Not looking for investors or trying to sell an expert :) just interested in opinions - page 12

[Rid]

Corwin_Volot:

i've started a new expert here is the initial trial

or is that bad too :( what to do with these drawdowns...

I solved this problem a long time ago. My aim is to make long and short positions independent of each other. And then you'll see what happens - the profit of the current long compensates for the loss of the current short and vice versa. Perhaps this will be the solution in your case. I got the result this way (well almost) - here in my last post here is the chart: 'Question for MT experts'.

[[Deleted]]

Mathemat:

Corwin, 1) if you think that large drawdowns are not a problem for you, then go for real and trade. You asked the question for a reason, but to ask for advice?

P.S. Heroism is just sitting, not taking a loss.

1) That's what I'm trying to figure out - "why a drawdown greater than so much is ALSO bad".

I do not quite understand why the Expert Advisor that goes through 1999-2007 with a profit in the test is considered bad only because the drawdown is higher, for example 20%. As far as I understand the drawdown is perceived as a percentage of risk, but if the deposit is calculated with allowance for drawdowns and the algorithm is ready for them the risk is lower than in many other Expert Advisors with lower drawdown.

Maybe you can explain it to me :(

2) But brave heroes always go the other way around :) Especially if losses are insured by the algorithm.

[[Deleted]]

goldtrader:

Corwin_Volot:

That is, create a problem and then heroically solve it :)

But thanks for the advice, I'll give it a try.

And 22 deals is nothing for statistics. A fitting and self-defeating. And you have to test with a constant lot. This is when you start working on real account (if you get to it), you may use MM.

So I wrote that this is the initial test case. I just needed to test the idea itself, I managed to make only a model of the algorithm in a few minutes and set the most forgiving conditions, in order not to bother with the parameters. Now when the test showed that the algorithm at least makes something, you can spend a few evenings to finalize it.

If I was involved in fitting I would not have sent six of my eight Expert Advisors to the trash :) Now I'm working on the ninth. And then something sensible will come out.
MM - what is it?

[[Deleted]]

rid:

Corwin_Volot:

i've started a new expert here is the initial trial

Or that's too bad :( what to do with these drawdowns...

I solved this problem long time ago. Make long and short entries independent of each other. And then you'll see what happens - the profit of the current long compensates for the loss of the current short and vice versa. Perhaps this will be the solution in your case. I thus (well, almost) got the result - here in my last post chart: 'Question for MT experts'.

I'll try the independent ones.

I tried to open in different directions simultaneously, but my algorithm has reduced the effect to zero because I don't really care where I open :)