Probability. - page 9

 
Maxim Kuznetsov:

We take the maximum and minimum of the price from the point where the start of the movement is marked, and between them we construct a standard deviation channel. The further from the middle of the channel the higher the probability of reversal. :-)


Show me that in percentages at the current moment. Then I will agree with you.

).

 
Uladzimir Izerski:

Show me that in percentages at the current moment. Then I will agree with you.

).

And no fantasy - just a property of the instrument. By constructing a channel you are making a hypothesis of a linear price movement and the limits that are drawn are the sigma of your hypothesis. The further the price goes beyond the limit, the less credible the hypothesis is.

 
Maxim Kuznetsov:

And no fiction - just a property of the tool. When you draw a channel, you are making a hypothesis of a linear price movement, and the limits you draw are the sigma of your hypothesis. The further the price goes beyond the limit, the less credible the hypothesis is.


The channel doesn't give me the right answer. I use channels, but for other purposes.

Can you give me any formula for calculating for a channel?

 
Uladzimir Izerski:

Here's the H1.

"Estimated points" - do you mean in perspective or from history?

I analyse price behaviour on history and give a future forecast on the output.

Analogous to AI without error checking.


Tester points are similar to Fibo, they are points of extremums used to set the Fibo grid, i.e. working range, and I was wondering how the indicator detects these points.

 
Uladzimir Izerski:

The channel doesn't give me the right answer. I use channels, but for other purposes.

Can you give me any formula for calculating for a channel?

this is the standard deviation channel in the terminal - the documentation (at the edge in the MQ source code) should tell you in detail how it is constructed.

 
Veniamin Skrepkov:

Calculated points are by analogy to Fibo, the points of extremums for which the Fibo grid is set, i.e. the working range, and I was wondering how the indicator defines these points.


If we rely on the calculated points, they are rather zigzag knees or in fact waves of some order.

 
Maxim Kuznetsov:

This is the standard deviation channel in the terminal - the documentation (at least in the MQ source code) should explain in detail how it is constructed.


I understand how the channels are constructed.

I'm interested in how you calculate the percentage probability.

I want to hear some practical advice. I have not come to argue and prove anything.

 

Why do I need all this?

To see the hidden capabilities of the machine brain.

To see on the price chart what we cannot see with our eyes, or rather our minds.

Yusufkhoja is doing the same thing, but he has a different approach and the goal is the same. To see what others do not see.

Not many people understand him, by the way, and for good reason.

Maybe they will not understand me either. But it would not upset me in the slightest).

 
Uladzimir Izerski:

I understand the channels as they are constructed.

I am interested in the formula for how you calculate the percentage of probability.

I want to hear some good advice. I did not come here to argue or prove anything.

First of all, I am not arguing with you :-)

Next, about the terminal's tools and probabilities:

when you build a standard deviation channel, it's assumed that :

- in the centre of the channel, our hypothesis about the linear movement is correct, i.e. the probability of white/black=1/2, i.e. no movement is at variance with the hypothesis about the channel

- on the border in the channel is sigma, i.e. white/black=1/2 *0.9=0.95

- at the two sigma level the hypothesis is fundamentally wrong :-)

but since we are not working with discrete data, and taking into account volatility, own error, market dynamics,
If we move from the %% to qualitative estimations, we may consider that from 0.8 to 1.5 probability of the reversal is high (>0.8 in favor of moving to the centre of the channel), higher than 1.5 - either we missed the reversal and it already happened or we built the channel incorrectly :-)

 
Maxim Kuznetsov:

First of all, I'm not arguing with you :-)

Next, about the terminal tools and probabilities:

when a standard deviation channel is plotted, it is assumed that :

- in the centre of the channel our hypothesis about the linear movement is correct, i.e. the probability of white/black = 1/2, i.e. no movement contradicts the channel hypothesis

- at the border of the channel is sigma, so white/black=1/2 *0.9=0.95

- at the two sigma level the hypothesis is fundamentally wrong :-)

but since we are not working with discrete data, and taking into account volatility, own error, market dynamics,
If we move from the %% to qualitative estimations, we may consider that from 0.8 to 1.5 probability of the reversal is high (>0.8 in favor of moving to the centre of the channel), higher than 1.5 - either we missed the reversal and it already happened or we built the channel incorrectly :-)


I see the logic in that. Thank you. I'll ponder how to attach useful qualities to my baggage.

I have a very different channel-building principle, but I'll think about it.

"About the argument" came out as a general phrase and wasn't directed specifically at you. Sorry for my stupidity).