Statistics combined with system. Profitable? What do you think | Forex Factory

HeroPoker_HeroPoker扑克_HeroPoker德扑圈官网

Attachments: Statistics combined with system. Profitable? What do you think

Statistics combined with system. Profitable? What do you think

Trading system: wick system 1.00
Type: trend and ranged
Timeframe: M1 - D1. In this example we take the H1 chart.
Entry:
- when: in the H1 chart buy or sell on every candle (this means, every hour you will open an order at exactly close of previous candle).
- buy: if lower wick of previous candle is bigger than the upper wick.
- sell: if lower wick of previous candle is smaller than the upper wick.
- TP/SL: 50 pips

Statistics:
If statistics say that wicks determine direction, this trading system should be profitable. I ask the readers of this thread to test this theory.

What do you think: will this system be profitable?

______________________
Discuss test case:
______________________

1. (stated by Tzamo) Find out if the wicks and candles are relative in the different timeframes of a particular symbol. (For example (Random numbers) in the H1, the candle bodies range from 100 to 200 units and the wicks are 40 to 80 units. While in the 15 min, the candle bodies are 25 to 50 units and the wicks are 10 to 20 units. If that is true, then the candle bodies are relative to the wicks and to run the EA on the other time frame just need to adjust the bodies/ wick limits.)
-- test case: select several different situations withing the symbol, like: 10 events of steep up trends, 10 events of steep down trends, 20 events of ranging, events with lots of spikes... and so on and then run the EA on each of those and see if it passes.

2.

3.

4.

5. [let's discuss and add more ideas]
Join our skype group.
Interesting! Here is a simple EA coded to just do that.
I've run couple tests and optimizations and i dont think going for 50 SL and 50 TP is working out very well
But... lets see...
Attached File(s)
File Type: ex4 wicksys.ex4   8 KB | 548 downloads
@fauzaanu on Telegram
Interesting! Here is a simple EA coded to just do that. I've run couple tests and optimizations and i dont think going for 50 SL and 50 TP is working out very well But... lets see... {file}
Thank you for the EA. I've done some testing with it. 72% seem to be short positions and 27% are long positions (h1 chart). Right now I'm optimizing the best tp/sl ratio and timeframe in strategy tester.

Could you update it with the following rules:
- add parameter: anti-trading (opposite) on/off: on means: selling (with the buy rules) instead of buying (argument: see results in image: 72% of the losses were sell)

- add parameter: trade on swing: [on/off]. On means:
(add parameter)*only trade [5] candles after previous higher high [on/off]
*only trade [5] candles after previous lower high [on/off]
*only trade [5] candles after previous lower low [on/off]
*only trade [5] candles after previous higher low [on/off]

- add parameter: [5] candles (argument: market phases have different effects on candle wicks)
- add parameter: duration: for [8] following candles

- add parameter: [on/off]only if previous candle is >150 times bigger than previous [4] candles
- add parameter: [4] candles

- add parameter: [on/off]only if previous candle is >[10] pips
- add parameter: [10] pips

- add parameter: [on/off]only if previous candle lower wick >3 pips
- add parameter: [3] pips

- add parameter: [on/off]only if previous candle lower wick is >150% bigger than upper wick
- add parameter: [150] percentage
Attached Image(s) (click to enlarge)
Click to Enlarge

Name: wicksys1.00.png
Size: 118 KB
Click to Enlarge

Name: Strategy tester.png
Size: 120 KB
Join our skype group.
you will have a good chance to be the first man or woman who touch the sun than finding a holy grail for this game a holly grail simply mean a system that work on fx, doesn't mean its a system with 0 loss so you will just lose in this world as far as i know till this day, but ofc me and you will always hope that i was wrong ! cheers
Thank you for your insight. Did you test it and what were your results? What can you add to the system?
Join our skype group.
Interesting! Here is a simple EA coded to just do that. I've run couple tests and optimizations and i dont think going for 50 SL and 50 TP is working out very well But... lets see... {file}
In Daily chart the results are: see image. (last 4yrs, TP/SL 500, every tick, EURUSD, 1% DD)
Attached Image (click to enlarge)
Click to Enlarge

