Flyer! EA - New Thread! | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

My trade explorer Flyer > 1000% return for May 3 replies

Flyer! EA - New Thread - Part 2 147 replies

Fozzy, Pip Boxer, Flyer 4 replies

Help Test my new EA, FLYER! 243 replies

Flyer! Version 2.00 is released. 0 replies

Attachments: Flyer! EA - New Thread!
Exit Attachments

Flyer! EA - New Thread!

  • #41
  • Oct 16, 2006 5:04pm Oct 16, 2006 5:04pm
Quoting Grahame
Hi Marci,

When backtesting what time frame do you use? Noooooooo! Just JOKING!

Just want to say having the entry and exit points joined by a dotted line is a nice touch. Haven't seen that anywhere else. It is so quick and easy to review the action.

Thank you so much for everything!

Kind Regards,

Grahame.

Oh, I almost forgot, my DEMO account (Real Time) is increasing! (He He)
Glad you're enjoying it!

Regards,
Marci
  • #42
  • Oct 16, 2006 5:47pm Oct 16, 2006 5:47pm
Hi everyone,

I'm attaching a file to this post that contains all the relevant optimum settings for all the currency pairs I have tested. It also shows the results achieved on back tests for those settings.

I hope this helps everyone out.

GirlFlyer

P.S. I hope tomorrow I can get the same out for Version B.
Attached File(s)
File Type: doc FlyerV105Asettings.doc   45 KB | 838 downloads
  • #43
  • Oct 16, 2006 6:52pm Oct 16, 2006 6:52pm
Quoting GirlFlyer
Chad,

I don't see any problems with your settings. I ran GBP/USD for 08/01/06 thru present with the same values and showed a 54,000 profit.

Typically when I see somebody unable to duplicate the results it is usually something to do with the data set or the strategy tester settings. It has to be run on the M1 setting. No other setting will work properly.

If you are unable to resolve it, please send me via PM a copy of the strategy tester report. I would be interested to see what is going on.

GirlFlyer
Marci,

this is the strategy tester report from 2 server Interbank and Strategybuilderfx.

showing loss in recent 1-2 months data.

i think i'm just too new to your EA, am i missing something in the first place?

and also,

i hope u can answer me this time, if i'm gonna run it on a mini account, what are the configurations ?

thank you,

Chad
Attached File(s)
File Type: zip StrategyTester.zip   20 KB | 447 downloads
  • #44
  • Oct 16, 2006 7:33pm Oct 16, 2006 7:33pm
ok say no more, I initiate forward testing immediately.

i've just looked in the code and find the configuration for mini account. sorry if my harsh question bothering u.

Chad

Life goes on
  • #45
  • Oct 16, 2006 7:47pm Oct 16, 2006 7:47pm
Hi, Marci

Can you explain what the point adjustment does?

Thanks, Dan
  • #46
  • Edited 11:49pm Oct 16, 2006 11:38pm | Edited 11:49pm
Quoting Gids
Hi, Marci

Can you explain what the point adjustment does?

Thanks, Dan
I am going to try to make up for my question earlier about historical data. Here is what i think you are looking for:


A positive or negative value that adjusts the decision process to enter a trade. Must be in increments of the currency pair being traded. i.e., EUR/USD = 0.0001, USD/JPY = 0.01, etc. Note: This factor is very sensitive and should not be adjusted very far from the suggested values.




That is from the Doc.

You are telling the EA what place value is equal to 1 pip. In the EUR/USD- 1 pip looks like 0.0001. In the USD/JPY- 1 pip looks like 0.01.

Im guessing if you increase it by 1 pip(0.0002), it creates sort of like a buffer zone? Maybe price must go up 1 more pip or down one more pip to give u the signal? Also does it being positive or negitive make this difforence;if it is negitive, a signal is taken 1 pip earlier? Taking a stab at the adjustment there.

Lookin out for you Macri! Keepin the blood pressure down

<3
  • #47
  • Oct 16, 2006 11:57pm Oct 16, 2006 11:57pm
Hi Marci and all.

