Jiva Dragon | Page 36 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Enter The Dragon - GBP/JPY 2 replies

Info on the Borihei-Dragon Scalping System? 15 replies

Jiva's 5,5 Trade Journal 5 replies

The Dragon 441 replies

***Dragon II*** 15 replies

Attachments: Jiva Dragon
Exit Attachments

Jiva Dragon

  • #701
  • Nov 1, 2010 12:23pm Nov 1, 2010 12:23pm
I just wanted to thank you guys for this wonderful EA. I have been trading version 11 with real money on FXDD, but to be safe just .10 cent pips, but it has won every trade over the last few weeks, except 1 that I closed in drawdown because it didn't look good, but of course it came back had I left it alone, and now with your new "safer" version maybe I can increase the size a bit. I would rather trade very small live than use a demo account and I beieve in your EA. I have bought many EA's that just don't work at all. I have been trading all pairs except pairs that have the CAD in them since on those I remember reading that further back tests than 1 year were not so good. So I just wanted to say a big thank you to Jiva and Taiyakixz and all the others who have contributed.
  • #702
  • Nov 1, 2010 1:06pm Nov 1, 2010 1:06pm
Quoting birdogent
I just wanted to thank you guys for this wonderful EA....
Nice!!
Happy to hear you have some success.
Yes, if a trader can afford it, micro gives a more realistic experience. after all demo there is nothing really at stake.
I think the entries are good. Now we are refining with several new tools. The 50/50 statistical edge test is superb to tell right away if the entries are any good or not, and then we can optimize those entries. Last thing to optimize is take profit.
With smaller stop loss we will be able tolerate the inevitable bad trade relatively painlessly. But we are not done!! I think we may never be really done.
We are not giving the green light go ahead for trading real money just yet. But if you feel confident to do so, more power to you as they used to say...

We have been asked if we would go commercial with this. But my feeling is this ea is a product of the free sharing of ideas and experience in the forum. It should remain free. It's not like there aren't enough pips for everyone in the forex market.
  • #703
  • Nov 1, 2010 3:28pm Nov 1, 2010 3:28pm
Please always refer to the title page for the latest updated files.

v14.1 EURUSD for Alpari
50/50 edge test win rate 72.73%
Stop loss 28 pips
Attached Image (click to enlarge)
Click to Enlarge Name: eurusdv14.1alpari11-1.gif Size: 9 KB
Attached File(s)
File Type: zip eurusdv14.1alpari11-1.zip   3 KB | 261 downloads
File Type: mq4 JIVA Dragon EA DSS T3MA MAs testv14.1.mq4   40 KB | 286 downloads
  • #704
  • Nov 1, 2010 4:46pm Nov 1, 2010 4:46pm
i got optimizer to work properly i think.
i tested eurchf 1-1-10 to 1-1-10
net profit 22,000
relative drawdown 7 %
only 18 trades.
What about overoptimizing ( curve fitting ) where it works good on back test but not on forward.
The optimization takes a long time even just changing one thing using every tick, can you use control point or open price to get a rough idea.
Trade 10 stayed open for 3 months with a 9,300 profit, not bad, a long time to stay opened but it kept modifying to keep the profit, so looks good.
All trades were sell i noticed.
I had a few 2 week gaps in history but 90% modeling.
Not sure on the test report about mismatched error i get 0, so that is good. But after that there is the green bar and on different tests sometimes there is more green, sometimes less, not sure if that is important.
Keep up the good work.
  • #705
  • Nov 1, 2010 5:21pm Nov 1, 2010 5:21pm
Quoting kauaibobby
i got optimizer to work properly i think.
i tested eurchf 1-1-10 to 1-1-10
net profit 22,000
relative drawdown 7 %
only 18 trades.
What about overoptimizing ( curve fitting ) where it works good on back test but not on forward.
The optimization takes a long time even just changing one thing using every tick, can you use control point or open price to get a rough idea.
Trade 10 stayed open for 3 months with a 9,300 profit, not bad, a long time to stay opened but it kept modifying to keep the profit, so looks good.
All trades were sell i noticed.
I...
The more we practice using the optimizer the more we will see how to use it effectively.
Good to see you making the effort.
Time consuming, cpu consuming, you name it, optimizing is a long slog for sure.
Sometimes I set the step to 5 or more and use control point and that speeds things up. Then go back and set step to 1 near where the optimal settings were found and use every tick to finalize it.
It's a matter of practice to get to know how to use it. Practice, practice, practice...
http://t0.gstatic.com/images?q=tbn:A...MGi06CewFiztU=
  • #706
  • Nov 1, 2010 5:28pm Nov 1, 2010 5:28pm
Quoting kauaibobby
i got optimizer to work properly i think.
i tested eurchf 1-1-10 to 1-1-10
net profit 22,000
relative drawdown 7 %
only 18 trades.
What about overoptimizing ( curve fitting ) where it works good on back test but not on forward.
The optimization takes a long time even just changing one thing using every tick, can you use control point or open price to get a rough idea.
Trade 10 stayed open for 3 months with a 9,300 profit, not bad, a long time to stay opened but it kept modifying to keep the profit, so looks good.
All trades were sell i noticed.
I...
>>>What about overoptimizing ( curve fitting ) where it works good on back test but not on forward.<<<