Name: winwin.png
Size: 97 KB
Join our skype group.
Some would say - body in direction of profit and wick in direction of loss.
9 pairs 1M TF TP 20 SL 5000 lets see for another hour
Attached Image (click to enlarge)
Click to Enlarge

Name: Screenshot1.png
Size: 196 KB
- add parameter: anti-trading (opposite) on/off: on means: selling (with the buy rules) instead of buying (argument: see results in image: 72% of the losses were sell)
Done

- add parameter: trade on swing: [on/off]. On means:
(add parameter)*only trade [5] candles after previous higher high [on/off]
*only trade [5] candles after previous lower high [on/off]
*only trade [5] candles after previous lower low [on/off]
*only trade [5] candles after previous higher low [on/off]
Done


- add parameter: [5] candles (argument: market phases have different effects on candle wicks)
Done


- add parameter: [on/off]only if previous candle is >[10] pips
- add parameter: [10] pips
Done

- add parameter: duration: for [8] following candles
Is this a new parameter? Is it the duration of the trade? or ...?

- add parameter: [on/off]only if previous candle is >150 times bigger than previous [4] candles
- add parameter: [4] candles

- add parameter: [on/off]only if previous candle lower wick >3 pips
- add parameter: [3] pips

- add parameter: [on/off]only if previous candle lower wick is >150% bigger than upper wick
- add parameter: [150] percentage
These three i havnt included yet

Im hoping there are some other people who can code in here as well who can add into this, and so im posting the mq4 file
Attached File(s)
File Type: mq4 wicksys.mq4   7 KB | 589 downloads
@fauzaanu on Telegram
I am testing the updated ea now.

Below parameters are not related to the duration. Below parameters are conditions for the entry.

- add parameter: [on/off]only if previous candle is >200 percentage bigger than previous [4] candles
- add parameter: [200] percentage
- add parameter: [4] candles

- add parameter: [on/off]only if previous candle lower wick >3 pips
- add parameter: [3] pips

- add parameter: [on/off]only if previous candle lower wick is >150% bigger than upper wick
- add parameter: [150] percentage

Conclusion:
So it means: if this parameter is turned on "only if previous candle lower wick >3 pips"
then the EA only opens a trade at candle close if previous candle wick is bigger than 3 pips. Every trade entry must be bigger than 3 pips wick, otherwise EA will not open the trade and waits for a new 'candle close' until there is a candle wick bigger than 3 pips. So the EA does not open orders every hour, because not all wicks are bigger than 3 pips.

The EA will only open an order if:
- the candle wick is bigger than 3 pips.

Same with 150 percentage and 200 times bigger.
The EA will only open an order if:
- previous candle is >200% bigger than previous [4] candles

The EA will only open an order if:
- if previous candle lower wick is >150% bigger than upper wick


____________________________________
The inheritance hierarchy (child/parent) is:
____________________________________

1. reverse logic
2. trade on swing: [on/off]
2a1. only trade [5] candles after previous higher high [on/off]
2a2. only trade [5] candles after previous lower high [on/off]
2a3. *only trade [5] candles after previous lower low [on/off]
2a4. *only trade [5] candles after previous higher low [on/off]
2b. [5] candles (start trading from 5 candles after swing)
2c. duration: for [8] following candles (so 'duration' (child) belongs to 'trade swing' (parent)) (stop trading 8 candles after [5] candles after swing)

-------------- general entry conditions ---------------------
3. [on/off]only if previous candle is >200% bigger than previous [4] candles
3a.[4] candles
3b. [200]%

4. [on/off]only if previous candle is >[10] pips
4a. [10] pips

5. [on/off]only if previous candle lower wick >3 pips
5a. [3] pips

6. [on/off]only if previous candle lower wick is >150% bigger than upper wick
6a. [150] percentage
Join our skype group.
Hi, what an interesting concept. Just by looking at wicks, so simple
Could you please confirm that you didn't take into account of the close of the bars proximity to the wicks? ie. a bearish bar with a larger lower wick vs a bullish bar with larger wick. I will play around and see if it makes a difference

I just worry about a large rejection wick from things like moving averages/ significant levels. Which would cause a buy entry according to standard rules.