Let me say this about my enthusiasm for this MAcross EA. As it is based on a manually tradable model, and one that is quite logical, i find it is a good fit because i would trade this way manually. All the more reason to dig into this one and see it's success for all.

As for my post above, and your kind replies; I have the point values set to those suggested in the original .doc files for both EAs (.0002/.02Jpy vB, .0001/.01Jpy). So if the point value creates a situation where there is more "play" i'll maybe go to zero?? I don't think i want it taking trades unless the lines really cross over for sure. I've just downloaded the new settings .doc and will make the according changes to vA.

There is an ea over at FXTSD that is using what NewDigital calls a "Delta", basicaly what it does is there is a user set filter off the primary MA of XX number of +/-pips. I don't know if this might help to eleminate some of these situations (i came home at there were two trades on vA that had similar close, but no actual cross entries). Anyway, I'm going to paste in the line of code here from the other MA, i'm not trying to "fix" anything that might not be broken, i'm just here to see Flyer work into what you've envisioned. My apologies if this post of possible additional code idea is inappropriate, just trying to help.

extern int Delta1=0;

-----the above lets the user set how many pips +/- from the main MA that acts as a filter for entering trades. The MAs must actually exit this channel (think of this as the "Levels" setting when placing an MA on a chart. Same idea)

if (
(Buy1_1 < Buy1_3+(Delta1*Point) && Buy2_1 >= Buy2_3+(Delta1*Point))

------So the above is looking at the Main Ma + the "Delta" filter before the trade will be taken.

double Buy1_1 = iMA(NULL, 0, SignalEMA_Period, 0, MODE_EMA, PRICE_CLOSE, Current + 1);
double Buy1_2 = iMA(NULL, 0, CloseEMA_Period, 0, MODE_EMA, PRICE_CLOSE, Current + 1);
double Buy1_3 = iMA(NULL, 0, MainEMA_Period, 0, MODE_EMA, PRICE_CLOSE, Current + 1);

------The above is just the details of what the different MAs are doing.


Good trading to all,
Thom
  • #48
  • Oct 17, 2006 2:06am Oct 17, 2006 2:06am
Quoting chad
Marci,

this is the strategy tester report from 2 server Interbank and Strategybuilderfx.

showing loss in recent 1-2 months data.

i think i'm just too new to your EA, am i missing something in the first place?

and also,

i hope u can answer me this time, if i'm gonna run it on a mini account, what are the configurations ?

thank you,

Chad
Hi Chad,

I figured out your problem....

I should have realized, but I'm not used to working with two different versions yet. The reports you show me are from version B. All the settings I have published so far, are for version A. Version B does not work well on the 15 minute time period, only on the 30 minute time period. I will be posting some settings for version B today. In the meantime, if you look in my old thread "Help test my new EA, Flyer!", you can find version 1.04 in there. The settings it has in the documentation will still be fairly accurate for use in Version B.

Also, you still have some gaps in your data history. If you have 100% of your data in your charts, you will get a modelling quality of 25%. 25% is the maximum value possible. If you're missing data, it will go below 25%. In looking at the trades, there seems to be a gap around 8/11 to 8/23. Check that out. You must make sure you can manually scroll on your screen through the entire history to be tested for your data set to be accurate.

Hope all this helps

Marci
  • #49
  • Oct 17, 2006 2:36am Oct 17, 2006 2:36am
Quoting accrete
Hi Marci and all.

Let me say this about my enthusiasm for this MAcross EA. As it is based on a manually tradable model, and one that is quite logical, i find it is a good fit because i would trade this way manually. All the more reason to dig into this one and see it's success for all.

As for my post above, and your kind replies; I have the point values set to those suggested in the original .doc files for both EAs (.0002/.02Jpy vB, .0001/.01Jpy). So if the point value creates a situation where there is more "play" i'll maybe go to zero?? I don't think i want it taking trades unless the lines really cross over for sure. I've just downloaded the new settings .doc and will make the according changes to vA.

