One for all. General exeprt. - page 3

 

Inside the bar...

What's a bar? What's in it that needs to be explored? For example, an hourly bar. It's just a 60 minute bar.

The hour bar is notable only because its start time is the same for all traders. But note: you could theoretically shift the start time of the hour bar by half a period and the bars would look very different, in spite of the fact that the tick history would be completely preserved.

What can be squeezed out of a bar at all?

For example, we can try to determine the characteristic ultra-low frequency of price movements during a period (for example, a day or a week), but then most likely it will appear that the length of one wave of oscillations does not coincide with any bar period (for example, the period will appear as 27 min, 43 minutes or 2,5 hours).

I think a bar is only useful in that it allows you to observe the market on a given scale. Just to look at and get a general idea.


 
SK. писал (а):


If anything is to be discussed, first of all - the idea itself, and not on the basis of indicators (which already have some idea in them), but on the basis of logical reasoning and common sense. If you find the idea and intuitively understand that it can work, then select (or make) indicators, filters, etc. for it.


Can you imagine MTS without indicators and filters? And you think that nothing happens inside a bar? Minutes?
 
SK. писал (а):


I think the bar is only useful in that it allows you to observe the market at a given scale. Just to watch and get a general idea.


Why do you think that the timeframe from which you are most comfortable observing the market allows you to actually observe it?
 
SK. писал (а):

Inside the bar...

What's a bar? What's in it that needs to be explored? For example, an hourly bar. It's just a 60 minute bar.

The hour bar is notable only because its start time is the same for all traders. But note: you could theoretically shift the start time of the hour bar by half a period and the bars would look very different, in spite of the fact that the tick history would be completely preserved.

What can be squeezed out of a bar at all?

For example, we can try to determine the characteristic ultra-low frequency of price movements during a period (for example, a day or a week), but then most likely it will appear that the length of one wave of oscillations does not coincide with any bar period (for example, the period will appear as 27 min, 43 minutes or 2,5 hours).

I think a bar is only useful in that it allows you to observe the market on a given scale. Just to look at and get a general idea.


I had one order for an Expert Advisor that was without a single indicator. And then there are trading systems that do not use indicators, so they are all based on statistics!
 
ExpertTrader писал (а):
There is nothing else about the indicator, at least here:Acceleration/Deceleration.
I want to know the opinion of traders who have been using this indicator for a long time.

I have made a dull arrow, maybe some inexperienced traders (in case they have looked through it) will understand what I'm talking about.
I have removed unnecessary arrows, but there are still a lot of signals.

//+------------+-----------------------------------------------------+
//| v.28.09.06 |                                     UpAndDownAC.mq4 |
//+------------+                 "Индикатор по теме by ExpertTrader" |
//|  эскизно   |                    Bookkeeper, yuzefovich@gmail.com |
//+------------+-----------------------------------------------------+
//+------------------------------------------------------------------+
#property copyright ""
#property link      ""
//+------------------------------------------------------------------+
#property indicator_chart_window
#property indicator_buffers 8
#property indicator_color1  Green
#property indicator_color2  Red
//#property indicator_color3  Silver
//#property indicator_color4  Black
//#property indicator_color5  Green
//#property indicator_color6  Red
//#property indicator_color7  Green
//#property indicator_color8  Red
//----
//extern int    FilterPeriod  =24; 
//extern int    Satisfaction  =5;
//extern bool   Pipsota       =true;
//----
double      StartUp[];
double      StartDown[];
double      ArrAC[];
//double      CountUp[];
//double      CountDown[];
//----
int         prevBars=0, xBar;
double      Price1, Price2;
bool        SignalUp=false, SignalDown=false;
string      ThisName;
bool        First;
//---------------------------------------------------------------------
void deinit()
{
//  Comment("");
  return;
}
//----
int init()
{
int    draw_begin;
   draw_begin=10;
   IndicatorBuffers  (8);
   SetIndexBuffer    (0,StartUp);
   SetIndexStyle     (0,DRAW_ARROW);
   SetIndexArrow     (0,233);
   SetIndexBuffer    (1,StartDown);
   SetIndexStyle     (1,DRAW_ARROW);
   SetIndexArrow     (1,234);
   SetIndexBuffer    (2,ArrAC); 
   SetIndexStyle     (2,DRAW_NONE);
   SetIndexEmptyValue(0,0.0);
   SetIndexEmptyValue(1,0.0);
   SetIndexEmptyValue(2,0.0);
   SetIndexDrawBegin (0,draw_begin);
   SetIndexDrawBegin (1,draw_begin);
//   SetIndexDrawBegin (2,draw_begin);
//   SetIndexBuffer    (6,CountUp);
//   SetIndexStyle     (6,DRAW_ARROW);
//   SetIndexArrow     (6,159);
//   SetIndexBuffer    (7,CountDown);
//   SetIndexStyle     (7,DRAW_ARROW);
//   SetIndexArrow     (7,159);
   ThisName=Symbol()+"_M"+Period()+"_";
   First=true;
   return(0);
}
//---------------------------------------------------------------------
int start()
{
int limit,i;
int counted_bars=IndicatorCounted();
   if(counted_bars<0)  return(-1);
   limit=Bars-counted_bars;
   if(First==true)
   {
      limit=limit+4;
      for(i=0;i<5;i++,limit--) ArrAC[limit]=iAC(NULL,0,limit);
      First=false;
   }
   for(i=limit;i>=0;i--) MainCalculation(i);
   return(0);
}
//---------------------------------------------------------------------
void MainCalculation(int Pos)
{
int    Shift, i;
   ArrAC[Pos]=iAC(NULL,0,Pos);
   if(SignalUp==false) {
   if( ArrAC[Pos]>0 &&
      ((ArrAC[Pos]>ArrAC[Pos+1] && ArrAC[Pos+1]>ArrAC[Pos+2]) ||
       (ArrAC[Pos]<ArrAC[Pos+1] && ArrAC[Pos+1]<ArrAC[Pos+2] && ArrAC[Pos+2]<ArrAC[Pos+3])
      ) 
     ) { StartUp[Pos]=Low[Pos]; SignalUp=true; SignalDown=false; }
   else StartUp[Pos]=0;
   }
   if(SignalDown==false) {
   if( ArrAC[Pos]<0 &&
      ((ArrAC[Pos]<ArrAC[Pos+1] && ArrAC[Pos+1]<ArrAC[Pos+2]) ||
       (ArrAC[Pos]>ArrAC[Pos+1] && ArrAC[Pos+1]>ArrAC[Pos+2] && ArrAC[Pos+2]>ArrAC[Pos+3])
      ) 
     ) { StartDown[Pos]=High[Pos]; SignalDown=true; SignalUp=false; }
   else StartDown[Pos]=0;
   }
   return;
}
//---------------------------------------------------------------------
//---------------------------------------------------------------------
//---------------------------------------------------------------------
 