Thank you for sharing the strategy and to those that contribute to coding the EA. I will try to suggest and add some functions as this looks really interesting.

Kind Regards,
Tzamo
"Work hard, put in the effort, and follow your dreams..." -Me
Some would say - body in direction of profit and wick in direction of loss.
Very good suggestion. This will be added to the EA in the form of: "only if previous candle is >200 times bigger than previous [4] candles"
Join our skype group.
Hi, what an interesting concept. Just by looking at wicks, so simple Could you please confirm that you didn't take into account of the close of the bars proximity to the wicks? ie. a bearish bar with a larger lower wick vs a bullish bar with larger wick. I will play around and see if it makes a difference I just worry about a large rejection wick from things like moving averages/ significant levels. Which would cause a buy entry according to standard rules. Thank you for sharing the strategy and to those that contribute to coding...
Thank you for your great suggestion! I really appreciate it! Thank you, yes the more input the better it is.

Yes, you can test it with the EA (settings: reverse logic on and off)(will be added: settings: only if previous candle upper wick is at least 150% bigger than its lower wick). Is that what you mean?
Join our skype group.
9 pairs 1M TF TP 20 SL 5000 lets see for another hour {image}
Thank you! I appreciate it. Is that 5000 pips or 500 pips? (the EA uses fractals/points so settings: 5000 is 500 pips in the EA)
Join our skype group.
Subs. Thanks and
Regards
{quote} Done {quote} Done {quote} Done {quote} Done {quote} Is this a new parameter? Is it the duration of the trade? or ...? {quote} These three i havnt included yet Im hoping there are some other people who can code in here as well who can add into this, and so im posting the mq4 file {file}
Could you fix this:
Inserted Code
bool swingCheck(Modes value,int candles)
{
if (value == HH)
{
if (High[candles+1] > High[candles+2])
{
return(true);
}
 
}
if (value == LL)
{
if (Low[candles+1] < Low[candles+2])
{
return(true);
}
 
}
 
if (value == HL)
{
if (Low[candles+1] > Low[candles+2])
{
return(true);
}
 
}
 
if (value == LH)
{
if (High[candles+1] < High[candles+2])
{
return(true);
}
 
}
return(false);
}

As I understand, above code is comparing the (5+1) 6th candle (from the right side (index bar)) with the 7th candle. Am I correct? It is comparing two candles, and not a 'swing'. Correct?
I tried to improve on it, but I failed (see below). Can you use the standard ZigZag indicator to get the LL, HL, HH and LH from?

My failed attempt:
Inserted Code
EA should not compare two bars, but instead compare 3 swings (for example: comparing LowestLow and HighestHigh).
double LowestLow = Low[iLowest(NULL, 0, MODE_LOW, 10, 1)]; //from the first bar to the 10th bar
 
double LowestLow2 = Low[iLowest(NULL, 0, MODE_LOW, 10, 10)]; //from the 10th to the 19th bar
 
double HighestHigh = High[iHighest( NULL, 0, MODE_HIGH, 10, 1 )];
//and now check if LowestLow occurs after HighestHigh
int checkSwingType()
{
if(LowestLow3 > LowestLow2 > LowestLow1){ return 1; //1 means: LL (lower low)
if(LowestLow3 > LowestLow2 < LowestLow1){ return 2; //2 means: HL (higher low)
if(HighestHigh3 < HighestHigh2 < HighestHigh1){ return 3; //3 means: HH (a new higher high occured)
if(HighestHigh3 < HighestHigh2 > HighestHigh1){ return 4; //4 means: LH (a new lower high occured)
}; // but the zigzag indicator still does a much better job to identify LL, HL, HH and LH (see last screenshot of zigzag).

