Art Phantom EA | Page 38 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Phantom 6 System 1,540 replies

Phantom 6 - Manual System 8 replies

Phantom Entry on ODL Japan. 2 replies

Tradeview CAD/JPY "Phantom" spike 0 replies

Phantom high, phantom low 0 replies

Attachments: Art Phantom EA
Exit Attachments

Art Phantom EA

  • #741
  • Nov 29, 2010 7:58pm Nov 29, 2010 7:58pm
Quoting timmy
I can get it to run with pretty good results using high mmtu (say 5) and low pp (say 0.1). You can run the SL at 100 pips.

Using 50:1 leverage.

Results (assuming the Single Pair EA_roundrock is working correctly) were, 25k profit, 22% max DD 22% a $50k account. (For 1 April 2010 to end of November)

The time filter was critical.


Tim
Dont you think 100 pips stop loss is less? Assuming that the basket places many trades? At present the EA ( art-phantom-full- EU ) has placed already four trades, and do not know hoe many more will be added.
Intraday only.
  • #742
  • Nov 29, 2010 8:10pm Nov 29, 2010 8:10pm
Just need to fix the Single Pair PhantomWithSL, roundrock EA so that it will continue to trade after the (open sequence) trade window.

It will be very interesting to see what results we will get then, with the single pair system.

Tim
  • #743
  • Nov 29, 2010 8:12pm Nov 29, 2010 8:12pm
Quoting timmy
As we discussed previously - any addition to the system that disrupts the functioning of the EA - such as cutting losers - will likely have a negative impact on the EA (almost guaranteed over the long term)

We need a serious coder like Jag to refine what we are doing - and not adhoc changes. (Or you Roundrock - but need focus to get the whole thing done without little spin offs here and there).

Come on guys lets get constructive and hopefully we will get some consistant programming but it will kill you when you go live as it is.

Tim
ummmm... not sure if you realize what a jab some of those comments are. I really don't appreciate the negative innuendos spread throughout your message here.

I'd say RoundRock is a pretty serious coder, wouldn't you?

Personally I wouldn't care if there were 50 serious coders here (in fact that would be really cool) as long as they are kindred sprits and want to go along with the intent of the thread and team members Caillou has put in place who follow his lead in general regarding this round of EA's. I mean it is his thread and if someone dosen't like it then they should go find a thread they do like or start one of his own instead of trying to hijack or through mud around this and complicate the messages with crap like "we need a serious coder".

Re: Your summary list:
1. agreed

2. Not so sure about this because it's a little too soon.
I know from experience that when you add filters to any code, even though you may alleviate some of the problems for which you added the filter(s), there's always an opposite cost, a price that insidiously destroys the original robust ingenuity of the code. We want and need the purest code we can have that works and you don't get that by piling on a bunch of filters on without lots of testing first, forward and backward. Rome wasn't built in a day so let's cool our jets a bit on the filters. First we need to have a good mechanically working ea that is what we want on the basic level. No one is going live with these EA's yet and if someone does they'll will be warned way before they get on line. If they still do it then that's their choice, they know the risks.

3. more filters

4. more filters

Funny, I already thought we were being very constructive - everyone has been getting along perfectly, we been together what? 3 days? - then suddenly you pop in and say, we're not serious, we're adding adhock changes and we're not being constructive (why you would try to sneak that one in is beyond me )

We already have a Coach, so if you want to play on the team with us, please understand to be apart of any team including this one is to become one with the members on that team - not another coach to lord over it or the team becomes destroyed.

It is true that We do need good ideas and people that genuinely want to help including all serious coders (jag too if he'd like to team up) but it's only common sense that we need the members to be apart of the whole group and not the second and third leader of it.

Okay back to work
TJ
Charts never lie! (everyone else does) ... now where's that FF Adult lounge
  • #744
  • Nov 29, 2010 8:17pm Nov 29, 2010 8:17pm
Quoting navk
Dont you think 100 pips stop loss is less? Assuming that the basket places many trades? At present the EA ( art-phantom-full- EU ) has placed already four trades, and do not know hoe many more will be added.
Is less than what?... not sure of your question. Are you suggesting that the SL should be bigger? If so I'd agree on the basis that when testing on the single pair EA it is most efficient without this sort of stop loss. You can see for yourself if you test it and make it bigger and bigger SL you'll get better results each time.

I don't like the individual trade SL , but at least is is something to test while someone gets to the other filters and fixes the single pair EA.

We would be better served with an overall max % DD SL, and other filters to improve the R:R and expectancy (and so reduce the chances of this overall max SL being hit).

Tim
  • #745
  • Nov 29, 2010 8:20pm Nov 29, 2010 8:20pm
Untill now i can say that the original EA s posted in first page works fantastic. The modifications/improvements suggested by many dont seem to be ok, but let us wait.

I only hope that the modifications will improve the performance and not spoil the fantastic EA created by Colliou.
Intraday only.
  • #746
  • Nov 29, 2010 8:33pm Nov 29, 2010 8:33pm
The only filters I would add to the original EA are as follows by priority:

1) PipStep
Most useful to prevent over-trading

2) Trade with the Trend
Again, most useful to prevent trading against the trend (if I can get it just right)

3) Trading Hours
A conservative filter allowing the EA to avoid opening a new trade sequence during hours considered outside normal trading hours.

