Grail. The puzzle is an interesting theme. - page 9

 
kharko писал (а) >>

I think the lot size distribution should be in reverse order, i.e. from maximum to minimum....

exit, as far as I remember, by trailing stop

You are correct.
Only I suggest the first lot be the minimum,
the second maximum, and then the rest on a descending scale.
First order we have to check if we are going in the right direction, and then we will earn if the market allows us to.
 
Only I suggest the first lot is the minimum,

the second maximum and then in descending order.

Depends on the entry signals, of course, but it probably won't work either. Most of the losses are incurred by the first or second order. If the 3rd-4th orders are clinging to it, then the market (most probably) has moved and will move pretty good. And the 1st-2nd is a run-up with an unpredictable result. So, this approach requires only a smart signalman.

 
SamMan писал (а) >>

Depends on the entry signals, of course, but it probably won't work either. Most of the losses are incurred on the first or second order. If the 3rd-4th orders are clinging, then the market (most probably) has gained some momentum and will move quite decently. And the 1st-2nd is a run-up with an unpredictable result. So this approach requires only a smart signal setter.

I have it on my demo for 2 weeks already. From my experience I say.

You need the right outputs either by the inductor or by the trailing stop.

 

The first transaction is attributed to the subsystem [1]

The second to subsystem [2]

etc.

Each subsystem is characterized by expectation and variance.

If expectation < 0, let's fire the system. In this way we get the maximum number of poses that can be opened.

By the same Vince (optimal f) we can calculate the size of each pose.

IMHO, maximal size will be somewhere in the middle.

( [1] will be characterized by high profit and frequent losing, [2] - more stable profit,

[N] - average profit and moose are about equal, both in size and frequency (maximum depends on output quality), MO is close to 0,

[N+1] - MO < 0 - this order is already unnecessary).

 
angarec писал (а) >>
I downloaded grider2_2. I just did it, but it does not work. I have given my EA permission to start trading. It shows icon that it is active. But it does not trade. What else can I do? How do I start it?

I've got the same thing, I've installed it, I'm trying to test it, no results, other EAs work, but this GRIDER doesn't work.

 
BROM писал (а) >>

I have the same thing - I have installed it and am trying to test it, no results. Other EAs work.

Go into the Expert Advisor code. Change falsh to trai. everything will work.

 
Fibo писал (а) >>

Go into the expert code. Change the faux pas to tray. Everything will work.

I have changed one line where: close all open orders and positions, no result.

 
BROM писал (а) >>

Thanks, I have changed one line where it closes all open orders and positions, but there is no result.

Replace the same line in the Expert Advisor properties and further in the input parameters (in the table where you set the order step, call order, etc.).

 
Fibo писал (а) >>

Replace the same line, in EA properties and further input parameters (in the table, where you set the order step, call order, etc.).

Fibo-thank you-it works.

 
Erics писал (а) >>

The first transaction is attributed to the subsystem [1]

The second to subsystem [2]

etc.

Each subsystem is characterized by expectation and variance.

If expectation < 0, let's fire the system. In this way we get the maximum number of poses that can be opened.

By the same Vince (optimal f) we can calculate the size of each pose.

IMHO, maximal size will be somewhere in the middle.

( [1] - probably will be characterized by high profit and frequent losing, [2] - the profit is more stable,

[N] - average profit and moose are about equal, both in size and frequency (maximum depends on output quality), MO is close to 0,

[N+1] - MO < 0 - this order is already superfluous).

Yes, that would be nice.

Reason: