From theory to practice - page 981

 
Evgeniy Chumakov:
There the distribution of increments turns out to be bimodal......early.

AAAAAAAAAA!!! Genius!!!!!

Except Warlock puts it in a neural network and has the Grail, and you know what to do with it???

 
Alexander_K:

AAAAAAAAAA!!! Genius!!!!!

Except Warlock puts it in a neural net and has the Grail, and you know what to do with it???

I don't know. All I've done so far is put the tics together a certain way.
 
Evgeniy Chumakov:
I don't know. I've only built ticks in a certain way so far.

Shit, Zhenya - don't lose that metod. 100% I'm telling you - it's the same schedule for Koldun and a couple of other people whose names I won't mention.

If you make the Grail, open the signal, please.

P.S. I myself have no idea what to do with it.... Even if I knew how to get one.... Try to ask neural networkers to run such a series in their models.

 
Evgeniy Chumakov:
That's where the distribution of increments turns out to be bimodal...... ears.

It is bimodal because increments are equally likely to be up or down.)

It's also probably because your vertical scale has room for zero increments.)

 
secret:

It is bimodal because increments are equally likely to be up or down.)

And probably also because your vertical scale has room for zero increments)

I can't show the originals, without the permission of the discoverers. But, their increment distributions look like this:

Can't you get anything out of them? Then why were these people bragging to me in the PM?!

 

About "extract" is a separate issue, but if the middle is zero increment, then this "bimodality" is apparent)

 
secret:

About "extract" is a separate issue, but if the middle is zero increment, then this "bimodality" is apparent)

Yes - zero... But, the triangles can be seen more than clearly. I'll have to think about it...

 

Forum on trading, automated trading systems and trading strategy testing

Miracle Grail... Myth or Reality!

Wizard2018, 2018.12.24 13:50

Second point. it is betterto enter from a pullback against the local movement, in the direction of the current trend defined in the first point. That way the stop is shorter and the potential profit is bigger. And hence the most important ratio for the system - profit/loss is higher.

So, we need to look for the indicator/method that allows us to easily detect the pullback in the current trend.

In other words, we should not just take an indicator, optimize it for some timeframe/instrument, see that the pullbacks are good and calm down. And then "the market changed" It's because the "indicators"/method grabbed the "form", a specific"realization" and a huge number of possible ones, but failed to "generalize".

It's unlikely to work that way. We should try to find a universal method, just likeEMA from the first itemalways and everywhere works correctly , so for the second "item" we need to find something that works as well. I can't suggest anything specific, I can only say that there is an excellent method for determining the pullback in the current trend that works anywhere and anytime. Info 100%, not a fairy tale. Search and find.


Here, strangely enough, our comrade has a point.

 
Alexander_K:

Yes - zero... But, the triangles are more than clearly visible. I'll have to think about it...

Without knowing what those increments are (how filtered, etc.), of course, it's hard to think. But still, why not try to see behind this distribution of increments the situation: a rebound or a level break. The modules of expected increments in these cases seem to me to grow as they move away from the broken/unbroken level, which looks like a failure of the module of expected increments at the level (see figure). Here would be the process memory.

P.S. Am I right in guessing that the figure is not about the distribution of increments, but their absolute values?

 
Martin Cheguevara:

oddly enough, our comrade here has a point.

Geez, Martin - that comrade always has a point. I'll be glad if his posts help you.

Even if my Grail doesn't work out, maybe you and Zhenya will... I do not need your TS, I need a trading signal from you, to which I would easily subscribe.

I believe and wait. Amen.