.am not yet where a lot are not even closer but that pip thing still ringing in mind as i do the candle counts on zigzag
For some reason I stopped getting emails about this thread.
My latest idea about ZigZag pip distance is to also use percentage of average daily range.
The new build will make this much easier as I can create data structures similar to the sql database.
Each node can hold counts, probabilities, number of success, number of failures, and ...
Then on each new candle we re-calculate the current node we are at and possibly add a new node.
Each time a true turning point is reached the data structure is used to follows the candle patterns, fractals, and zigzags as they occur.
Kind of like the start of a chess game. If the first move is P-K4, that rules out checking openings using P-Q4. If Black replies P-K4 then the opening book is even more reduced.
  • #407
  • Feb 7, 2014 10:17pm Feb 7, 2014 10:17pm
Quoting gg53
Our target is to find major SWING points in the forex market, to pinpoint accurately entry and exit point for trading. BackgroundUnlike gambling, lottery, Coin toss, etc. which are (hopefully) purely random - the Forex market is biased.It does have a "memory" and it’s based on needs, speculation, and human behavior.No one here (I hope) is looking for 100% sure "bet". Just to learn a bit more on Forex and its internal "DNA" - and maybe to get a little "edge". We know that the ZigZag indicator (ZZ) best describe those SWING points, but it does it...
Very interesting video! Subscribed and will be waiting for more
  • #408
  • Feb 8, 2014 4:21am Feb 8, 2014 4:21am
@ gg53
I have doubt in calculating probability of "next bar".
I believe that you calculate the probability and set to "actuall bar".
If you have high probability "Current bar" and the market that gives you direction, taking into account the probability of "next bar" it would signal.
If we have high probability "Current bar" and the market does not give us that direction, then we must correct that pattern.
My first question is:
apply "Monty Hall" a whole sequence of bars of length 12?

If I have 40% bearish, 60% bullish, we consider high probability bullish?

If we have high probability in an address on "actuall bar" as we interpret the probability of "next bar". I almost all inflection points, studying probability of "next bar" have eg

100% bullish in actuall bar
100% bullish in next bar

By law, we would have to take "next bar" opposite "actuall bar" bar.

thank you very much
  • #409
  • Feb 9, 2014 8:28am Feb 9, 2014 8:28am
Quoting Thoughts
{quote}I really do appreciate where youre coming from twoblink..but I'm not entirely convinced that is the case. I think the price tells us what its going to do before it does it..I think such windows, however small, exist even in the case of black swan events. (This is just my opinion) @GG, I appreciate this thread G; your heart and your time. 100 Million Thanks.

So the concepts I struggle with are these; there are the "known unknowns", and the "unknown unknowns" a la black swans.

But there's a different category that I'm addressing altogether that I think might be referred to as "Gladwell Swans".

Interest Rates will move, by how much I don't know, but will be announced next Friday <== Known (Short term) Unknown known

Some kid in his garage invents a cryto-key that can hack any computer <== Suspected Known Unknown

Aliens who all look like Gisele Bundchen land <== Unknown Unknown

Major Earthquake rocks Japan <== Known (Medium Length, Inevitable) Unknown known

At current spending Rate, America will become Greece basically <== Known (Known) Known

Since in theory we can't prepare for the unknown unknown, because if we could, they would be known unknowns; I don't worry about them. They of course will have the biggest impact; but I don't know about them so Gladwell and Tilab is wrong; I can't spend my entire life searching for an unknown unknown just to make it a known unknown. The category that I think is the second biggest impact; and one that I'm not sure we can do about also; is the one I have marked a Known Known Known.

If you are the ONE kid that knows that the Emperor has no clothes on, and everybody else keeps pretending otherwise, then it almost makes no difference the fact that you know.

So let's take a look at it for a moment; Currently in the USA; 50%+ do not pay taxes, so 50%+ are supported by < 50%. Obama has spent more than all past presidents COMBINED. Mathematically impossible to sustain. So knowing this; has the US economy crashed? Not really (yet). But you know it will, SO WHAT? Knowing about it, and being able to do something about it; are two different things; and while I'm of the "price will tell you the whole story" camp, we already know the whole story; the emperor has no clothes; but everybody is STILL PRETENDING HE DOES.

So the paradox here is; is the price reflecting "reality" or reflecting the everybody perceived reality? While I would love to say it's the former; data seems to support that it's the latter... So to me a Gladwell swan; you go to Australia, take a pic of a black swan, go back to England, tell everybody there's a black swan; and they tell you "I'm sure that's just photoshopped" and go on living their "white swan only" lives.
google:
  • #410
  • Feb 9, 2014 9:32am Feb 9, 2014 9:32am
http://www.forexfactory.com/showthread.php?p=7269538#post7269538

As promised, a semi-comprehensive look at the use of Pascal's Triangle as it applies to Forex.

