Tell me how to prepare a martin so as not to drain... - page 2

[Сергей Криушин]

Renat Akhtyamov:

"so that you don't lose..." - that is, how do you get a profitable EA?

That's the question on everyone's mind here

;)

Oh, yeah, I didn't understand what I wanted to get modestly - it's sure to turn out to be a grail... Of course there is no answer and is not expected, but yes, at the reversals two minuses spoil everything... I thought to filter them out somehow, but nothing comes to mind, except the anti-martin... i'll try another execution delay... or put both martin and anti-martin and let them slow each other down... see what happens...))

[Vitaly Muzichenko]

Сергей Криушин:
I don't understand what I wanted to get modestly - it could be a grail for sure... I don't have an answer and don't expect one, but at reversals two minuses spoil everything... I thought to filter them out somehow, but nothing comes to mind, except the anti-martin... i'll try another execution delay... or put both martin and anti-martin and let them slow each other down... see what happens...))

It's a "Chumbley" - a flat-out loss due to a lack of margin).

[Сергей Криушин]

Vitaly Muzichenko:

It's a "Chumbley" - a flop in a flat because of a lack of margin)

What can't you think of... again, it's a bicycle, so you go around in circles to get around forced losses...((

[Renat Akhtyamov]

aleger:

Wrong approach - should beOrderProfit()>Spread

fine, but what if the volume is different?

not yours? :

https://www.mql5.com/ru/forum/277549/page2#comment_8571475

[Sergey Vradiy]

Сергей Криушин:
Can you please tell me how to skip 1 or 2 negative deals (not to apply Martin to them) that usually occur at an unsuccessful start flat... or rather an unsuccessful entry into a flat by the first trade or it often happens on a reversal to a new trade...

Obviously, the challenge is not just to skip, but to determine whether to skip or not. I.e. we need a preliminary theanalysis to distinguish between a false-break and a true breakdown. If we have a flat channel, then upon a false breakout the price returns to the channel borders. If it is a true breakout, the price breaks through the boundary, then comes back to it, but then bounces from it again. This is the check that should be performed to include further code in case of true breakout.

[aleger]

Renat Akhtyamov:

fine, but what if the volume is different?

not yours? :

https://www.mql5.com/ru/forum/277549/page2#comment_8571475

Not really my thing...

[Mickey Moose]

what if we make a martin using this algorithm

open all trades with sl and tp

1. at two points the third opens with lot x2

2. in case of a loss -50% lot

3. if the next losing trade also takes a loss, then lot -50% of point 2

in case of two profitable steps one step higher

[Renat Akhtyamov]

Mickey Moose:

what if we make a martin using this algorithm

open all trades with sl and tp

1. at two points the third opens with lot x2

2. in case of a loss -50% lot

3. if the next losing trade also takes a loss, then lot -50% of point 2

in case of two profitable steps one step higher

there's a tester

[Mickey Moose]

Renat Akhtyamov:
there's a tester

I'm a long way from that, I've forgotten how to edit my code.

maybe someone wants to make one for themselves

[Сергей Криушин]

Mickey Moose:

what if we make a martin using this algorithm

open all trades with sl and tp

1. at two points the third opens with lot x2

2. in case of a loss -50% lot

3. if the next losing trade also takes a loss, then lot -50% of point 2

in case of two profitable steps one step higher

No, I figured it out: we need such an excursion that should trigger the signal after the reversal, i.e. we need to cut off all high-speed false signals, which are often triggered on small TFs exactly at the moment of the reversal, we need a delay in triggering, which happens only on large TFs, because there the signal comes when the middle of the trend is already coming... Or the next reversal is close but reliable... so it seems... Again, the question is where to find such a reliable robust Ex... the circle is closed...))