The similarity system | Page 534 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Similarity: A Practical Approach 859 replies

Similarity Trading - The ZZBB Method 215 replies

Rainbow similarity 256 replies

Similarity Stuff 17 replies

Directional Similarity 0 replies

Attachments: The similarity system
Exit Attachments

The similarity system

  • #10,661
  • Oct 21, 2014 9:04am Oct 21, 2014 9:04am
Quoting vlady1974
{quote} Excellent!!! Here's a thought.... are you able to do the same thought experiment but have an Uneven TZ as the premise, with 1 bar to the right? Just want to see what you come up with and how your mind works. I'm borrowing a leaf from Dr Van K Tharp.
  • #10,662
  • Oct 21, 2014 9:13am Oct 21, 2014 9:13am
I have read about half this thread and I find it fascinating. But from a scientific perspective I am amazed that even halfway through the thread that nobody has tested any of these patterns on random data. Every theory must be tested in this scientific manner with a control group of data, otherwise..... Not to disprove eurusdd of course, but to gain more in depth knowledge. Can someone please point me to the right section of the thread. Thanx.
  • #10,663
  • Oct 21, 2014 9:21am Oct 21, 2014 9:21am
Quoting vlady1974
{quote} Excellent!!! Here's a thought.... are you able to do the same thought experiment but have an Uneven TZ as the premise, with 1 bar to the right? Just want to see what you come up with and how your mind works. I'm borrowing a leaf from Dr Van K Tharp.
You mean that rust bucket of mine? Check you know where mate
I want more brain power!
  • #10,664
  • Edited 9:42am Oct 21, 2014 9:29am | Edited 9:42am
Quoting Porkpie
I have read about half this thread and I find it fascinating. But from a scientific perspective I am amazed that even halfway through the thread that nobody has tested any of these patterns on random data. Not to disprove eurusdd, but to gain more in depth knowledge from such hypotheses. Can someone please point me to the right section of the thread. Thanx.
Hi Porkpie,

For the random data the transient zone (TZ) theory would in some cases even work better.

Eurusdd introduced the TZ concept on the Markov process. That is the process with no memory, i.e. the subsequent candles are independent events (imagine flipping a coin every time unit to decide on the market behaviour).
A simple strategy: The probability of the subsequent transient zones for a Markov process will be (given a judicious choice of h resulting in 97% recurrence) 3%*3%=0.09%, so, when one observes two TZs on subsequent candles, it is almost a safe bet for a clearance of at least one of them, because a probability of them remaining uncleared is basically vanishing.

One of the problems often encountered here is exactly the opposite: that the market is not sufficiently random, it does not behave like a Markov process, it has memory and therefore the events on two candles are not independent. Hence the probability of having two TZs on subsequent candles is not ignorable, particularly when strong breakouts occur.

Here also the answer is hidden (and Eurusdd mentioned it in the thread somewhere) as to why the market will not be able to be adapted to prevent the exploitation by the TZ approach. If it becomes more noisy: we'll have more recurrence, yay! If it goes the other direction: there will be more trends, yay!

Cheers,
k
  • #10,665
  • Oct 21, 2014 9:34am Oct 21, 2014 9:34am
Quoting cfdtaleong
{quote} You mean that rust bucket of mine? Check you know where mate
LOL

I was just curious as to whether you would half the number of scenarios you had on your first pic
ie. not the top half
A pip is only worth it if you know how much you risked to earn it
  • #10,666
  • Oct 21, 2014 9:52am Oct 21, 2014 9:52am
Quoting Porkpie
I have read about half this thread and I find it fascinating. But from a scientific perspective I am amazed that even halfway through the thread that nobody has tested any of these patterns on random data. Every theory must be tested in this scientific manner with a control group of data, otherwise..... Not to disprove eurusdd of course, but to gain more in depth knowledge. Can someone please point me to the right section of the thread. Thanx.
In fact, TZ says that the market favors recurrent prices rather than transient prices. Think about it. Will you have a problem with forex if transient prices have 97% weight? Certainly no! Because most of the time your chart will look like this!!!
Attached Image (click to enlarge)
Click to Enlarge Name: Capture.PNG Size: 17 KB

