Looking for patterns - page 83

New comment
 

Look, Makar, you said levels and I thought:

if the indicator shows a dark histogram and the following bars are light, that means the movement has slowed down. And if the candle of the last movement had a big body, then there is a horizontal level in that place. Further on, you can check this place once again using other candlesticks. And the level indicator is ready!

 

Aleksei Stepanenko:

And the levels indicator is ready!

I agree, but it's not like we're looking for levels here, but the patterns associated with levels would be interesting...

I'll try to draw it now

 
MakarFX:

but the patterns associated with the levels would be interesting...

There is definitely a pattern! Levels slow price movement. Three or four levels in the path of a wave and there is a pullback.

 

Completion of each move creates an extremum

I have noticed that if price does not return to 50% FIBO on a breakout, it will pass 1.61

and if it reaches 50% then it is either flat or reversed

In the picture the first and third Fibo are in plus and the second one is in minus.

Dschinghiz 2.5% ADR20 (by the way, this parameter is adequate for all the majors)


 
MakarFX:

In the picture, the first and third Fibo are on the plus side and the second on the minus side.

Try it, I'm not good at Fibo. I think the real levels may not coincide with it somewhere.

 
Aleksei Stepanenko:

When several conditions are superimposed, for example, how long the trends lasted after the opposite trend of 500 pips. There can be many such conditions, and interesting findings are possible.

You are missing one thing. Yes, we were able to identify the trends but it's not bad. But you have forgotten about the flots. I showed in the pictures the second indicator showing what is going on at the moment, the trend component is trending or flat (one chaos subtracted from the other one by time or cluster volatility) that's all.
Otherwise you will "hear the ringing, but don't know where it is".
In some situations pulling back on trends will be justified, and in others it seems that the mechanism of separation has not been studied.
 
Aleksei Stepanenko:
It is necessary to make a time-dependent indicator, it is necessary to estimate something in relation to something, otherwise it is impossible to estimate the infinite variety of quotation movements in a one-sided way. Now you cannot tell how strong the trend is and what is the probability of its reversal.
 
Aleksei Stepanenko:

Try it, I'm not strong on Fibo. I think the actual levels may not coincide with it somewhere.

The point is not FIBO, as they say "everything is relative".

In this case the percentage can lead to a pattern regardless of the asset or its volatility.

I.e. if price broke through the level and came back, but didn't go 50% back, it will go forward from the level another 61%(something like that)

 
CHINGIZ MUSTAFAEV:
... for this you need a time-based assessment.

or by levels.

Genghis, you and Alexei have my deepest admiration.

 
multiplicator:

on channel reversals at all, it works horribly.

I'll continue...

You can do regression on a straight line.

you can do regression on a polynomial, on a logarithmic function, on an exponential function. all functions are somehow smoothly curved.

are there any broken functions in mathematics?


1...767778798081828384858687888990...306
New comment