Fractals, ZZ & the Pissing Dog | Page 19 | 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

  • #361
  • Jan 30, 2014 7:36am Jan 30, 2014 7:36am
Quoting Michael_Lnr
Hi, I think my algorythm is performing pretty much the same as GG53's. Here are the results for current m5 eur/usd. Still having doubts about renko. Working on that. Kind regards, Michael {image}
Error handling has not been discussed yet...
How do you not get early signals?

Carl
  • #362
  • Jan 30, 2014 7:46am Jan 30, 2014 7:46am
@ GG53

When are you planning to release the final part "Handling Prediction errors".

Dennis
  • #363
  • Edited 2:54pm Jan 30, 2014 11:18am | Edited 2:54pm
The "Three door puzzle" - (handling predictions "errors") & Calculating probabilities :

The "three door puzzle" is an interesting and unusual probability question. Here's how it works:
You are a contestant in a game show, and the game show host tells you there is a prize behind one of the three doors you face.
You have to guess which door to open.

But when you make your guess, instead of opening the door you picked, the game show host opens a different door...one that he knows has nothing behind it.
So now you're down to two doors. And the game show host says: "I'll let you change your choice, if you want to."

And the question is, do you change your guess? Or keep your original choice?

The natural assumption is that it makes no difference, and so people tend to keep their original guess - reasoning being that we all know you should never "second guess" yourself. But is that really a wise choice?
(quote from "The problem site")

This puzzle is also called "Monty Hall problem" (http://en.wikipedia.org/wiki/Monty_Hall_problem (http://en.wikipedia.org/wiki/Monty_Hall_problem))

Relating to our "puzzle" -
We also have 3 options: use the current pattern, use longer pattern ending with "0", use longer pattern ending with "1".
We know our current pattern status and we know the other optional pattern probabilities.
Try to apply the above logic to "next bar" and "fractal pattern longer". Use the math that is clearly explained.

To be continued...

G.
  • #364
  • Edited 2:57pm Jan 30, 2014 11:27am | Edited 2:57pm
(Continued...) The "Three door puzzle" - Calculating probabilities:

If your prediction turns out to be wrong (before zz leg completion) - look for other patterns that will produce higher probability for the new sequence.
In other words - recalculate probabilities with similiar patterns.
The emphasis is on this: ..."(before zz leg completion)".
The target direction of the ZZ remains the same.

here is the question:
So far you got 0101 fractal pattern.
You predict 01010 fractal pattern, and wait.
Now you get 01011 - which is a prediction "error", and the zz is not finished yet.
Now what? do you stay with the initial fractal pattern?

No.
You go with what the market shows you.
But you have to find another pattern, as similiar as possible to what you got so far, which will give you high probability of "1" at the end and not "0".
...And from now on - that will be your pattern for future predictions.

There are 3 options WHY this "error" happened:
1 - the actual current market of "1" is a "mistake" (overshooting of some kind)
2 - the first element of our pattern is a "mistake" (we didn't have enough confirmed data when establishing it)
3 - some signal in between was a "mistake" (overshooting/undershooting of some kind)

Options 2 & 3 suggest you recalculate/change your pattern.
Option 1 suggest you stay with what you got now.
That's 66%/33% suggestion to change pattern to fit the last "1" presented by the market.

Do you want to ignore such a massive suggestion?

To be continues...

G.
  • #365
  • Jan 30, 2014 11:43am Jan 30, 2014 11:43am
(Continued...) The "Three door puzzle" - Calculating probabilities:

The MOST likely scenario is that one of the previous elements in the fractal pattern is an overshoot - which means shouldn't exist at all.
The LESS likely scenario is that one of the elements of the fractal pattern should be an opposite fractal (i.e. "0" instead of "1").

So go with the most likely scenario - start eliminating one fractal at a time from your pattern - to get the highest probability of "1" (or "0") at the end of your pattern, as the market suggest.

My statistics shows that you should start eliminating from the first element (meaning that the first element of the fractal string is wrong in most of those cases) - but that's not fully validated.

G.
  • #366
  • Jan 30, 2014 12:57pm Jan 30, 2014 12:57pm
Thanks G, that is perfect.

Carl
  • #367
  • Jan 30, 2014 2:37pm Jan 30, 2014 2:37pm
Quoting gg53
(Continued...) The "Three door puzzle" - Calculating probabilities: The MOST likely scenario is that one of the previous elements in the fractal pattern is an overshoot - which means shouldn't exist at all. The LESS likely scenario is that one of the elements of the fractal pattern should be an opposite fractal (i.e. "0" instead of "1"). So go with the most likely scenario - start eliminating one fractal at a time from your pattern - to get the highest probability of "1" (or "0") at the end of your pattern, as the market suggest. My statistics shows...
Thank you gg53, my brain hurts I need refreshment. I am not sarcastic, I am really enjoying reading this different perspective of trading. Million miles away to put this together in form of indicator. Keep up the good work!
  • #368
  • Jan 31, 2014 1:35am Jan 31, 2014 1:35am
Hello all,

As you probably know metaquotes will release "new" MT4 the 3rd february, with lots of problems in it.

For those working with offline chart, be sure to replace your renkos producer, because all change with new metatrader.

As a latest security, you must copy your MT4 folders in another partition, because if your install is in C:\program files, sometimes with windows and their new MT4 there is security conflicts, and no indicators could be written in folders.

So, a simple task, just copy your MT4 folder in another partition. And replace your targets shotcuts.

Cross fingers for the update go right !
  • #369
  • Jan 31, 2014 2:52am Jan 31, 2014 2:52am
Hi Everyone,

Herewith 3 screenshots showing EUR/USD, GBP/USD & USD/JPY.