We are running a 50/50 statistical edge test the very first thing to determine if the entries are any good at all.

The 1st profit, 2nd profit and stop are all set to the same amount.
No trailing stops. The other settings are left as they are from the downloaded set file for the pair.

Run the optimizer. If the win rate is more than 50%, preferably 65% or better we know we have a statistical edge. We will win more times than lose. That we are not simply fitting a curve. Jm2110 in his critical analysis last week brought this point out. We are seeing how to use this 50/50 to best advantage.

Then when we optimize for take profit etc, we have reassurance that the basis from where we are performing optimization is sound. So if the profit is set larger than the stop loss, and the win rate is more than 50%, we know we will come out ahead even if there a few losses along the way.

So we try to do the 50/50 test with as small a stop loss as possible. So we can optimize the take profit higher starting from a small stop loss. Make sense?
:nerd:
  • #707
  • Nov 1, 2010 5:36pm Nov 1, 2010 5:36pm
Here is what i was reading about EAs and optomizing, very informative, at least for me anyway http://www.easyexpertforex.com/backt...imization.html
  • #708
  • Nov 1, 2010 5:47pm Nov 1, 2010 5:47pm
Quoting kauaibobby
Here is what i was reading about EAs and optomizing, very informative, at least for me anyway http://www.easyexpertforex.com/backt...imization.html
Thanks for posting the link..
  • #709
  • Nov 1, 2010 5:52pm Nov 1, 2010 5:52pm
Jiva thank you
You explained what the 50/50 is, I wasn’t sure what you meant by that.
It is starting to make sense all the hard work you are doing, I am slowly figuring it out.
I know brokers vary a lot, but I was wondering what we should experiment with our broker.
I know the offset is different for brokers. What about changing other setting what should we try first.
How much difference between FXDD and Alpari do you find you need to change settings.
I hope I am not asking to much of you to help me figure this stuff out, but I want to help and make money also with your great ea.
  • #710
  • Nov 1, 2010 6:13pm Nov 1, 2010 6:13pm
Quoting kauaibobby
Jiva thank you
You explained what the 50/50 is, I wasn’t sure what you meant by that.
It is starting to make sense all the hard work you are doing, I am slowly figuring it out.
I know brokers vary a lot, but I was wondering what we should experiment with our broker.
I know the offset is different for brokers. What about changing other setting what should we try first.
How much difference between FXDD and Alpari do you find you need to change settings.
I hope I am not asking to much of you to help me figure this stuff out, but I want to help...
No problem. Communicating about trying to get this right gets the creative juices flowing.
Have to be away from the trading desk for a little while. Sorry 'bout that. Will get to back to you later.
have settings for EURCAD v14.1
  • #711
  • Edited 9:10pm Nov 1, 2010 7:13pm | Edited 9:10pm
I tryed to post in the thread some hours ago, and somehow it was not aprooved again . Do you know at what numbers of posts my posts are not going to be manually checked by moderator?


