Fractal theory - page 7

New comment
 
Now we can calculate the structure of the fractal, I will think about how to do it. There is a suggestion that we should take a piece on a large timeframe, measure the ratio of the pullback to the trend and look for this ratio on a smaller timeframe. Then it will be possible to understand in which part of the fractal pattern we are now.
 

Fractal is more of a philosophical notion than a concrete trading one

Bill Williams devalued the original term "fractal" by substituting an ordinary chart kink as the meaning

We could have just said "5-bar kink".

Why did Bill stop there?

Because even a single bar on the chart is a fractal (in the sense of scalability and similarity)

here we may also note that high and low of any bar are the breaks of a chart of a smaller TF

having connected high and low (i.e. two nearest breaks) we obtain a wave

it is not far from here to the Elliott Wave Theory and the scalability and similarity of the structure of the wave formations

i.e. any "fractal" of Williams is the moments of transition from one wave to another

there is such a fractal-wave dualism

 
223231:

Made fractal charts from tick charts. It turns out that as the timeframe decreases, the detail of the chart increases, if we take very rough, then up to a straight line.

The grouping by points is taken as the basis- when the price has made a swing at N points the bar is closed. It turns out to be a point timeframe.

And then you compare it with bars sliced purely by time?

I think the logic is strange.

 
In this thread you should put in red on the front page: it's about fractals, NOT about Billy's fractals.
 
Silent:

And then you compare that to bars sliced purely on time?

The logic seems strange to me.

Why then compare it to time bars? I wasn't supposed to be comparing time charts with each other, but specifically point charts. By the way, they are much more stable than the timebars and are easier to analyse.
 
transcendreamer:

Fractal is more of a philosophical notion than a concrete trading one

Bill Williams devalued the original term "fractal" by substituting an ordinary chart kink as the meaning

We could have just said "5-bar kink".

Why did Bill stop there?

Because even a single bar on the chart is a fractal (in the sense of scalability and similarity)

here we may also note that high and low of any bar are the breaks of a chart of a smaller TF

having connected high and low (i.e. two nearest breaks) we obtain a wave

it is not far from here to the Elliott Wave Theory and the scalability and similarity of the structure of the wave formations

i.e. any "fractal" of Williams is the moments of transition from one wave to another

that's the fractal-wave duality

Indeed, there is something common with the wave theory, the only difference is that the wave theory does not go into the essence of what is going on, but tries to analyze what is there. I want to get to the essence of what is going on. Williams' fractals do not reflect the essence of what is going on at all, they simply mark local extremes which do not mean anything in themselves.

In radio engineering there is a method of fractal noise filtering, based on the fact that noise does not change with increasing time. In essence, it also predicts future noise values.

 
223231:

There is indeed something in common with wave theory, the only difference is that wave theory does not go into the essence of what is happening, but tries to analyse what is there. I want to get to the essence of what is happening. Williams' fractals do not reflect the essence of what is happening at all, they just mark local extremes, which do not mean anything in themselves.

In radio engineering there is a method of fractal noise filtering, based on the fact that noise does not change with increasing time. In essence, it also predicts future noise values.

I think it can only be applied to the radio; on the market the noise/signal ratio changes with time.

I mean there's more noise in the flat areas, just visually

 
transcendreamer:
The flat areas are mostly caused by time sampling. Collected ticks for the week, plotted on the basis of the pips sampling (a candle closes after the price passes a certain number of pips), there are no flots as such, there are trends of different size, if the price is not moving up, it is moving down, but not standing still, only the scale of the movement changes!
 
223231:
Fleeting patches are generated mostly by temporal sampling. Collected ticks for the week, built charts based on the point sampling (a candle closes after the price passes a certain number of points), there are no flotsam as such, there are trends of different size, if the price is not moving up, it is moving down, but not standing still, only the scale of the movement changes!

You can't argue with that!Every bar has some kind of intra-bar movement

and how are punctual bars better than temporal ones? does this turn out to be some sort of renko?

 
transcendreamer:

You can't argue with that!Every bar has some kind of intra-bar movement

Why are punctual bars better than temporal ones?

Because as the timeframe decreases, the behavior of the chart barely changes, in contrast to the timeframe. Plus, with time sampling the period of oscillations is less stable than with point sampling. It turns out that timeframe depends not on an abstract time, but on the transactions performed in the market, and roughly speaking, on the traded volumes, which is equivalent to the volume of resources traded in the market. In my opinion, this is much more correct.
12345678910
New comment