MADTrader

[MathiasJ]

I don't want to "mess up" this thread with too many alternative versions of source code.

I am a strong believer in version control. This MADDY/MADTrader combination, with all their settings, already provide an enormous amount of configuration and flexibility, that can be a tester's nightmare. If we are going to throw into this mix a plethora of different versions of the source code, we are completely going to lose ourselves and it will be very hard to attribute a certain test result to a certain setting/option/version.

I would encourage you to keep testing your version, and if you can demonstrate with tangible performance reports that your version has merit over the core version, I will include your changes into the core.

Hi Renée!
It sounds like we are on the same page.
I am myself an Enterprise/IT Architect and fully understand your standpoint.
I will use my versions as a playground to evaluate things.
I will explain my settings that I have added in another post just to give you an idea what I try to explore.
BR
Mathias

[MathiasJ]

Hi Renée!
It sounds like we are on the same page.
I am myself an Enterprise/IT Architect and fully understand your standpoint.
I will use my versions as a playground to evaluate things.
I will explain my settings that I have added in another post just to give you an idea what I try to explore.
BR
Mathias

Here are the settings I currently explore:
UseMADDYSignals=true/false (Odd one, but used when exploring the CSS part to temporarily exclude MADDY signals)

entryRule (here I have added CSS as a new entryRule in the enumeration)

CSSTimeFrame=15
CSSAlignWithD1=true (flag if also testing currencies towards CSS D1 so they are aligned)
CSSMinLevel=0.2 (the normal border for ranging)
CSSMaxLevel=0.8 (upper border for level limit for when to test "Angle")
CSSMinAngle=0.05 (It is actually not a real angle, it is delta of CSS for curr on shift 0 compared to shift CSSShift)
CSSShift=2 (Mentioned above)


Then I have added a separate list of filters:
AddCSSFilter=false (I think this is like in MADDY but here only tested when there is a signal from MADDY)
flt_CSSDiffLimit=0.3
flt_CSSTimeFrame=240

filterEntryRule=4 (new enumeration with extra filtering possibilities if needed)
FILTER_MA_STACKING, // Use proper MA Stacking for filter
FILTER_MA_CROSS, // Use MA Cross for filter
FILTER_HGI, // Use HGI Trend Arrows for filter
FILTER_STOCH, // Use Stoch OB/OS for filter
FILTER_NONE, // Do not use filter
FILTER_MA_SINGLE, // Use single MA for filter
FILTER_MA_SINGLE_CROSS // Use single MA Cross for filter

I actually added possibility to select both HGISignalTrend and HGISlopeTrend.

Then regarding exit criterias I added the following to be able to have another and maybe more tight setting for close of trades than for open of trades.:
exit_MATimeFrame=5
exit_SlowMAPeriod=35
exit_SlowMAMethod=3
exit_FastMAPeriod=20
exit_FastMAMethod=3
exit_StochTimeFrame=15
exit_StochKPeriod=5
exit_StochDPeriod=3
exit_StochSlowing=3
exit_StochMethod=3
exit_StochOSLevel=30.0
exit_StochOBLevel=70.0
exit_HGITimeFrame=240
exit_HGIActOnSignalTrend=true
exit_HGIActOnSignalRange=true
exit_HGIActOnSignalRAD=true
exit_HGIActOnSlopeTrend=true
exit_HGIActOnSlopeRange=true
exit_HGILookBackCandles=10

And a new flag for closing trades at a specific time:
doCloseTradesAtHour=false
CloseHour=17

I have yet to add a News filter, because that can be needed sometimes.

I can imagine that some of those settings may not be needed but that is what I try to find out.
In the end it can even be so that my CSS ideas may evolve into another EA, like a new CSS Trader
But that is yet to find out.
Imagine having CSS D1 levels giving the trend direction.
Then follow up with eg CSS on m15 to try the level and Angle to find movement.
Add to that MADDY signals confirming the CSS15.
After that I am not sure we need more signals like MA Cross/Stack etc to filter or we can just go with what we have.