Not perfect yet as you can see.

Kind regards,
Michael
Attached Image(s) (click to enlarge)
Click to Enlarge Name: GBP-USD-m5.PNG Size: 30 KB
Click to Enlarge Name: UJ-m5.PNG Size: 30 KB
Click to Enlarge Name: eur-usd-01-31.PNG Size: 32 KB
  • #370
  • Jan 31, 2014 5:31am Jan 31, 2014 5:31am
hello gg. i'm at a loss understanding your EA entries. i thought it was based on your forex gt arrow on a renko chart. but why did it open that gbp/nzd short 120 pips from the top? could you please shed some light?

thanks
  • #371
  • Jan 31, 2014 6:21am Jan 31, 2014 6:21am
Quoting bubola
hello gg. i'm at a loss understanding your EA entries. i thought it was based on your forex gt arrow on a renko chart. but why did it open that gbp/nzd short 120 pips from the top? could you please shed some light? thanks
could be he is doing a larger timeframe, because he has only 3 % down with -140 pips, while previously he has gained 3% with 22 pips on a trade.
  • #372
  • Jan 31, 2014 7:41am Jan 31, 2014 7:41am
Quoting bubola
hello gg. i'm at a loss understanding your EA entries. i thought it was based on your forex gt arrow on a renko chart. but why did it open that gbp/nzd short 120 pips from the top? could you please shed some light? thanks
bubola,

Because it still have some bugs... and it's not the only one...
Again, it's the first time that this system is running live - with all it's complexity of choosing currency pairs.
We still have 3 more weeks of performace and weekends of debuggings.

G.
  • #373
  • Jan 31, 2014 7:45am Jan 31, 2014 7:45am
Quoting Starry
{quote} could be he is doing a larger timeframe, because he has only 3 % down with -140 pips, while previously he has gained 3% with 22 pips on a trade.
Starry,

This specific trade is a bug.
I'll try to post a summary later this weekend.

G.
  • #374
  • Jan 31, 2014 8:36am Jan 31, 2014 8:36am
Quoting gg53
{quote} bubola, Because it still have some bugs... and it's not the only one... Again, it's the first time that this system is running live - with all it's complexity of choosing currency pairs. We still have 3 more weeks of performace and weekends of debuggings. G.
i see. good luck with your work, then.
  • #375
  • Feb 1, 2014 6:38pm Feb 1, 2014 6:38pm
I'll give you straight up, the problem I haven't been able to solve and I'm not sure there's a solution .. Federer.. an easy solution I mean.

Let's say you wanted to avoid traffic accidents with your car, and so you install a spiffy new AI . It sees you are driving you fast, or regulates the tire pressure to compensate for wet roads etc. It can even listen in on police scanners to see if there is an accident.

But physics being what they are, you only control half the variables, our someone wants to hit your car, will, AI be damned.

Let's say, super duper turning point predictor indicator works, how would it know about planned assassination of a major world leader or an earth quake that would leak radiation? These outliers a la Malcolm Gladwell, are the more devistating part of forex yet nothing we can do to predict them.

Question, would you be interested in my original hockey puck fractal stuff? I tried not so much to predict turning points so much as entry and reentry points that yielded best risk to reward, and they often end up close to turning points..

Great thread.
google:
  • #376
  • Feb 2, 2014 4:32am Feb 2, 2014 4:32am
Quoting twoblink
I'll give you straight up, the problem I haven't been able to solve and I'm not sure there's a solution .. Federer.. an easy solution I mean. Let's say you wanted to avoid traffic accidents with your car, and so you install a spiffy new AI . It sees you are driving you fast, or regulates the tire pressure to compensate for wet roads etc. It can even listen in on police scanners to see if there is an accident. But physics being what they are, you only control half the variables, our someone wants to hit your car, will, AI be damned. Let's say, super...
twoblink,

Yes, I'm interested.

G.
  • #377
  • Feb 2, 2014 7:18am Feb 2, 2014 7:18am
Hi, GG53!
I am an mql4/mql5 programmer. My question is: is there a completely automated EA, based on your strategy?
  • #378
  • Edited 9:58pm Feb 2, 2014 4:37pm | Edited 9:58pm
Quoting Korshun-Mike
Hi, GG53! I am an mql4/mql5 programmer. My question is: is there a completely automated EA, based on your strategy?
Korshun-Mike,

Yes, but in the near future I might need your programming abilities.
Thanks.

G.
  • #379
  • Feb 2, 2014 5:36pm Feb 2, 2014 5:36pm
Quoting twoblink
..Let's say, super duper turning point predictor indicator works, how would it know about planned assassination of a major world leader or an earth quake that would leak radiation? These outliers a la Malcolm Gladwell, are the more devistating part of forex yet nothing we can do to predict them...
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.
  • #380
  • Feb 3, 2014 8:09am Feb 3, 2014 8:09am
hi everyone,

im still far behind of gg53's indicator because of some contradictions.

the logic is :

"current fractal pattern"
"current length"
"current +1 bar (which is the next bar) in opposite direction"

right?

but if you look at the attatched picture, you can see:

fractal pattern ----- length
[] ----- 4 (the red circle in the picture)
[] ----- 1 (the blue circle in the picture)

this means that the prediction is confirmed for "null fractal".

i can see many "null fractal prediction" at gg53's pictures.

my understanding is that if there is no fractal, we cannot predict but wait.

is this because of the 3 doors problem? or other tactics not revealed yet?

im really confused


thanks
jim,
Attached Image (click to enlarge)
Click to Enlarge Name: eur.PNG Size: 230 KB
1 trader viewing now