Each bar there is an hourly bar! You will certain trade with your eyes closed!!!
  • #10,667
  • Oct 21, 2014 9:57am Oct 21, 2014 9:57am
Quoting Eurusdd
{quote}In fact, TZ says that the market favors recurrent prices rather than transient prices. Think about it. Will you have a problem with forex if transient prices have 97% weight? Certainly no! Because most of the time your chart will look like this!!!{image} Each bar there is an hourly bar! You will certain trade with your eyes closed!!!
This is why you say price is King and needs protection
I want more brain power!
  • #10,668
  • Oct 21, 2014 9:59am Oct 21, 2014 9:59am
Quoting kprsa
{quote} Hi Porkpie, For the random data the transient zone (TZ) theory would in some cases even work better. Eurusdd introduced the TZ concept on the Markov process. That is the process with no memory, i.e. the subsequent candles are independent events (imagine flipping a coin every time unit to decide on the market behaviour). A simple strategy: The probability of the subsequent transient zones for a Markov process will be (given a judicious choice of h resulting in 97% recurrence) 3%*3%=0.09%, so, when one observes two TZs on subsequent candles,...
nice read! Love this comment!
  • #10,669
  • Oct 21, 2014 10:00am Oct 21, 2014 10:00am
Quoting Eurusdd
{quote}nice read! Love this comment!
I had a great teacher.
k
  • #10,670
  • Oct 21, 2014 10:02am Oct 21, 2014 10:02am
if 5,5 of 100 candles are transient and the last tz was 20 candles ago whats the probability of the 21th candle to be transient? 5,5/(100-20) ?
follow the flow
  • #10,671
  • Oct 21, 2014 10:10am Oct 21, 2014 10:10am
Quoting Atihun
if 5,5 of 100 candles are transient and the last tz was 20 candles ago whats the probability of the 21th candle to be transient? 5,5/(100-20) ?
Please. The probability that an entire candle will contain only transient prices is zero, almost surely. A candle with only transient prices is a candle where the high=open/close and low=open/close. Very unlikely!!!

So, that is not the precise way to deal with transient prices!!!
  • #10,672
  • Oct 21, 2014 10:20am Oct 21, 2014 10:20am
Quoting Eurusdd
{quote} Please. The probability that an entire candle will contain only transient prices is zero, almost surely. A candle with only transient prices is a candle where the high=open/close and low=open/close. Very unlikely!!! So, that is not the precise way to deal with transient prices!!!
so it doesnt make sense to look at the frequency in which the boxes occure?
follow the flow
  • #10,673
  • Oct 21, 2014 10:25am Oct 21, 2014 10:25am
Quoting Atihun
{quote} so it doesnt make sense to look at the frequency in which the boxes occure?
See K's post above. The market isn't truly Markovian, so be careful trying to make it so - your theory will work most of the time, until it doesn't, in a very big way.
  • #10,674
  • Oct 21, 2014 10:27am Oct 21, 2014 10:27am
i was thinking about the following.
if we get a possible transient zone there is a percentage it will be cleared within h bars of 90+% depending on your h value.
lets say we had 7 possible zones which were cleared before h right was completed. at the same time we have a frequency of confirmed zones where h right didnt clear the zone.