That was my thinking so far
I may have misunderstood some concepts as I am quite new on CSS, but please direct me then as I do what I can to contribute here.
BR
Mathias

[windorz]

Thank you Leon!!

What a amazing approach

[windorz]

Ok Guys due to some requests I received via Pm I am posting some set files here to be attached to Mad Trader v 1.1 and only because my birthday is coming up

Here we go, and please share ideas:

Tc

Hi Leon

Thank you for sharing your knowledge!

I see your files and No5 is only FastMAMethod=3.
Is it intended ?

 5- Mad Trader 15 min Ma stack.set
- SlowMAPeriod=15
- SlowMAMethod=3
- FastMAPeriod=8
- FastMAMethod=3
^^^

other files
-SlowMAPeriod=15
-SlowMAMethod=3
-FastMAPeriod=8
-FastMAMethod=1
^^^

No1(M5 stack) close condition is different from No5(M15 stack)
Is it intended ?

 1- Mad Trader 5 min Ma stack.set
doExhaustTrades=false
doCloseOnMACross=false <<<
doCloseOnOppositeHGI=true

5- Mad Trader 15 min Ma stack.set
doExhaustTrades=false
doCloseOnMACross=true <<<<<
doCloseOnOppositeHGI=true

[Boulder]

Hey guys, apologies for hopping on the party train here so late
Rene, George, Leon and all the others involved here; you guys are doing a fabulous job.. Thanks for all the efforts! Brilliant piece of sophistication right here..

So I thought to jump in and help out a bit too. I'll be using Leon's open and close configurations (also brilliant btw :hi: ) for Maddy as well. Things I tweaked are:

• LookBackMinutes= 120
• Stricter StrengthSignal conditions
• Exhaustion signals included

The idea of exhaustion really struck me as utterly impressive. To try to bring a bit of consistency I've been thinking of maintaining a net (strength - exhaustion) condition of at least 4. So for M5 I have at least 6 strength and exhaustion at most 2. M15 at least 5 strength and at most 1 exhaustion. M30 and H1 same as M15. I'll be back with some results soon! With Brexit on the agenda, this should be an interesting week

Cheers and take care,

Aram

[Wavegarrick]

Hi Windorz,

I see your files and No5 is only FastMAMethod=3.
Is it intended ?

Thanks for pointing this out. Set file 5 must be set to exponential ( Fast Ma method )

No1(M5 stack) close condition is different from No5(M15 stack)
Is it intended ?

Yes this is correct. I was getting too many small closures and I find that having just Hgi set to true sufficient for the 5 min time frame.

Hi Boulder, thanks for coming in. Looking forward to your test results.

Cheers
Leon

[windorz]

Hi Rene

Simple question
Why MadTrader does not support STOC OB/OS exit ?

[renexxxx]

Why MadTrader does not support STOC OB/OS exit ?

There is no particular reason for. I had to stop somewhere with adding exit rules. There are an unlimited number of entry rules and exit rules. You have the source code, so you can add your own. But having more options is not always better. If you can demonstrate that the OB/OS exit rule is crucial for this EA, I will add it in a flash.

[windorz]

There is no particular reason for. I had to stop somewhere with adding exit rules. There are an unlimited number of entry rules and exit rules. You have the source code, so you can add your own. But having more options is not always better. If you can demonstrate that the OB/OS exit rule is crucial for this EA, I will add it in a flash.

Yes, You are right.

I download source codes on post 1 and place each file to right place(I think).

But I have Errors when compiling MadTrader1.1.mq4.

How can I fix this error ?
Anything wrong ?

[Wavegarrick]

Hi Guys,

As always I am inspired by Nanningbob's Css...... Here is a set I am trying right now and as mentioned before I am starting to look at including Css with Maddy based on the open and close of a candle on diff time frames, so far steady results today... Trades do not come too often but a steady supply of trades.....good enough for a days trading..

Please feel free to try and feedback welcome.

Tc
Leon