Analyse the most important STATISTICAL characteristics of the pattern and choose a method of trading on it. - page 2
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
You have to step it up gradually. First just find candles of a given size. Visually see what you get:
Search of a pattern, version "1.000"
Oh no, I don't play these games :D
Oh no, I don't play those games :D
Why did you even ask and start a topic then?
Why even ask then and start a topic in the first place?
maybe someone has worked with different statistical distributions of time series and given them characteristics
P.S. oh-ho-ho, the forum seems to have prompted the link itself, but not sure yet :)
So you want to cluster the patterns as well (calculate) ....
In other words, I don't want to fit into a pattern - here we open, here we close, and the high probability ones are closed... So, I don't need a pattern fit. That is, I don't need to fit the pattern - opened here, closed here, but the high probability trades in the pattern, they may not necessarily coincide with the "best" deal by the number of points or with the pattern high or low.
And what there is to cluster inside the pattern I don't know, I don't know anything
When a pattern is triggered, we carefully save the data about further price behaviour into a file, for example, we can record the minimum and maximum price changes after a certain period of time.
If the pattern signal is shown on several adjacent bars, it is better to take the data from the first triggering. If a series of triggers is long, then data from the first one should be taken first, and then half of the selected time interval.
Then we draw distributions and calculate exponential distributions for up and down movement. Depending on position of TP and SL, the average profit is calculated, of course, we should correctly calculate probabilities and take into account that the price can stand still at night, for example.
You can also use percentile, it's easier to calculate, you need more data to avoid surprises...
Gave you a direction where to dig). Although, there's a lot you can do....
You have to step it up gradually. At first, just find candles of a given size. Visually see what you get:
Search of a pattern, version "1.000"
So big and you believe in fairy tales...
When a pattern is triggered, we carefully save the data about further price behaviour into a file, for example, we can record the minimum and maximum price changes after a certain period of time.
If the pattern signal is shown on several adjacent bars, it is better to take the data from the first triggering. If a series of triggers is long, then data from the first one should be taken first, and then half of the selected time interval.
Then we draw distributions and calculate exponential distributions for up and down movement. Depending on position of TP and SL, the average profit is calculated, of course, we should correctly calculate probabilities and take into account that the price can stand still at night, for example.
You can also use percentile, it's easier to calculate, you need more data to avoid surprises...
Gave you a direction where to dig). Although, there's a lot you can do....
It's obvious that it's easier with a percentile than without a percentile.
maybe someone worked with different statistical distributions of time series and gave them characteristics
P.S. oh-ho-ho, the forum seems to have prompted the link itself, but not sure yet :)
ehhh... "what are you doing there...digging gaussians? then dig deeper - there are many more like them..." :-)
although the link is quite good, at least better than the ones seen here in the last couple of years
ehhh... "what are you doing there...digging gaussians? then dig deeper - there are many more like them..." :-)
although the link is pretty good, at least better than what we've seen in the last couple of years
No, I just need to trade not the pattern itself, but statistically the most probable direction... so-so... and the more trades the better, for statistics it's always better... Then I will always be able to see if the pattern is correctly predicted or if something goes wrong, and stop trading. But since profits and losses will be spread out over a large number of trades, I will not lose much if the prediction goes wrong, as I will trade the 10th part of them. And even in the 10th part you can get close to a loss probability of only 50%. Understand? I'm just not quite there yet myself. ) There we may get randomaysis, we may competently mix buy and sell deals and they will be opened not by waves inside the pattern, but by how much profit is possible here and now from every deal; thus even if the pattern changes but the probability inside the changed pattern is approximately the same, we will still obtain a satisfactory trading result.