Searching for market patterns - page 46
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
Why is that?
And the code is really not cleaned up, I would say horrible, difficult to understand. In addition, there are errors in order modification.
But even though it is "semi-working", it worked on the section I gave it.
I did it in 15 min. and I didn't make any optimizations or adjustments, I just know that this theme works and that there is no pity to show it. He trades m5, close stops, etc. - And according to you noise. Do you have a real argument that "noise" is small timeframes or just blah blah blah)))? Or show me an example of an EA that you made on higher timeframes and consider it an argument for predictability of higher timeframes:)
Well, I sketched it out in 5 minutes, I just know this theme works. He trades m5s, close stops, etc. - and you say noise. Do you have a real argument that "noise" is small timeframes or just blah blah)))) Or show me an example of an EA that you made on higher timeframes and consider it an argument for the predictability of higher timeframes:)
I don't know what to say exactly about your "outline" so far - few tests have been done. You need to fix the bugs and tidy up the code (not for you, but for someone who is interested), to do tests at different parts of history, then we can talk about "noise" or "not noise".
In addition, I never said that you cannot make money on noise. I said that noise cannot be analyzed and predicted on its basis (at least I do not know how, but I admit the possibility of creating some "mechanical", without analytical part of TS, that can earn on market noise - using fitting and the property of market inertia). Maybe your exp exploits the "noise" features of M5, and will not work on other TFs - tests are needed.
Unfortunately, I cannot show my work - too much time and effort has been put into it, and I do not yet have any "five-minute" ideas that show less noise in higher TFs.
So far it is difficult to say anything specific on your "outline" - few tests have been carried out. You need to fix the bugs, comb out the code (not for you, but for those who are interested), conduct tests at different parts of history, and then you can talk about "noise" or "no noise".
In addition, I never said that you cannot make money on noise. I said that noise is impossible to analyse and predict based on it (at least I don't know how). Maybe your exp exploits the peculiarities of M5 "noise" and it won't work on other TFs - you need tests.
Unfortunately, I can't show my work - too much time and effort has been put into it, and I don't have any "five-minute" ideas that show less noise in higher TFs yet.
Well, then at least give me a link to where
joo:
It was clearly shown by Mathemat and his namesake that the patterns decrease as you go down in the TF, starting from H1 and rising above H1.
I'm sure there is no such correct proof in nature
The problem is that they do not have any logical or practical arguments when it comes to the question :)).
S.I. You have a unique understanding of "noise" and that you can make money on it but you can't analyse it)))
There was a discussion here. But that was just the tip of the iceberg, as I understand it.
ZZZ Analyse means that by studying the properties of an object, it is possible to form an opinion about its further behaviour. Noise cannot be analysed and has only the simplest of statistical characteristics. It is on use of statistical characteristics of noise (and not on its analysis) that TS may be built. But as experience of this forum shows the life time of such TS is very short.
ZZZY
просто получается, что некоторые с умным видом говорят что можно торговать, а что нельзя, кто фигнёй мается, а кто реальный трейдун))), а аргументов не логических не практических :)
Show us with your fingers.
There was a discussion here. But that's just the tip of the iceberg, as I understand it.
there's no proof, not even a nemecker. ;) The author of the thread, by the way, admitted that the ox completely predetermined the result there
Thank you for remembering. I've checked the 5 minute, hour and day lags. The maximum mutual information between the lags is on the hourlies; the daylies and 5-minute lags are less. But I haven't closed my thread yet either. Although, I checked for myself that when randomly mixing a sign of increments while keeping the initial volatility, the mutual information obtained does not statistically differ from the initial series. That is, the dependencies attributable to the signs are not significant, in other words, the volatility of 99.9% provides the non-random dependencies found. But I counted the mutual information from system to system, and you can also count for each element of the alphabet.
ZZZ Analyse means to study the properties of an object to be able to form an opinion about its further behaviour. Noise cannot be analysed and has only the simplest statistical characteristics. It is on use of statistical characteristics of noise (and not on its analysis) that TS may be built. However the life of such TSs, as the general experience of this forum shows, is very short.
show me where you found such a definition of noise))) It's good noise that you think you can make money on.
ZZZY
Point your finger.
If you intend to go on trumpeting that "there is no noise", be so kind as to tell me how to make sure that "there is no noise".
This mantra is annoying.
Are you a billionaire pipsqueak like Paukas?
Not a word about Sweene. But this is a clinical case of systemic non-smoking :)
1. there's no proof, not even a little bit. ;) The author of the thread, by the way, admitted that the ox completely predetermined the result there
2. show me where you found such a definition of noise)))
3. You for example)))1. If you don't like the proof in the link, well, I have no other for you. What I have, empirical, doesn't suit you either.
2 It's not a definition, it's my thoughts on the possibility of noise analysis.
3. In my post I politely "if you would be so kind as to prompt me please" asked for guidance on how I can be convinced of what they say. This request does not bind anyone to anything. And you're asking me for proof. I'm not going to prove anything to anyone. And to those who voluntarily post their "proof" I say thank you very much (unlike you), whatever they are (proof).
ZS On my some developments are based profitable TS dozens of people who write me letters of gratitude, they do not require proof from me, but just say "Thank you". Although, I don't need anyone's gratitude, and I wouldn't like to be accused of being a chatterbox.
1. If you don't like the proof in the link, well, I have no other for you. What I have, which is empirical, doesn't suit you either.
2 It's not a definition, it's my thoughts on the possibility of noise analysis.
3. In my post I politely asked for instructions on how to verify their words, "If you would be so kind, please tell me. This request does not bind anyone to anything. And you're asking me for proof. I'm not going to prove anything to anyone. And to those who voluntarily post their "proof" I say thank you very much (unlike you), whatever they are (proof).
didn't get it, the hit-and-run))) What evidence was there and where was it posted and I boorishly didn't say "Thank you very much"? :)
SZS on my some developments are based on a profitable TS dozens of people who write me letters of gratitude, they do not require me to prove, but just say "Thank you".
We are discussing the subject, not sparring. I.e. we need arguments or proof, or at least logical reasoning, and not just - somewhere someone showed something :)
And basically I just don't like being categorical, especially when one understands the topic in a very superficial way. Otherwise it is absolutely clear that the evidence that some timeframe is better predicted does not exist in principle.
And essentially I just don't like being categorical, especially when one understands the topic in a very superficial way. Otherwise, it is absolutely clear that there is no proof that any timeframe is better predictive in principle.
How categorical? What are you talking about? Everyone has an opinion here. Or do you need to put an IMHO at the end of each post, just in case?
And concerning the best TFs I've stated quite accurately and specifically, based on purely personal experience, without referring to anyone - H1 is predicted by neurocells better than other TFs.
Smaller than H1 is worse - because the noise is lower. Higher is worse, because there are no patterns useful for making profit, which can be detected by neuro-sets.
This is my experience, MY experience. I hope that's clear? That's what I've said before and I'm saying it now.
Are there other ways to find out which TF is better for specific types of TS? If there are, please do, I've already suggested my way. If not, then no, and let's end this useless conversation.