I personally don't use the Trading Hours filter as the PipStep is enough (if big enough) to filter out the chop during the Asian Session, plus if there is no chop, the EA will still be able to trade normally.


The following filters, I personally will NOT use as any pair can reverse at any time and become profitable:

1) Cut Loser Multiplier

2) Cut Biggest Loser when Margin is less than Alarming Margin Level

3) Stop Loss


Just my $0.02 on filters.
  • #747
  • Nov 29, 2010 8:39pm Nov 29, 2010 8:39pm
Quoting TraderJoe
ummmm... not sure if you realize what a jab some of those comments are. I really don't appreciate the negative innuendos spread throughout your message here.

I'd say RoundRock is a pretty serious coder, wouldn't you?

Personally I wouldn't care if there were 50 serious coders here (in fact that would be really cool) as long as they are kindred sprits and want to go along with the intent of the thread and team members Caillou has put in place who follow his lead in general regarding this round of EA's. I mean it is his thread and if someone dosen't...
My hurried typing and poor sentence structure. Not meant to have a go at Roundrock at all - just happened that the thought of focused developement came into my head after mentioning him as the other coder that could do the work... sorry for that if you were offended. (And to you roundrock).

I know these guys can code I am not trying to offend them at all. My point still holds, though change the words if you like.... could be changed for 'serious development work still required'. Was not meant as an attack on anyone.

Reply to rest;

I am only suggesting filters that will not inhibit functionality - so your comments are in line with my thinking.

I accept that we need to fix a couple of issues with base EA - but a variation needs to be settled on. All of these filters etc can be added with flags so they can be tested individually and combined after this.

i will leave off on development if you want to go slow. Sorry if you think I am over imposing. Not intentional at all. Just trying to be constructive.

Moving on.

Tim
  • #748
  • Nov 29, 2010 9:04pm Nov 29, 2010 9:04pm
for the-game,

Concerning trading with the trend. I trade a phantom 6 concept manually on my Oanda account. This isnt the place to get into what I do but it follows the basic concepts of these EA's.

One filter I use is the 24 EMA on the hourly chart. After trading for a month I can tell you it works well to only take a trade that is on the correct side of that MA. Im not a coder so I couldnt figure how to add that to an EA so thats why I trade it manual. Now it works good for a 6 pair basket, but 3 pair or 1 pair it may limit the recovery trades if prices reverse I dont know. Maybe something to try. I had made a post about a month ago on this very subject in the phantom 6 thread but no one seemed to think it was a good idea.

But I know on the EA's Im testing now (mostly 6 pair baskets) the majority of the losing trades were taken on the wrong side of the trend. You guys testing, eye ball it yourself you will see what I mean.
  • #749
  • Edited 9:57pm Nov 29, 2010 9:20pm | Edited 9:57pm
I'm trying the ADX indicator (yeah, yeah, I know I said I don't like indicators), but I'm using it a little different.

I'm just taking into account the +DI and the -DI lines on the H4 time frame.

Example:
EJ is a buy trade, so if +DI > -DI then allow the EA to trade EJ
EU is a sell trade, so if -DI > +DI then allow the EA to trade EU
UJ is a sell trade, so if -DI > +DI then allow the EA to trade UJ

So far it is working as I have expected, but time will have to tell.

EDIT: thorn, sorry, but I absolutely loathe MAs with a passion
  • #750
  • Nov 29, 2010 9:25pm Nov 29, 2010 9:25pm
I made a new filter based on PipStep called PipStepStart.

Basically it won't allow a trade on any phantom in profit until the phantom is in X pips profit

Rough example:
EU phantom is in profit 1 pip, but my PipStepStart is 5 pips = NO TRADE
EU phantom is in profit 5 pips, and my PipStepStart is 5 pips = LET'S TRADE!

I made this filter because a pair would sometimes get into 1 pip profit when it was time to trade then reverse as soon as the trade was made.

But I'm thinking if used with my trend filter, it may restrict the EA from it's full potential.
  • #751
  • Nov 29, 2010 9:37pm Nov 29, 2010 9:37pm
first of all thanks to all who are doing back/forward testing of this. there are known issues and bugs in this, i am going to work with Calliou to fix this. only after we fix these issues we will release this here

again appreciate everyones effort here
  • #752
  • Nov 29, 2010 9:38pm Nov 29, 2010 9:38pm
Quoting roundrock
first of all thanks to all who are doing back/forward testing of this. there are known issues and bugs in this, i am going to work with Calliou to fix this. only after we fix these issues we will release this here

again appreciate everyones effort here
Noted.
Intraday only.
  • #753
  • Edited 10:14pm Nov 29, 2010 9:48pm | Edited 10:14pm
Quoting the-game
The only filters I would add to the original EA are as follows by priority:

1) PipStep
Most useful to prevent over-trading

2) Trade with the Trend
Again, most useful to prevent trading against the trend (if I can get it just right)

3) Trading Hours
A conservative filter allowing the EA to avoid opening a new trade sequence during hours considered outside normal trading hours.