{quote} Done {quote} Done {quote} Done {quote} Done {quote} Is this a new parameter? Is it the duration of the trade? or ...? {quote} These three i havnt included yet Im hoping there are some other people who can code in here as well who can add into this, and so im posting the mq4 file {file}
Attached Image (click to enlarge)
Click to Enlarge

Name: HH HL LH LL.png
Size: 151 KB
Attached Image
Join our skype group.
{quote} Could you fix this: bool swingCheck(Modes value,int candles) { if (value == HH) { if (High[candles+1] > High[candles+2]) { return(true); } } if (value == LL) { if (Low[candles+1] < Low[candles+2]) { return(true); } } if (value == HL) { if (Low[candles+1] > Low[candles+2]) { return(true); } } if (value == LH) { if (High[candles+1] < High[candles+2]) { return(true); } } return(false); } As I understand, above code is comparing the (5+1) 6th candle (from the right side (index bar)) with the 7th candle. Am I correct? It is comparing two candles,...
Since u didnt mention about zigzag i assumed HH and LL based on Candles.
Ill see what i can do using the Zigzag
@fauzaanu on Telegram
{quote} In Daily chart the results are: see image. (last 4yrs, TP/SL 500, every tick, EURUSD, 1% DD) {image}
Dear michaellobry,
Lets not get too carried away with trying to optimize your EA. If we do, we will just end up 'curve fitting' into the historical data. We can approach this more statistically as you have started by observing the wicks

What we need to find out is if the wicks and candle bodies are relative in the different timeframes of a particular symbol (eg EUR/USD) For example (Random numbers) in the H1, the candle bodies range from 100 to 200 units and the wicks are 40 to 80 units. While in the 15 min, the candle bodies are 25 to 50 units and the wicks are 10 to 20 units. If that is true, then the candle bodies are relative to the wicks and to run the EA on the other time frame just need to adjust the bodies/ wick limits.

Now to approach the testing: My idea is to select several different situations withing the symbol, like: 10 events of steep up trends, 10 events of steep down trends, 20 events of ranging, events with lots of spikes... and so on and then run the EA on each of those and see if it passes. Anyways, just an idea, if it helps, great and if not I will not be offended

ZigZag is an amazing indicator, will wait for latest wicksys before adding any suggestions

Kind Regards,
Tzamo
&quot;Work hard, put in the effort, and follow your dreams...&quot; -Me
Lets not get too carried away with trying to optimize your EA. If we do, we will just end up 'curve fitting' into the historical data. We can approach this more statistically as you have started by observing the wicks

I agree.

What we need to find out is if the wicks and candle bodies are relative in the different timeframes of a particular symbol (eg EUR/USD) For example (Random numbers) in the H1, the candle bodies range from 100 to 200 units and the wicks are 40 to 80 units. While in the 15 min, the candle bodies are 25 to 50 units and the wicks are 10 to 20 units. If that is true, then the candle bodies are relative to the wicks and to run the EA on the other time frame just need to adjust the bodies/ wick limits.

I'm pondering on the different ways to find out if there is a relation between the size of wick-candle ratio in different timeframes:
- the EA could calculate the average candle size and wicks for each timeframe, and check if the 'ratio of candle and wick size' is the same in other timeframes in the same currency pair.
- the EA could categorize the results into: timezones (for example, London Open has a different candle/wick average than Asia Open)


Now to approach the testing: My idea is to select several different situations withing the symbol, like: 10 events of steep up trends, 10 events of steep down trends, 20 events of ranging, events with lots of spikes... and so on and then run the EA on each of those and see if it passes. Anyways, just an idea, if it helps, great and if not I will not be offended

"10 events of steep down trends". I take it you mean checking 10 'upswings/downswings' on the chart with a sharp slope. It's a very good suggestion, I agree.
Join our skype group.