Hello Guys
I have no trades placed by EA ,since I first activated V14, (like 9 hours ago)

Is this normal guys? it still makes few trades ? I will update when new trades are placed


Thanks guys

__________-
Sequoia
  • #712
  • Nov 1, 2010 10:16pm Nov 1, 2010 10:16pm
Quoting Sequoia
I tryed to post in the thread some hours ago, and somehow it was not aprooved again . Do you know at what numbers of posts my posts are not going to be manually checked by moderator?


Hello Guys
I have no trades placed by EA ,since I first activated V14, (like 9 hours ago)

Is this normal guys? it still makes few trades ? I will update when new trades are placed


Thanks guys

__________-
Sequoia
Sheesh... Why so many problems for you to post? shouldn't be. I don't think it is moderator. Maybe some technical problem with computer...
You can contact the administrators, to see if the problem is on ff end, they're okay people. They do a darn good job of keeping forex factory in line... and focused with very little ads cluttering up things.
Bottom of any web page, click contact ff..
It's hard enough trying to get trades right without all the other technical stuff making it even more difficult. Let's get this straightened out.
  • #713
  • Edited Nov 2, 2010 9:58am Nov 1, 2010 10:27pm | Edited Nov 2, 2010 9:58am
Quoting Sequoia
I tryed to post in the thread some hours ago, and somehow it was not aprooved again . Do you know at what numbers of posts my posts are not going to be manually checked by moderator?


Hello Guys
I have no trades placed by EA ,since I first activated V14, (like 9 hours ago)

Is this normal guys? it still makes few trades ? I will update when new trades are placed


Thanks guys

__________-
Sequoia
9 hours and no trades is not unusual at all. Especially while the market has no clearly defined direction which is often the case for days before significant data is going to be released.
Sometimes just before the data there are big moves in anticipation and then after the release a strong trend forms in many of the pairs. That is why we want to be set up as much as possible before Tuesdays US election results and Weds fed announcements.
I have noticed that v11 has taken so many trades last few days while v14.1 has not taken any at all. It may be because the settings for v14.1 are safer with smaller stop losses they will not be opened until there is a really strong move in the direction of the trend. So Tues and Weds will be perfect test time for v14.1. It will be interesting to see how v11 and v14.1 both perform.
v11 just opened short EURUSD during Asian session.


Update several hours later:
here's what happened.
this is using v11 and shows why we want to develop v14.1 with the small stop losses.
The Dragon sold a bottom............on the 1h
the Dragon was also set up with a 15min setting which opened long a few minutes later.
The long position acted as a hedge to lessen the losses.
Total loss -126 pips
Without the hedge loss would have been the full -220. It is an idea that has been planted. A setting with a large stop loss could be accompanied by a setting on lower time frame. Like this, the 1hr accompanied by a hedge setting on the 15 min.
Interesting.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: eurusd.gif Size: 76 KB
Click to Enlarge Name: eurusd3.gif Size: 66 KB
  • #714
  • Edited 10:42pm Nov 1, 2010 10:40pm | Edited 10:42pm
Quoting jiva34
Sheesh... Why so many problems for you to post? shouldn't be. I don't think it is moderator. Maybe some technical problem with computer...
You can contact the administrators, to see if the problem is on ff end, they're okay people. They do a darn good job of keeping forex factory in line... and focused with very little ads cluttering up things.
Bottom of any web page, click contact ff..
It's hard enough trying to get trades right without all the other technical stuff making it even more difficult. Let's get this straightened out.
Yes , and this Forum is by far the best in all the web about Forex.
Maybe I just have to wait until I have the right number of posts to be automatically authorized, I really don't want to boder anyone, they should be very bussy , and just to track the problem is going to take too much time for them. (I know is not my system's problem, because I have seen my post online ,with the little red box in the bottom part saying my post first should be manually authorized...)

A quick qestion, are you still thinking about moving the system to lower TFs, or do you think this can be best in H1 as it is now?

Thanks Jiva for your time and kindness, green pips for you and everyone in the thread.