@GG, I hope this helps!
google:
  • #411
  • Feb 10, 2014 3:18pm Feb 10, 2014 3:18pm
Interesting thread, what is it, lindenmayer?
  • #412
  • Feb 11, 2014 2:53am Feb 11, 2014 2:53am
Quoting twoblink
http://www.forexfactory.com/showthread.php?p=7269538#post7269538 As promised, a semi-comprehensive look at the use of Pascal's Triangle as it applies to Forex. @GG, I hope this helps!

twoblink,

Many thanks. Going to read...

G.
  • #413
  • Feb 11, 2014 3:02am Feb 11, 2014 3:02am
i really admire your work of all in this and past thread(ultimate truth of eurusdd even he seems a bit weird guy altho he did gave a nice idea of DNA) .i am new in forex and i have no programming skills but i am following you i mean all of you for some time now.GG's killing zone is my favorite .
now the point i write this post is that since you are comparing past with present and it seems the past has memory i see one of my(not mine,gratz who ever write it) ) fractal indicator could work with what you looking for of when to start counting new ZZ leg .
it seem he has also a zz inside .
But if i am talking silly dont mind me at all.
Signal it appears 2 bars in the past and is showing the way .I think it counts 5 or 7 bars to find which one is the turning point.now price cant go opposite direction of that signal unless a new one comes out .So maybe price goes 3-4 pips away but it will evantually return and makes new signal at the ooposite direction .
I hope the coders can get something usefull out of this code.If not dont blame me cause i want to help but i cant offer coding knowlenge.
Keep up the good work.
this is no repaint but is lagg indicator.so if you could use it to compare this indi with your DB sets maybe you could get the signal a bar earlier which is a nice bonus.
Attached File(s)
File Type: mq4 ReversalFractals.mq4   7 KB | 626 downloads
  • #414
  • Feb 11, 2014 3:07am Feb 11, 2014 3:07am
(Continued...) On our way to "double within a month" (or not...)

I went for a long weekend since thursday morning, came back yesterday to find that the system "stuck".
MT4 was frozen solid... big loses accumulated, internal SL wasn't triggered and no new trades.
I considered canceling the test and starting the whole thing from fresh.

The end decision was to continue as-is, and try to recover the loses and the lost time.

G.
  • #415
  • Feb 11, 2014 3:12am Feb 11, 2014 3:12am
forgot to post the pic .here it is.
Attached Image (click to enlarge)
Click to Enlarge Name: eurusd.m5.png Size: 84 KB
  • #416
  • Feb 11, 2014 3:17am Feb 11, 2014 3:17am
pff i cant even make a proper post ... well the aqua color arrows are the reverse fractal.
end of posts.
oo by the way GG , i found it good call to keep as it is.finishing your goal with all the ups and downs is life ,is everything.
  • #417
  • Feb 11, 2014 10:08am Feb 11, 2014 10:08am
Quoting zarkon
i really admire your work of all in this and past thread(ultimate truth of eurusdd even he seems a bit weird guy altho he did gave a nice idea of DNA) .i am new in forex and i have no programming skills but i am following you i mean all of you for some time now.GG's killing zone is my favorite . now the point i write this post is that since you are comparing past with present and it seems the past has memory i see one of my(not mine,gratz who ever write it) ) fractal indicator could work with what you looking for of when to start counting new ZZ...

zarkon,

Thanks. I'm going to look what's inside this "present"...

G.
  • #418
  • Feb 11, 2014 2:50pm Feb 11, 2014 2:50pm
Quoting twoblink
http://www.forexfactory.com/showthread.php?p=7269538#post7269538 As promised, a semi-comprehensive look at the use of Pascal's Triangle as it applies to Forex. @GG, I hope this helps!

twoblink,

I've read, double and triple read - amazing...
The next version of my bot will definitely implement trade management based on those ideas.

T H A N K S !!

G.
  • #419
  • Feb 12, 2014 4:26am Feb 12, 2014 4:26am
Quoting twoblink
http://www.forexfactory.com/showthread.php?p=7269538#post7269538 As promised, a semi-comprehensive look at the use of Pascal's Triangle as it applies to Forex. @GG, I hope this helps!
excellent post
  • #420
  • Feb 12, 2014 10:50am Feb 12, 2014 10:50am
GG,

For your test, forget the most exotics pairs, you are paying BIG spread for nothing ...
0 traders viewing now