There is an ea over at FXTSD that is using what NewDigital calls a "Delta", basicaly what it does is there is a user set filter off the primary MA of XX number of +/-pips. I don't know if this might help to eleminate some of these situations (i came home at there were two trades on vA that had similar close, but no actual cross entries). Anyway, I'm going to paste in the line of code here from the other MA, i'm not trying to "fix" anything that might not be broken, i'm just here to see Flyer work into what you've envisioned. My apologies if this post of possible additional code idea is inappropriate, just trying to help.


Good trading to all,
Thom
Hi Again,

I think I'm going to answer your post and Liquid11's all in one post. First, I don't mind people making changes to the code for their own usage. That's what its here for. Second, as to the Point Adjustment. The point adjustment feature allows the EA to jump in a trade early or late depending on how close everything is to crossing. (There is one other setting that has some affect on the entry price which I will discuss further down). I have also found this factor is very sensitive for the profit factors. Some currency pairs have better performance with significantly different values. Btw, I've tested these pairs on various values ranging from either -0.0003 to 0.0003 or -0.03 to 0.03 and the ones I have given are statistically what seem to give the best profit results over time. However, this value (the Point Adjustment) must be balanced with the Stop Loss Value and the Trailing Stop Value. For instance, I found a more profitable setting for the Eur/Usd using different values with the point adjustment set to .0001 instead of -0.0001. However the Stop Loss value was over 75. On a 15 minute timeframe, I felt a stop loss over 75 was a bit excessive so I went with the lower Stop Loss.

Additionally, the Trailing/Stop value also helps determine the entry price. The EA takes 25% of the Trailing/Stop value and compares that amount with the difference between the actual price of the pair vs. the value the MA's at cross. If the difference exceeds 25% of the T/S, it doesn't trade because the the price has moved too far too fast. (Prevents news announcements from causing trades!!!!!) I will probably add a new variable in future versions to separate this function so as to have more control and less interdependency.

Finally, Accrete... a word about the situation that occured last night for you. This situation seems to be very rare. This is the first time I've seen it. For this to happen, essentially the price movement on a pair has to be almost zero for an extended period of time for all 4 MA's to be within 2 or 3 pips of each other all at the same time. Yes, changing the point adjustment to 0 will eliminate this problem. However, you statistically over time change your profit margins. Run a few strategy testers with various values in the point adjustment and see the difference. It's quite amazing how much difference 1 pip on your entry price can do.

Your Delta filter seems interesting, but I'm curious to wonder how well it will work over time and how wide the tunnel is that it creates. If the tunnel is too wide, more than 4 pips top to bottom with the centerline at the point of cross over, I doubt you will get good results if it forces the trade to occur only once it is outside of that tunnel. Anyway, feel free to test it, and let me know if you get any success.

Liquid11,

I hope this answers most of your question on the Point Adjustment. Essentially what you said in your post is correct. Negative values allow the signal to become valid earlier while Positive Values force it to become valid later. So, yes! you got it and thanks for examining it in detail first before just jumping in with the questions. This is the kind of situation I feel most inclined to answer!

Happy Trading to All!

GirlFlyer
  • #50
  • Oct 17, 2006 2:38am Oct 17, 2006 2:38am
Quoting Gids
Hi, Marci

Can you explain what the point adjustment does?

Thanks, Dan
Gids, See Post #49
  • #51
  • Oct 17, 2006 2:55am Oct 17, 2006 2:55am
Hi Marci,

It seems to me that what you are doing is harder than playing a dozen games of CHESS at the same time.

Good on You!

Kind Regards,

Grahame.
  • #52
  • Oct 17, 2006 3:16am Oct 17, 2006 3:16am
Quoting Grahame
Hi Marci,

It seems to me that what you are doing is harder than playing a dozen games of CHESS at the same time.

Good on You!

Kind Regards,

Grahame.
More work anyway!!!!

Marci
  • #53
  • Oct 17, 2006 4:43am Oct 17, 2006 4:43am
GirlFlyer

I am currently testing Version A with a 50,000 demo account using all the settings from the document. Some pairs have large profits, but I do not see the TS adjusting itself to protect the gains thus far. For the EAs, MinVal was left at 5000. If the SL is ever hit for all the profitable trades, all would be at a loss.... I will let the EA run to see how it does without my intervention.
  • #54
  • Oct 17, 2006 4:50am Oct 17, 2006 4:50am
Quoting GirlFlyer
Hi Chad,

I figured out your problem....