SK. писал (а):

If anything is to be discussed, first of all - the idea itself, and not on the basis of indicators (which already have some idea in them), but on the basis of logical reasoning and common sense. If you find the idea and intuitively understand that it can work, then select (or make) indicators, filters, etc. for it.


Also not bad. Any suggestions?
 
ExpertTrader писал (а):

Personally, I see AO and AC as indicators that can indicate an undesirable entry, i.e. you can't go there, but the signal to enter by them is really "worthless". I do not insist on AC and AO, I want everyone to suggest their own ideas, not without proof of course. And in general I was just driving along and thought: I should first make an analysis based on intra-bar movement statistics for large periods, which I will do now.


Yep, then AO is a good filter, ready to keep out when it shouldn't.
I daresay there is movement within a large bar and it is no different from movement within a small bar. Tell me, what do you specifically expect from the statistics? What hypothesis? Maybe someone already knows the answer to what you are looking for.
 
SK. писал (а):

Inside the bar...

What's a bar? What's in it that needs to be explored? For example, an hourly bar. It's just a 60 minute bar.

The hour bar is notable only because its start time is the same for all traders. But note: you could theoretically shift the start time of the hour bar by half a period and the bars would look very different, despite the fact that the tick history would be completely preserved.

.....
I think the bar is only useful in that it allows you to observe the market at a given scale. Just to look at and get a general idea.


Well, SK (Sergey Kovalev, if I'm not mistaken) has already joined the discussion. I thought about it myself: to shift the chart by half a period of a candle and see how the chart and indicators will change. But since the paradigm of using standard indicators has been already rejected, I lost interest in it. I think it will prove that relying on indicators in trading is inappropriate (i.e. not my interest, but the chart shift).
And one more thing. MQL4, MT4 and its history tester not only provide an opportunity to build your own strategies and check them on the real market, but also debunk many popular, advertising and even classical illusions and misconceptions, common to people who see (or show by example) a small piece of history. The tester sees the whole story and immediately presents us with a lot of examples where our rule does not work, but leads to the exact opposite results. In other words, scientifically speaking, there is no correlation between our hypothesis and the real market behavior. This is also confirmed by the statements above in this thread.
 
Also, it seems to me that with the involvement of many small traders in trading in recent years and the emergence of a network of DCs around the world, the forex market has changed a lot. And the indicators that worked in the slow market days (without the current personal units) are no longer relevant. Today we need new benchmarks in trading.
 
Gorillych писал (а):

Why do you think that the timeframe from which you are most comfortable observing the market allows you to actually observe it?
Gorillych wrote (a):

Can't you imagine MTS without indicators and filters? And you think nothing happens inside a bar? Minutes?


It's like in the joke:
- You think badly of me!
- I don't think of you at all.

As for me personally, I use all TFs for analysis. The question was not about what is convenient or not, but about the fact that in my perception a bar is an entity that makes no practical sense to analyse. Both oscillation time and duration of bars are based on developers' will and, in terms of physics, the absolute values of these characteristics are useless.

I can imagine MTS without indicators and filters. It's not about filters, indicators, libraries, Expert Advisors.
At the root of any technology must be the underlying idea. Until there is no idea, it is silly to discuss technical means (for its implementation). When an idea appears, depending on the requirements of the idea: if we need an indicator - use (or create) the indicator, if we need a filter (a library, functions, DLL) - create the filter.

-------------------

A little bit about ideas.

The market is a complicated phenomenon. Much more complicated than many people think.
I'm sure when speaking about indicators we can find only partial solutions for a narrow range of elementary tasks.
If we are talking about creation of MTS, then at least it is necessary to process large amounts of information.

I will hardly be able to participate in this discussion.
I have only one idea of my own (which, by the way, doesn't use built-in indicators), and I still have some work to do.
And I don't have a ready (not so much a solution, but even) suggestion.
I've been in the market since 2002. During this time I've written hundreds of different Expert Advisors and indicators in MQL2. However, I have only one written here, and it's a semi-intuitive one based on empirical formulas. The rest of them are not worth any attention. I do not want to criticize anyone.

I can still tell you what not to do.
Why? Just because it's all been done before.
What you should do... - I'd love to hear something meaningful.

Reason: