Damned Martin - page 13

[Mickey Moose]

Sergey Gritsay:
Explore, I don't mind.

https://www.mql5.com/ru/charts/6468353/audusd-m15-metaquotes-software-corp

Prefitting on any pair without changing within the code

[Vitalie Postolache]

trader781:

https://www.mql5.com/ru/charts/6468353/audusd-m15-metaquotes-software-corp

pre-fit on any pair without changing within the code

Cool. Does it draw the same beauty outside of the fit period? And on the demo?

[Mickey Moose]

Vitalie Postolache:
That's cool. And outside the refinement period the same beautiful drawings? And what about the demo?

The idea here is that this code trades at a profit of about 20-40% per annum (tester) Of course not much, but it survived and did not lose money.

The rest is to finalize all parameters, but it is everyone thinks it over

It is not good for the demo, I have too much work to do there.

[Vitalie Postolache]

trader781:

The idea here is that this code trades at a profit of about 20-40% per annum (tester) Of course not much, but it survived and did not lose.

The rest we need to fine tune.

The point is that almost any EA does not fail in the tester for a period with optimization, if there is something to tweak. But on forward or real time most of them fail.

That's why we should adjust on one part of the history and test on another.

[Mickey Moose]

Vitalie Postolache:

The point is that almost any EA does not lose in the tester on a period with optimisation, as long as there is something to adjust. But on forward or real time most of them fail.

That is why we should adjust on one part of the history and test on another.

I don't know, I've never traded robots at all and don't intend to.

[Vitalie Postolache]

trader781:
I don't know, I've never traded robots at all and don't intend to. it's just the fact that he is able to survive and not lose in a year.

So was it adjusted in the tester optimiser for the year it "didn't fail", or with default parameters? And for another year of testing? Will it survive as well?

[Mickey Moose]

Vitalie Postolache:
So, was it adjusted in the optimizer of the tester during the year when it "failed", or with default parameters? And for another year of testing? Will it survive as well?

I do not know how to work with the optimizer

I just took the data of my bot which I've been writing for the second month and ran it on a dozen charts, and this is the average result.

But it was not designed for them from the very beginning. But it was not originally designed for it.

[Sergey Gritsay]

Added dynamic TP setting, calculated from ATR indicator value.

[Ivan Butko]

Sergey Gritsay:
added setting of dynamic TP, calculated from ATR indicator value.

Sergey, what is the point? We don't have a SL, there is no need to calculate a different TP every time, because we don't have a trading strategy as such. There is only one logical option - to set the TP so that it would take away as much as possible any movement. For example, if we have a flat on H4, the price goes back and forth and we can fill our wallet on such movements and an order with a larger TP than the flat will simply miss such moves.

You need to set TP in such a way that it does not eat the spread, that it does not miss small movements and that it corresponds to the timeframe. It is enough to do some digging and find an appropriate one. But you don't have to change it all the time. If you have better results with it, please send me a screenshot of the results.

[Ivan Butko]

Sergey Gritsay:
Added a dynamic TP, calculated from the value of the ATR indicator.

USDCAD from 2010 to date. The most important - from 100(!) dollars.

For eu pours every year.

I have also added a dynamic TP (calculated from the ATR indicator value) for the yen.