Is martin so bad? Or do you have to know how to cook it? - page 9

 
notused:

In general, for anti-martingale I have not found any application (but that does not mean there is none :))

With limited (no matter how much) capital for any MM there is an application only when the initial system is winning. And this statement has "proven" status, no imho.
 
alsu:
With limited (no matter how much) capital, there is a use for any MM only when the initial system is winnable. And this statement has "proven" status, no imho.
a system may be loss-making (i.e. not have positive expected payoff), but in combination with a profitable system -> may produce more profit than just a profitable system. A necessary condition for this is negative correlation of systems (when one wins, the other loses), so your statement is only true for "single" isolated trading systems.
 
alsu:
With limited (no matter how much) capital, there is a use for any MM only when the initial system is winnable. And this statement has "proven" status, no imho.
If this reasoning is correct, then we can say that any system is losing, because there will come the drawdown, which the deposit will not survive. If MM improves performance of profitable system, then why it cannot improve performance of loser system? Imho, the worse the system, the deeper the drawdown. Who proved it and where? Can you give me the link?
 
220Volt:
If you think like that, then you can say that any system is a loser because there will come a drawdown that the depo will not survive. If MM improves performance of profitable system, then why can't it improve performance of loser system?
Maybe the drawdown will happen later.)
 

A simple example (MT4 code):

#define UP     1
#define DOWN   -1
//+------------------------------------------------------------------+
//| script program start function                                    |
//+------------------------------------------------------------------+
int start()
  {
//----
      MathSrand(54);
      int handle = FileOpen("File.csv", FILE_CSV|FILE_WRITE, ";");
      int handleMM = FileOpen("FileMM.csv", FILE_CSV|FILE_WRITE, ";");
      
      double point = 0, mmPoint = 0;  // По значениям этих переменных, строим графики.
      int curVal;
      int lastVal = 0;
      int direct;
      double startLot = 1;
      double curLot = 1;
      double step = 1;
      for(int i = 0;  i < 31000;  i ++)
      {
         //begin//  Блок установки направления текущей точки график (относительно прошлой). Добиваемся необходимого МО.
            curVal = MathRand();
            if(curVal > lastVal)
            {
               curVal = MathRand();
               if(curVal > lastVal)
                  direct = UP;
               else
                  direct = DOWN;
            }
            else
               direct = DOWN;
            lastVal = curVal;
         //end//

         point += startLot * direct;   // Значение в на графике с постоянным лотом.
         mmPoint += curLot * direct;   // Значение на графике с непостоянным лотом.
         
         //begin//  Изменяем лот, в зависимости от исхода, для графика с непостоянным лотом.
            if(direct == DOWN)
               curLot += startLot * step;
            else
               curLot = startLot;
         //end//         

         FileWrite(handle, DoubleToStr(point, 3));       // Пишем в файл.
         FileWrite(handleMM, DoubleToStr(mmPoint, 3));   // --//--
      }
      FileClose(handle);
      FileClose(handleMM);
      
//----
   return(0);
  }
//+------------------------------------------------------------------+

Fixed lot (point) chart:

Not constant lot (mmPoint):

On the first chart the mathematical expectation is clearly <0, but MM has pulled the system to profit.

 
220Volt:

A simple example (MT4 code):

Constant lot chart:

Not a constant lot:

On the first chart the mathematical expectation is clearly <0, but the MM has pulled the system into a profit.

What is the longest duration of "losing" trades for the above figures (if I am not mistaken, the D'Alamber principle applies)?
 
notused:
What is the longest duration of "losing" trades for the above figures (if I am not mistaken, D'Alamber's principle applies)?
Biggest series of 8. Not familiar with D'Alamber's principle, so maybe )).
 
220Volt:

A simple example (MT4 code):

Constant lot chart:

Not a constant lot:

On the first chart the mathematical expectation is clearly <0, but MM has pulled the system to profit.

I beg your pardon, I was drawn into stylistics of 18-19 centuries: "You, my dear, should entertain the townspeople with curious card opuses in a marquee, but at a gaming table in a fair campaign for such liberties may be a bronze candlestick on the head (on its motherland, you crazy bastards!!!)".
And if it's a little simpler, then the question is: why display pictures that have no meaning? For example I did not see any initial deposit or time interval in the pictures, so one can make various conjectures (the only question is why?), maybe the MM system did not take a profit, but just gave a push before it died. You are talking to me in the realm of subjectivity ...
And speaking of the culture of communication, I'm convinced that it's mauvais ton to expose code in this form, absolutely no enthusiasm to understand other people's "kurval and kurloty", when the author's thoughts are not commented elementary (even if it's simple and short, I mean code)... All the more, there are enough good examples in this regard, t. that, culture is more the rule than the exception.
P.S. Nothing personal...
 
220Volt:
Biggest series of 8. Not familiar with D'Alamber's principle, so maybe )).

It does seem to be one after all. And it is profitable if most (with nuances that I'm too lazy to get into) of the losing streaks do not exceed 4 consecutive losses at a win rate of twice the initial bet.

Well, having 8 losses says you should have had at least 45 times the equity of the initial bet.

 
Wangelys:
I beg your pardon, I was drawn into the style of the 18th and 19th centuries: "You, my dear, should entertain the bourgeois with curious card opuses in the marquee, but at the gaming table in a fair company for such liberties may be a bronze candlestick on the head (on its dear, oh my God!!!!)".
And if it's a little simpler, then the question is: why display pictures that have no meaning? For example I did not see any initial deposit or time interval in the pictures, so one can make various conjectures (the only question is why?), maybe the MM system did not take a profit, but just gave a push before it died. You are talking to me in the realm of subjectivity ...
And speaking of the culture of communication, I'm convinced that it's mauvais ton to expose code in this form, absolutely no enthusiasm to understand other people's "kurval and kurloty", when the author's thoughts are not commented elementary (even if it's simple and short, I mean code)... All the more, there are enough good examples in this regard, t. that, culture is more the rule than the exception.
P.S. Nothing personal...

I was counting on that whoever is interested will get into the code and understand the point. If you start to describe and comment on everything, it will be very voluminous, I don't want to do that.