neural network and inputs - page 25
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
specifically on paterns - yes they are on history and of course they are no longer profitable!!! ))))
But seriously, I searched for ways to identify patterns in patterns, of course I didn't find them like everyone else, the only thing I did was to calculate on all timeframes how many and what patterns are repeated on the history. There is nothing to search for - everything that is repeated at least 10 times on history is less than 20% of all historical data, i.e. all possible patterns take about 1/5 of the data, and 4/5 is what never repeats (or extremely rarely).
So the question is, are we looking there? We are looking for patterns where the pattern is the lack of copy in history.
specifically on paterns - yes they are on history and of course they are no longer profitable!!! ))))
But seriously, I searched for ways to identify patterns in patterns, of course I didn't find them like everyone else, the only thing I did was to calculate on all timeframes how many and what patterns are repeated on the history. There is nothing to search for - everything that is repeated at least 10 times on history is less than 20% of all historical data, i.e. all possible patterns take about 1/5 of the data, and 4/5 is what never repeats (or extremely rarely).
So the question is, are we looking there? We are looking for patterns where the pattern is the absence of a copy in history.
Finally, at least one bothered to calculate the statistics. That's exactly the kind of numbers I was getting.
Only 10% up and 10% down... and the rest is 80% rubbish.
IgorM:
.... We look for patterns where the pattern is the absence of a copy in the story.There will never be an exact copy (or there will rarely be, roughly similar, as you wrote). Just above, I wrote about the effect of internal market timing on patterns, it makes the task very difficult, plus different volatility (the size of a single "brick" - a pattern), plus the formation of patterns on different dimensions (not to be confused with timeframes) - the market is fractal. God forbid that one should confuse these dimensions. There are a lot of complexities, and nobody promised it would be easy.
You can compare it with leaves in the forest, for example a maple leaf, no two leaves are the same, but we can easily recognize that it's a maple leaf, for example, and not an oak leaf. The leaves may be large or small, but they are still maple leaves.
Or, for example, a tea cup. The cups may vary in size, colour and shape, but they are all cups with handles.
Breathing in and out, also a pattern, you can breathe often, slowly, deeply, but it's still a breathing in and out.
A pattern is a systematically repeated stable element (fragment) or sequence of elements (fragments) of behaviour.
There will never be an exact copy (or there will rarely be, roughly similar to what you wrote). Just above, I wrote about the effect of internal market timing on patterns, it makes the task very difficult, plus different volatility (the size of a single "brick" - a pattern), plus the formation of patterns on different dimensions (not to be confused with timeframes) - the market is fractal. God forbid that one should confuse these dimensions. There are a lot of complexities, and nobody promised it would be easy.
You can compare it with leaves in the forest, for example a maple leaf, no two leaves are the same, but we can easily recognize that it's a maple leaf, for example, and not an oak leaf. The leaves may be large or small, but they are still maple leaves.
Or, for example, a tea cup. Cups may be different in colour and shape, but they are all cups with handles.
Inhale and exhale, also a pattern, you can breathe often, slowly, deeply, but it's still an inhale and exhale.
A pattern is a systematically repeated stable element (fragment) or sequence of elements (fragments) of behaviour.
Not once have I heard the opinion that pattern trading does not work in forex, only in equities. I think I heard it even from Neo. Do you succeed in finding them (patterns) in forex? Or are you talking about stocks?
More than once I have heard the opinion that pattern trading does not work in forex, only in equities. I think Neo even said that once... Are you able to find them (patterns) in forex? Or are you talking about stocks?
I'm talking about forex. the search works but there's always a lot of unforeseen mess.
Well, I kind of trumpeted it on the forum back in spring, but people are looking because they want to look - well, there is no predominance of paternals over the so-called chaos, even visually looking at the chart and not too lazy to pay attention to see that to the left of the pattern and to the right of the pattern there are no patterns, and if you count the number of bars before the pattern and after and compare how many bars were in the paternals it turns out >the market is in chaos 60% of the time. If you calculate it using the script, the numbers you have given or me will be the same, but it is obvious that >2/3 of the time the market is in chaos
i see, nobody is looking for 100% coincidence, i generally just analyzed the movement "up- down- up-down", i.e. just the nature of the movement and the different number of simultaneously analyzed bars, if you search here on the forum you may find a pattern search using zig-zag and other methods
yes it is repetitive, but if you compare the history that does not bear patterns with the time of existence of patterns, the stable element is just the absence of patterns
Well, I kind of trumpeted it on the forum back in spring, but people are looking because they want to look - well, there is no predominance of paternals over the so-called chaos, even visually looking at the chart and not too lazy to pay attention to see that to the left of the pattern and to the right of the pattern there are no patterns, and if you count the number of bars before the pattern and after and compare how many bars were in the paternals it turns out >the market is in chaos 60% of the time, and if you calculate the script, the numbers you have given or I have, but it is clear that >2/3 of the time the market is in chaos
it's clear, no one is looking for 100% coincidence, i generally just analyzed the movement "up - down - up - down", i.e. just the nature of the movement and the different number of simultaneously analyzed bars, here on the forum if you look you can find search patterns using zig zag and other methods
yes it does, but if you compare the history which does not bear patterns with the time of existence of patterns, the stable element will be just the absence of patterns
i'm talking about forex, but i think also on stocks it will turn out that patterns won't have 50/50 with chaos, i will try to analyse stocksSo... where did this limit of happiness come from in the first place (50/50)? It should be about... 10-20%. It just doesn't mean that the other 80% is exactly chaos. It is quite possible that it is possible to squeeze something out there, just by completely different methods.
As for me the ideal MTS - is a forecast at each bar (just how many points price will pass during some period of time (maybe market time), or for example where the price will be in N points - +N or -N) - plus estimate of probability of such forecast. To this probability it is possible to attach MM, or enter after a certain threshold (for example probability of 0.65 of successful realization of a scenario). But the point is, I'm for this approach - simply put, "there is always a pattern", it's just at each moment the probability of successful realization fluctuates, it is an oscillator with extreme values 0 and 1.
Anyway, this is the old rubric about whether a pattern can be parametric, or whether it is always binary (in Neo terms - "you can't be pregnant just a little bit" or something like that).
More than once I have heard the opinion that pattern trading does not work in forex, only in equities. I think Neo even said that once... Are you able to find them (patterns) in forex? Or are you talking about stocks?
In FOREX it is more difficult, for several reasons, you can see how to do it in Felix's thread: http://forex.kbpauk.ru/showflat.php/Cat/0/Number/265918/an/0/page/0#Post265918 And there are also several threads related to it.
Or in SpartaFX's blog :http://spartafx.livejournal.com/
The patterns Neo wrote about are of course only on stocks. I've written about both those patterns ( Neo) and the "universal" ones (like Adverse Tactics models), but I've always shared what we're talking about.
In FOREX it is more complicated, for several reasons, you can see how to do it in Felix's thread: http://forex.kbpauk.ru/showflat.php/Cat/0/Number/265918/an/0/page/0#Post265918 Also, there are several threads related to it there as well.
Or in SpartaFX's blog :http://spartafx.livejournal.com/.
I have read it all. By the way, about HOW it's done, there's not a word, just vague allusions :).
By the way, as far as I understand, Felix has no patterns, but rather a sanction (which is not the same thing). At least he thinks it's a sanction (probably right). Or to put it another way - he has exactly what I was talking about - "there is always a pattern" (a parametric pattern, not a binary one so to speak). There is no "yes/no pattern" situation, but there is a situation when the oscillator gives a probability of successful moving up or down.