Let's build a mini grail!? - page 34

 
Petros Shatakhtsyan:

Don't try very hard to show any nonsense here.

Here is an article from the developers of MQ that explains how you can get the "Grail":https://www.mql5.com/ru/articles/239

It shows how to do it.

You are not being practical. It's just a programmers toy, it's obvious and useless!

Programmers don't make money in the market...
 
Михаил:

You are not practical. It's just a programmers' toy, it's obvious and useless!

Programmers do not make money in the market...

Why don't they make money because you are not a programmer ?

 
Petros Shatakhtsyan:

Don't try too hard to show some nonsense here.

Here is an article from the developers of MQ that explains how to get the "Grail":https://www.mql5.com/ru/articles/239

It shows how to do it.


Also read:

In 2011 there was no "All ticks based on real ticks" mode yet

There are 100++1 ways in the tester other than this to show the crooked grail. ))

 
Михаил:

But since you're showing it, it's not a grail for some reason))

It's a candidate for further testing - checking it on demo, putting it on real.

 
Serqey Nikitin:

No, it's not a grail!


The grail works with the same settings on ANY pair!

I highly doubt this. Different pairs are different samples from the population, so their estimation parameters are different. It's like finding the parameters of a Regression line on one sample and then fitting them to another, the error will be immediately visible.

 
Serqey Nikitin:

No, it's not a grail!

So making a test report on this pair on a six-month site is very easy:

The dependence of the stability of the estimation on the sample size is a task I'm still going to do, but haven't got around to it yet. I have a feeling that 3-4 months will be enough for some, and for some a whole year is needed. We'll see what to guess.

 
Михаил:

But since you are showing it, it means that for some reason it is not a grail.)

So why did you open this thread even if you admit that no one will show you the grail? Is it for nothing?

 
How many Grail branches have scattered (:o)
 
Martingeil:
How many Grail branches are there (:o)
People are searching for the Truth...
 
Serqey Nikitin:

No, it's not a grail!

...


The grail works with the same settings on ANY pair!

Any abstraction expressed in words can be studied down to a "non-abstraction" i.e. an ordinary entity (perhaps unattainable, marginal, but understandable)

it is possible to assume the weak sign - if we decompose a"grail" trading series into limits, then removal of any two adjacent ones does not cause drawdown and does not change its "grail".

If the premise is correct, then the unexpected conclusion is: grail-like algorithms trade on an expanding zigzag.