Once again, it's about the eternal: trend/flat. - page 7

New comment
 
Uladzimir Izerski:

Trending down and up is tops here or there.

Into the annals!
 
Дмитрий:
In the annals!
You ghoul!!!
 
Andrey Dik:

In my opinion, to create an indicator of something (a flat or other pattern) you need to open a chart and look for this pattern. If you have determined for yourself that a flat from 0 to 8 (roughly), open a chart of the desired instrument in the desired TF and search at this time with the parameters that you consider optimal (height/width). As soon as your brain says: "Yes, it looks like a flat" - mark it on the chart with lines as you see it, search further, find a new one - mark it and compare with the previous ones. The idea is to define clear criteria that will satisfy most of the patterns. This process is not quick, but then the pattern can be automated.

 
Alexander Laur:

What is the Root difference between a flat and a trend? The answer is in the direction of the position to be opened!!! In a flat, the direction is NOT important, while in a trend, the direction is VERY important.

As far as I understood, it's not a question of which way to open, but of determining the moment to open. This is where determining whether we are in a flat or trend helps.
 
Alexander Laur:

In my opinion, all arguments about "candles", "fractals" etc. etc. - This is just poor man's talk, because the toolkit is not the same.

What is the Root difference between a flat and a trend? The answer: in the direction of the open position!!! In a flat, the direction is NOT IMPORTANT, while in a trend, the direction is VERY IMPORTANT.

I'm sure everyone understands this very well, but then what do candlesticks have to do with it .... A trader should calculate the risks, not talk about fractals and candlesticks. :)

For successful trading it is important to know the f/o, that is why joo created this thread.

Before the release of important news the price may stay in flat for several days. At the news release, which is still a flat, IMPORTANT will show itself in full glory.

A candlestick on the D1 is nothing but a wave on the H1. You can also consider candlesticks, but they are tied to time segments, and do not carry complete information about the wave.

 
Alexander Laur:
Gentlemen, whether you are in a trend or in a flat is determined by your risks. On the same instrument, on the same TF, with some risks you will be in a flat, and with other risks you will be in a trend! If you are in a flat, the direction of opening is not important, if the trend is important. You need to understand it clearly.
You're confused... Is t/f important or not?
 
Alexander Laur:
Gentlemen, whether you are in a trend or in a flat is determined by your risks. On the same instrument, on the same TF, with some risks you will be in a flat, and with other risks you will be in a trend! If you are in a flat, the direction of opening is not important, if the trend is important. You need to understand this clearly.
It's a common sense, no one argues, but you are speculating from the risk position (MM), and TS - from the position of technical analysis, he needs the criteria of the model, from which he will then calculate the risks, and not vice versa.
 
Alexander Laur:
It all depends on your risks. :)
And if we are out of the market and our risk is therefore zero, what's on the price chart: a trend or a flat? :)
 
Alexander Laur:

I'm sure everyone understands this very well, but then what do candlesticks have to do with it .... A trader should calculate the risks, not talk about fractals and candlesticks. :)

Can you give an example of the calculated risk? What exactly do you need to calculate - what to subtract/add and multiply from what?
 
The simplest and most effective way to test a system for robustness is the ratio of the number of profitable optimization options to unprofitable ones. In other words, it is more likely that a profitable variant will be chosen for future work if this ratio is higher. On the other hand, this indicator is influenced by the history length. If you choose the history length so that during optimization the ratio of profitable to unprofitable optimization variants was 3/1 (try to get this ratio using trading robots - it will not work) at the amount of deals of some 1000 and more, you can say that the system is robust. But by applying a simple t/f filter on time I got a 5/1 ratio on the same optimization area! But if I had a tool that numerically determines t/f, then the ratio can be even higher!
1234567891011121314...21
New comment
Reason: