Woodie's choppy zone indicator - page 4

 

Rename(StepRSI_v5.2 instead StepRSI_v[1].2) and compile these indicators

 

Hi Igor,

TNX for the help!..now everything works properly.

I would like to point to some litlle bug in this indicator. Sometimes it repaints the past better then it was. As you can see on the image were I drew a vertical line, that was a litlle up bar. But no reason to change color because there was no significant step taken to the upside. Price was still lower then a lott of previous bars. But as you can see the color of the indicator changed allready to blue.

If one looks to this situation you would think that you have an ideal entry. If you would be short before the blue bar shows and turn in to long at the close of this first blue bar it makes a differance of 26 pips sooner out of the short and 26 pips quicker in the long. So that makes a total differance of 52pips that the indicator gives the impression that you are better then in reality.

The 2 indicators are the 1.2 and the 1.2a. It happens not every time the indicator changes color but it happend in severall important situations when one overlooks the historie.

friendly greetings..iGoR

Files:
bug.gif  56 kb
 

If the step indicators by IgorAD are variable, then the past data of the indicators will reflect the current conditions of the indicator. The past condition might have been step size "N", but current condition might be step size "N*2" which would give different visual results on past data.

This is not the same as repainting to fit old data, it is adjusting to current conditions. The indicator changes it's variable for the entire data history as it adjusts to the current market fluctuations.

 
spec101:
If the step indicators by IgorAD are variable, then the past data of the indicators will reflect the current conditions of the indicator. The past condition might have been step size "N", but current condition might be step size "N*2" which would give different visual results on past data. This is not the same as repainting to fit old data, it is adjusting to current conditions. The indicator changes it's variable for the entire data history as it adjusts to the current market fluctuations.

spec1, Thank you very much for your quick answer.

But we are still faced with a problem.

As most of us know, the back tester of MT4.0 is absolutly not reliable. So if we want to do some back testing we need to do this manually (wich already is a time absorbing work). If we want to do this manually, we need to look to the indicators in what position they are (long or short) but if we can not rely on what we see then it is again not reliable.

And I think you agree that it is not possible to do forward testing every time one has an idea or strategy. So there probably must be a way to programm this so one can rely on what he sees.

friendly greetings...iGoR

 

Igor, I wish I had a solution for you. You are right that the Metatrader backtesting function is at best misleading-outstanding results do not prove worthy in forward testing, and perhaps poor results in backtesting are worthy strategies for trading.

Maybe the Tradestation tester is reliable? I wonder if IgorAD gets radically different results while testing the same system on these different platforms.

 

iGoR, Maybe if the indicator had an input that permitted the user to determine a step size, like the stepma, then manual testing would be possible. The stepma defaults to "0" to be variable, but the user can choose some size that will set the ma for it's entirety, right?

This will "hold" the past results for visual inspection, but at the expense of the indicator adapting to current conditions. Without the adapting ability the stepma seems to be rather like point-and-figure charting.

 

First of all thanks to Igorad for making these indicators, they are very well done.

I noticed the StepSize is dependant on the number of bars you have loaded into your chart. What do you guys think is an optimal number of bars to use to calculate stepsize on 30 minute charts? I am thinking 2000 but i havent tested enough find a good number....I want to use the same number of bars for all my charts to use these indicators with.

Igorad do you think it would be possible to allow user defined settings for StepChoppyBars_v1? I tried to change some by loading StepMA_v7 to my chart and changing Number of Bars and StepSize, hoping that it would change the calculation for StepChoppyBars, but it didnt.

TIA,

Rupp

 

iGoR, I attached stepchoppy and stepma to charts, then deleted data to simulate market change over the weekend.

I did not get different values for the indicators on past data. Step size had changed, but the indicators held their positions. My little theory is wrong, apologies to igorad.

Rup, I do not know an optimal value. If you are trading Catfx50, then your choice of 2000 is probably fine.

I did go and change the input values to look at differences on stepchoppybars, but I think that igorad's settings are good.

I think that Igorad makes such good indicators that a system could be built, using only igorad's indicators, that would be as close to a "sure thing" system as could be found. The weakness of the system would be the psychological issues of the trader.

 

Hi Spec I tried a similar experiment to you, to see how past data would change based on changing stepsize. I did this by simply loading more bars into my chart (aud/usd for this test). At first I noticed no change at all, even as stepsize increased....however I continued loading data (all the way to 10,000 bars, started at 2000), at this point, I started to notice some very small changes, stepsize had increased from 12 to 16..........in some areas, data had actually improved, repainting an area that would have been a false signal......in other areas, the data was not as good as the original, giving later signals. So I think in most cases, the data will be very consistent....only until stepsize changed significantly did I start to see small changes.

 

What is stepsize actually? Is it number of bars ?