From theory to practice - page 1558

 
Anatolii Zainchkovskii:

Honestly - I'm sick of this forum. There was a specific question - how to determine if there is a giant outlier, by what criteria. If the only criterion is OI, how to get it into the terminal, or how to calculate it.

The answer to this question was a lot of Vysotsky's waffling, high philosophy with accusations of my signals and just plain stupidity.

What is there to talk about?

Anatoly - this is not specifically addressed to you, but simply a reasoning about the pointlessness of all endeavours on the forum.

 
Aleksey Nikolayev:

The main point is that discord is not predicted in the future, but sought in the recent past.

By the time you find it using statistical methods, it will be too late) The best discontinuity is a stoploss, or a breakdown.
 
Alexander_K:

Honestly - I'm sick of this forum. There was a specific question - how to determine if there is a gigantic outburst, by what criteria. If the only criterion is OI, how to get it into the terminal, or how to calculate it.

The answer to this question was a lot of Vysotsky's waffling, high philosophy with accusations of my signals and just plain stupidity.

What is there to talk about?

Anatoly - this is not specifically addressed to you, but simply a reasoning about the pointlessness of all endeavours on the forum.

Well, not completely pointless. The forum helps to find like-minded people and sometimes helps to discover something new. In the post above I suggested to make a dynamic timeframe based on reversals of a zig-zag indicator. So maybe a dynamic timeframe reference point will not necessarily be a Zig-Zag. You can take another signal indicator and make bars based on its signals.
 
Or you could try something even simpler, set the duration of each new bar with a random number generator. I think the picture would be interesting.....
 
Anatolii Zainchkovskii:
The forum helps to find like-minded people and occasionally helps me discover something new. So maybe a dynamic timeframe reference point will not necessarily be a Zig-Zag. You can take another signal indicator and make bars based on its signals.


Implement the above, there are a lot of timeframes already tried.


And judging by the way Sasha is tearing his W shirt, Lambert is also off the mark.

 
Evgeniy Chumakov:


Implement the above, there are a lot of timeframes already tried.


And judging by the way Sasha is tearing his W shirt, Lambert is also off the mark.

I will definitely implement it programmatically. At the moment I'm happy with the visual perception. When you have imagination and understanding of your idea then you can see everything on a naked graph. And the software only needs to collect statistics for a couple of years, to make sure of what the idea is based on.
 
Evgeniy Chumakov:


Exactly. That's where illusion trading comes in... On a transformed chart - almost 100% of trades are in profit, but on real data - so-so...

 
Alexander_K:

There was a specific question - how to determine if there is a giant emission, by what criteria.

As usual, with the help of MO. We learn to predict abnormal spikes in volatility, of course we cannot predict spikes related to socio-political events like Brexit and some war, but "planned" spikes can be a couple of % more accurate than random.

 
Alexander_K:

The answer is simply stupidity.

Stupidity is on your part. You've been answered a thousand times and you keep looking for the keys where the light is, not where you lost them.
 
secret:
By the time you find it out using statistical methods, it will be too late) The best breakdown - stoploss, or breakdown.

I have already written to him, he is looking for meaning in a certain hypostasis, then with a bias towards everything he remembers from high school, then Gann, then Witches, then the Almighty trying to call him to account.

Igor Makanu:

HH: I like to solve all problems by searching for analogies in the physical world - so I read the TV forum, more and more convinced that the search for a mathematical formulaE the market is like searching for a way to guess the next card in a pack of cards in a card game..... but without knowing what the next card in the deck one can often beat the opponent, right? ;)