[i]I personally don't use the Trading Hours filter as the PipStep is enough (if big enough) to filter out the chop during the Asian Session,...
My friend, I have to disagree this time, rmb the EA you made me few days ago? with the cut looser and time filter, I compare it with the pure Full, the result is so different, and if anyone think the Full is already so good, then you all will think the mod version the-game made is fantastic amazing. Please, the game, the cut looser is a big help, I dont say it because of some theory I made up with, I say it because I tried it. Now please, be kind again, please combine the Art_full_triming_cut_losser with the pip step in it. Then I will do further analysis. I will post everything in detail when i test them long enough, dont want to confuse anyone or make up theory, I only go for practical approach. trial and error.

the-game, can you kindly code it with the full version this time as you did it with the extreme version last time.

Thanks
  • #754
  • Nov 29, 2010 9:56pm Nov 29, 2010 9:56pm
LP, do you have a statement to show me?

I want to analyse it first before committing any more time to your request.

Please include what Cut Loser Multiplier settings you had used.

Thanks.
  • #755
  • Nov 29, 2010 10:14pm Nov 29, 2010 10:14pm
Quoting the-game
LP, do you have a statement to show me?

I want to analyse it first before committing any more time to your request.

Please include what Cut Loser Multiplier settings you had used.

Thanks.
yes sir, i'll do so when I get home. again, thanks for your help again.

sincerely,

LP
  • #756
  • Nov 29, 2010 10:21pm Nov 29, 2010 10:21pm
LP, uh, don't worry about it, after reading what you wanted, it was an easy mod.

See attached. Go for GOLD!

TH = Trading Hours
CL = Cut Losers
PS = PipStep
Attached File(s)
File Type: mq4 Art_Phantom_Full_TH_CL_PS.mq4   56 KB | 363 downloads
  • #757
  • Nov 29, 2010 10:28pm Nov 29, 2010 10:28pm
Quoting the-game
LP, uh, don't worry about it, after reading what you wanted, it was an easy mod.

See attached. Go for GOLD!

TH = Trading Hours
CL = Cut Losers
PS = PipStep
great, gona try that when i get home, btw, the setting is default, then multiplier is 3, 4 works too, reason i chose 3 because 2 was too tight, doesn't perform well. dont know why, but i'll go for what the result tells me to. again, trial and error, now i have 4 platform running, 2 months later, I will choose the best two, and compete with this one u just posted, and after another 2 months, i'll be able to pick the best, and then go live with 3000$ USD. for 3 months, better if i could test it for one full year, experience both the quiet and volatile or in other words, ranging and volatile market conditions.

best regards.

LP
  • #758
  • Nov 29, 2010 10:39pm Nov 29, 2010 10:39pm
Hi LP,

I'm testing this Trade with Trend filter if successful (looking good so far), you probably won't need to use the Cut Losers Multiplier filter any more as it won't be trading against a pair's trend.
  • #759
  • Nov 29, 2010 10:43pm Nov 29, 2010 10:43pm
I'm also trying this filter where I half the PP during Asian Session (AS).

Reasoning is because there is usually not much Price Action/Volume during the AS.

BUT there is still money to be made during the AS, especially during those rare times there is actually enough Price Action/Volume.
  • #760
  • Nov 29, 2010 10:49pm Nov 29, 2010 10:49pm
Quoting the-game
The only filters I would add to the original EA are as follows by priority:

1) PipStep
Most useful to prevent over-trading

2) Trade with the Trend
Again, most useful to prevent trading against the trend (if I can get it just right)

3) Trading Hours
A conservative filter allowing the EA to avoid opening a new trade sequence during hours considered outside normal trading hours.

[i]I personally don't use the Trading Hours filter as the PipStep is enough (if big enough) to filter out the chop during the Asian Session,...
the-game,

You're on the money for me. (Although not sure yet about pip step...) I agree with you.

Cutting out losers: .... sure its theory Low Profile - but that does not mean that it can't be fact.

The 'theory':
Why should taking out losing trades from the basket be any more effective on the net result than leaving them in... after all the EA is designed as it is since we are making the statement that we do not know the direction of the market and we will let the market decide which way it will go (weighted trade distribution). So why would you want to take out some of the trades just because they are losers at a given moment in time. Six hours later they may be the winning trades that get you to your pp.

Some trade sequences it will help - others it will casue them to take much longer to complete their equity pop. All it will do in the long run is cause a greater chance of a possible death trade DD on one of the trade sequences it destabilises because the market reverses back into the direction of the loosers you cut.

I can understand if you are desperate for margin - but then if it is that bad we should have killed of the trade, and reset for a better mmtu setting to suit the account size and broker side leverage.

i am sure we can see this theory demonstated on the single pair EA over a decent sample period if you wanted to try it.

Tim
PS Trader Joe please don't think I am coaching here - merely trying to be constructive in the development process.
0 traders viewing now