EDIT : AWSOME !!! this post was authomatically uploaded, no more problems

______________-
Sequoia
  • #715
  • Nov 1, 2010 10:49pm Nov 1, 2010 10:49pm
Quoting Sequoia
Yes , and this Forum is by far the best in all the web about Forex.
Maybe I just have to wait until I have the right number of posts to be automatically authorized, I really don't want to boder anyone, they should be very bussy , and just to track the problem is going to take too much time for them. (I know is not my system's problem, because I have seen my post online ,with the little red box in the bottom part saying my post first should be manually authorized...)

A quick qestion, are you still thinking about moving the system to lower...
The Higher Powers have seen fit!!

One thing at a time. Let's get the 1hr series up and running first before going on to anything else. The Dragon is extremely versatile and I have no doubt there are other ways to set the options but for now we are concentrating, let's stay focused. The grass is very green on this side of the fence...
the 1 hour has a lot going for it.
  • #716
  • Nov 1, 2010 11:20pm Nov 1, 2010 11:20pm
Please always check the title page for the latest updates.

Settings for EURCAD v14.1 FXDD

stop loss 50 pips
50/50 edge test win rate 85.71%
Attached Image (click to enlarge)
Click to Enlarge Name: eurcadv14.111-1.gif Size: 8 KB
Attached File(s)
File Type: zip eurcadv14.111-1.zip   3 KB | 235 downloads
  • #717
  • Nov 1, 2010 11:20pm Nov 1, 2010 11:20pm
Quoting jiva34
9 hours and no trades is not unusual at all. Especially while the market has no clearly defined direction which is often the case for days before significant data is going to be released.
Sometimes just before the data there are big moves in anticipation and then after the release a strong trend forms in many of the pairs. That is why we want to be set up as much as possible before Tuesdays US election results and Weds fed announcements.
I have noticed that v11 has taken so many trades last few days while v14.1 has not taken any at all. It may...
Very true , with safer conditions less trades are taken, thanks Jiva
  • #718
  • Nov 1, 2010 11:36pm Nov 1, 2010 11:36pm
Okay here's what happened. I have 3 different settings for EURUSD set up on the demo. The original v11 1h setting with the 110 stop loss entered short.
Then the v11 15 min setting entered long. so now the account is hedged.
This, simply by accident, not planned, has given an interesting idea how to possibly manage a trade with a large stop loss. It will be interesting to see how this plays out.
Attached Image (click to enlarge)
Click to Enlarge Name: eurusd2.gif Size: 69 KB
  • #719
  • Nov 2, 2010 12:51am Nov 2, 2010 12:51am
Hi Jiva,

I agree v14 looks very promising BUT unfortunately when I run the same set file from Jan09 you can once again easily visually spot the period where you started your optimization from. This is using your .set file with a corrected GMT offset for Dukascopy 99% accurate tick data. As you can see at about trade 74 the P&L starts its upward march but before that it isn't great at all so I still am very concerned that with so many variables to optimize for we can effectively curve fit your model to any data set and make it look fantastic.

As you know from my past comments I am a little dubious on rolling optimizations as I see them as just rolling curve fitting but I am happy to agree to disagree on that. I still fancy trying to optimize this EA over much longer data sets and seeing what the "average" performance is like.

Given the speed at which you are pumping out these .set files you have obviously managed to work out a series of variables you optimize in steps rather than all at once as otherwise the time taken to do each optimization would take weeks. Therefore could you please explain what process you are going through i.e. which variables you target first and with what steps and then 2nd 3rd etc. This would help me greatly when doing much longer runs over larger data sets after which I can return with my findings.

Thanks in advance.

Jono
Attached Image (click to enlarge)
Click to Enlarge Name: eu.jpg Size: 54 KB
  • #720
  • Nov 2, 2010 1:01am Nov 2, 2010 1:01am
Quoting jm2110
[color=black][font=Verdana]Hi...
Glad to see you back. Thought maybe you had abandoned ship. There seems to be less and less hours in a day lately. Will get back to you asap with the sequence I am currently using and our ongoing discussion of optimizing.
0 traders viewing now