isnt it likely after 9 cleared zones and a period of time that there will be a zone which will not be cleared because of the frequency % ?
follow the flow
  • #10,675
  • Oct 21, 2014 10:33am Oct 21, 2014 10:33am
Quoting kprsa
Here also the answer is hidden (and Eurusdd mentioned it in the thread somewhere) as to why the market will not be able to be adapted to prevent the exploitation by the TZ approach. If it becomes more noisy: we'll have more recurrence, yay! If it goes the other direction: there will be more trends, yay! Cheers, k
Sounds like a paradox... :-)
IMO, TZ is kind of a principle. There could be million specific ways/strategies to apply that principle.
Market could be able to adapt a specific strategy but could not be able to adapt to every strategies because the number of these strategies is unlimited.
So the principle could not be change unless market destructs itself...
  • #10,676
  • Oct 21, 2014 10:42am Oct 21, 2014 10:42am
Quoting Atihun
i was thinking about the following. if we get a possible transient zone there is a percentage it will be cleared within h bars of 90+% depending on your h value. lets say we had 7 possible zones which were cleared before h right was completed. at the same time we have a frequency of confirmed zones where h right didnt clear the zone. isnt it likely after 9 cleared zones and a period of time that there will be a zone which will not be cleared because of the frequency % ?
Perhaps, but even if so, what will be the magnitude of the bar, and what part of it will be transient? This is why I started looking at range and renko bars when I first saw the concept. I was hung up on EURUSDD saying NOT to use the TZ's as S/R - and while it may look like that is what many are doing, in fact, their trades are simply probability based - they are waiting for Rare, before trading against Rare*Rare.
  • #10,677
  • Oct 21, 2014 10:51am Oct 21, 2014 10:51am
Quoting Barbaresco
{quote} Perhaps, but even if so, what will be the magnitude of the bar, and what part of it will be transient? This is why I started looking at range and renko bars when I first saw the concept. I was hung up on EURUSDD saying NOT to use the TZ's as S/R - and while it may look like that is what many are doing, in fact, their trades are simply probability based - they are waiting for Rare, before trading against Rare*Rare.
A pip is only worth it if you know how much you risked to earn it
  • #10,678
  • Oct 21, 2014 10:54am Oct 21, 2014 10:54am
if rare is trans. whats rare rare? trans. after trans?
follow the flow
  • #10,679
  • Oct 21, 2014 12:00pm Oct 21, 2014 12:00pm
Quoting ninpojutsu
1: When one indicator shows start of dis-similarity and the other doesn't, the zone that will be created is a very good target. I think most of you use the method described in posts 1 Can anyone clarify this doubt? I see two new areas, which one should I accept it? Natalia. {image}
I see that no one can confirm me, I just want to learn .. have to talk myself answering me my own questions!

The signal came in EURUSD, gave good pips, can see how the price went to the green zone.
Someone is kind enough to confirm whether it is good entry?
I look for on the green areas?

Thank you.
Natalia.
Attached Image (click to enlarge)
Click to Enlarge Name: Sin título.png Size: 51 KB
  • #10,680
  • Oct 21, 2014 12:26pm Oct 21, 2014 12:26pm
Quoting ninpojutsu
{quote} I see that no one can confirm me, I just want to learn .. have to talk myself answering me my own questions! The signal came in EURUSD, gave good pips, can see how the price went to the green zone. Someone is kind enough to confirm whether it is good entry? I look for on the green areas? Thank you. Natalia. {image}
To answer your question properly.....

1)You have your targets correct.

2)What are you doing? You are trading a return to a level you identified.

3)When do you enter? In your picture, why buy now? What is your trigger? By indicator or price action?

4)Your templates only tell you where the targets are. They DO NOT tell you when to enter.

5)Just because you identify a target, it does not mean it will be returned to 100% or in the near future. The green area can be there for days or weeks.

6)You decide if your entry is good or bad. If you must have an answer then base on PA yes the entry is good because it looks like price has bounced of support but this is just ONE factor of consideration. There may be other factors that tell you the entry could be bad.

7)Learning to identify targets is one thing. How to trade them is another.

8)In your picture you label "good sell" and "now buy" Should be "sell target", "buy target"

Please digest this before asking more questions.
I want more brain power!
12 traders viewing now