Looking for patterns - page 4

[Aleksei Stepanenko]

Here's an old thread the administrators keep bringing up that's interesting, just on the subject of beneficiaries and more: https://www.mql5.com/ru/forum/12342

[Aleksei Stepanenko]

Макс:
Three whole pages of patterns. :)

OK, Max, don't poke fun, say something on the subject.

I have thoughts too, I'll write some more.

[Alexander_K2]

There's no fooling around looking for something... We need the Grail. A destiny and a holy one. The rest is nothing.

[khorosh]

Vitaliy Maznev:

Likewise. About the bottle, if you only knew how lacking it is :)

Yeah, to use it to shut offmost of my brain and talk nonsense and nonsense while getting a buzz out of it, marvelling at how great a smart conversation I'm having).

[Vitaliy Maznev]

By the way, on patterns. I really didn't initially understand your goals specifically. And therefore approached the question from the wrong angle. You would describe in more detail the patterns you are already using, and I and others could add to them based on private experience. The fact is that a huge number of patterns emerge with the use of indicators and tools. Quite a few trading systems are built on this. And so if you are aiming to implement this in indicators or Expert Advisors, then really set your direction by describing your approach specifically. That way it will be easier to get others involved.

[Vitaliy Maznev]

Here is an example: There is one global moving average (233). It clearly appears on all timeframes. And if we take higher timeframes, the first touch of it on the daily, almost always leads to a powerful bounce. On the 4h at strong news it may occur without a bounce, but not so often. Is this roughly the kind of pattern you are interested in?

There are a huge number of such patterns. I am also interested in this topic, so let's develop it further.

[[Deleted]]

Alexander_K2:
There's no fooling around looking for something... We need the Grail. A destiny and a holy one. The rest is nothing.

The cuckoo flew away after all, sadly.

[Aleksei Stepanenko]

Yes, I wanted to talk about technical analysis, those are the things that can be formalised more concretely.

I know a few patterns, or I think I do. I'm not narcissistic or self-righteous, and I'm willing to admit mistakes without a problem. Well, not the point. The following are the obvious patterns I've been thinking about:

- well the fact that trends exist is silly to say, they do exist,

- there is some periodicity of movement,

- There are levels,

- a real chart is different from a random wandering, the presence of spikes,

- others that I want to ask you all about.

Knowing that these patterns exist is not helpful. So yes there is, so what? Benefits are only possible with statistical analysis of these and other patterns. That is what I wanted to talk about, to share experiences, to learn.

And what are you guys afraid to talk about? Everyone thinks their knowledge is unique. Now imagine how many of us on earth are so unique? Were? Will there be? What is the probability of your unique knowledge coming into someone else's head? I think a big one, we are all thinking the same thing. And not smarter or stupider than each other.

[Vitaliy Maznev]

Aleksei Stepanenko:

What are you guys afraid to tell? Everyone thinks their knowledge is unique. Now imagine how many of us are so unique on earth? Were there? Will there be? What is the probability of your unique knowledge coming into someone else's head? I think a big one, we are all thinking the same thing. And not smarter or stupider than each other.

I, too, am inclined to think that there cannot be any secrets here, just as we will not find anything new. Another question is what and how to use. I myself am not a coder, but it would be very interesting to work with someone who is.

The second point is that my subjective opinion leans towards the fact that not indicators were written for regularities, but vice versa: regularities are often adapted to existing tools. So that the movements do not scare away and look understandable and orderly.

[Aleksei Stepanenko]

Here are the statistics of the ratio of useful wave movement to its full movement. Here and below we are talking about daily waves, the duration of which is on average 10-20 hours.

The next wave on the chart with the movement size from point A to point C consists of two parts. The first part from point A to B is in the shadow of the previous low S. And this part of the wave cannot be called a real advance, because price has already been here recently. But the movement from point B to C is precisely that movement. So, if we look at the ratio of BC/AC on the entire chart of the EURUSD, for example, there were 5436 such waves in total, and the frequency of repetitions

So it turns out that the ratio of the real wave movement to the full wave movement,

- which is from 10 to 70% happens in 90% of cases,

- which is between 20 and 60% occurs 80% of the time.

And the probability estimate:

There is also a skewness from the first wave in which it is not clear from which point to count the useful movement. I didn't think about it right away, I should have excluded the first wave from the calculation. Well, so far, that's a start.