I should have realized, but I'm not used to working with two different versions yet. The reports you show me are from version B. All the settings I have published so far, are for version A. Version B does not work well on the 15 minute time period, only on the 30 minute time period. I will be posting some settings for version B today. In the meantime, if you look in my old thread "Help test my new EA, Flyer!", you can find version 1.04 in there.........

haha ...

i just start running it for 10 hours now, doing great, making 200 hundreds over pips, even the settings are wrong. will correct and restart again.

Millions Thanks.

Chad
  • #55
  • Oct 17, 2006 4:52am Oct 17, 2006 4:52am
Quoting et_phonehome_2
GirlFlyer

I am currently testing Version A with a 50,000 demo account using all the settings from the document. Some pairs have large profits, but I do not see the TS adjusting itself to protect the gains thus far. For the EAs, MinVal was left at 5000. If the SL is ever hit for all the profitable trades, all would be at a loss.... I will let the EA run to see how it does without my intervention.
ET, The T/S doesn't kick in until the profit exceeds in pips the value you set for the Trailing Stop. Hope that helps you understand.
  • #56
  • Oct 17, 2006 6:36am Oct 17, 2006 6:36am
Quoting GirlFlyer
ET, The T/S doesn't kick in until the profit exceeds in pips the value you set for the Trailing Stop. Hope that helps you understand.
Thanks for explaining it. If the profit is never reached prior to kicking in, don't you feel that it is risky in losing all those gains? I am not sure if you will find many trading like this live since one want to at least break even - I know there pros and cons to it.

Keep up the good work....

ET
  • #57
  • Oct 17, 2006 6:47am Oct 17, 2006 6:47am
Quoting et_phonehome_2
Thanks for explaining it. If the profit is never reached prior to kicking in, don't you feel that it is risky in losing all those gains? I am not sure if you will find many trading like this live since one want to at least break even - I know there pros and cons to it.

Keep up the good work....

ET
ET,

That's why I started version B. It will take profits at the top/bottom of the curve. Drawback is if the trend is not done and takes off again, it will not catch the whole thing. Your choice.

GF
  • #58
  • Edited 8:29am Oct 17, 2006 6:54am | Edited 8:29am
Hi Everyone!

******** Correction Everyone ***** Everything inside these asterisks is a correction to the earlier message. Had an error in this post earlier... I had an incorrect setting on the usd/jpy and the eur/jpy (%equity was wrong) and falsely gave higher than normal profit readings.

I've corrected the settings file to reflect the true 8% equity I have been testing with all along.

Sorry, Eur/Jpy is not truly more profitable on Version B as I originally thought.

The file in this message is changed from earlier.

*****************************************


As promised, here are the settings and test results for Version B. Some interesting findings with Version B. *** Part of original message deleted here ****

PLEASE, PLEASE remember that the settings for Version A do not match those with Version B. Version A works better on the 15 minute period whereas generally, Version B works better on the 30 minute period. I say generally because the AUD/USD is an exception. It works on Version B better with the 15 minute period.

Anyway, see the the attached file for the settings, results, and comments.

Good luck to everyone and I'm glad to see everybody seems to be overall getting some good results.

Marci (GirlFlyer)
Attached File(s)
File Type: doc FlyerV100Bsettings.doc   51 KB | 707 downloads
  • #59
  • Oct 17, 2006 8:58am Oct 17, 2006 8:58am
Quoting GirlFlyer
I should have realized, but I'm not used to working with two different versions yet.
May I sugest breaking this thread into two ... one for each version to avoid confusion.

May turn out to be easier to track the two variants of this EA?
Just my 2 pips worth
  • #60
  • Oct 17, 2006 9:12am Oct 17, 2006 9:12am
just wondering which broker would run this EA? I have tried on neuimex and interbankfx platform but seems not functioning...

please advise.
{Link Removed}]
0 traders viewing now