Mibl, this is why most people lose at Roulette, it's been shown statistically that one should rather bet WITH the trend and double those bets as the trend works in your favour vs the Martingale stratagy. The problem with an EA that has a PT=40 and SL=100 or 300 (in some EA's) is that you have a long line of winners and only a single trade to set you back to the beginning or even worse, losses. Is there an EA that agressively buys on a confirmed trend with a 2:1 or 3:1 risk reward stratagy.? I'm having good success with with the ASC trend indicator and using a williams %R as a stop, it's not without it's losss of course but at least i'm working with the trend.

If you switch over to 1.5 progression your first argument on SecureProfit is no longer correct.
Take a quick view on the following (regarding a mini account, round up to a valid lotsize <=> a little more that 1.5 and spreads are ignored in this model):

trade / lotsize => overall pips on closing on a retracement of 17 pips
#1 / 0.1 => -43 * 1 = -43
#2 / 0.2 => -28 * 2 = -56
#3 / 0.3 => -13 * 3 = -39
#4 / 0.5 => +2 * 5 = +10
#5 / 0.8 => +17 * 8 = +136 ==> total +8 pips

trade / lotsize => overall pips on closing on a retracement of 18 pips
#1 / 0.1 => -42 * 1 = -42
#2 / 0.2 => -27 * 2 = -54
#3 / 0.3 => -12 * 3 = -36
#4 / 0.5 => +3 * 5 = +15
#5 / 0.8 => +18 * 8 = +144 ==> total +27 pips

=> Using 5 trades drawdown (out of Maxtrades) you need a move of 18 pips to overstep SecureProfit of 10.


trade / lotsize => overall pips on closing on a retracement of 19 pips
#1 / 0.1 => -58 * 1 = -58
#2 / 0.2 => -43 * 2 = -86
#3 / 0.3 => -28 * 3 = -84
#4 / 0.5 => -13 * 5 = -65
#5 / 0.8 => +4 * 8 = +32
#6 / 1.2 => +19 * 12 = +228 ==> total -33 pips

trade / lotsize => overall pips on closing on a retracement of 20 pips
#1 / 0.1 => -55 * 1 = -55
#2 / 0.2 => -40 * 2 = -80
#3 / 0.3 => -25 * 3 = -75
#4 / 0.5 => -10 * 5 = -50
#5 / 0.8 => +5 * 8 = +40
#6 / 1.2 => +20 * 12 = +240 ==> total +20 pips

=> Using 6 trades drawdown (out of Maxtrades) you need a move of 20 pips to overstep SecureProfit of 10.


trade / lotsize => overall pips on closing on a retracement of 21 pips
#1 / 0.1 => -69 * 1 = -69
#2 / 0.2 => -54 * 2 = -108
#3 / 0.3 => -39 * 3 = -117
#4 / 0.5 => -24 * 5 = -120
#5 / 0.8 => -9 * 8 = -72
#6 / 1.2 => +6 * 12 = +72
#7 / 1.8 => +21 * 18 = +378 ==> total -36 pips

trade / lotsize => overall pips on closing on a retracement of 22 pips
#1 / 0.1 => -68 * 1 = -68
#2 / 0.2 => -53 * 2 = -106
#3 / 0.3 => -38 * 3 = -114
#4 / 0.5 => -23 * 5 = -115
#5 / 0.8 => -8 * 8 = -64
#6 / 1.2 => +7 * 12 = +84
#7 / 1.8 => +22 * 18 = +396 ==> total +13 pips

=> Using 6 trades drawdown (out of Maxtrades) you need a move of 22 pips to overstep SecureProfit of 10.


So if you want to have the effect needing smaller retracements, when the trade goes against you, you must enhance the profitfactor between 1.5 and 2.

mibl

__________________
Never trade with money which you need to make a living *** I trade with my money based on my decisions *** Please decide for yourself.

My testing to date has been mainly done on a doubling progression and my first comments reflected that with the attached gif previously posted as an example.

I have not tested the effects of the 1.5 progression but agree that the retracement needed to close profitably will be increased and that would affect my argument comparing this EA with Roulette.

Thank you for your analysis I have not done the work to see exactly how much more retracement will be needed I am prepared to forward test the 1.5 progression and after time decide on it's effectiveness.

I have however been forward testing a 1.64 progression and have included a current screenshot indicating some increased retracement needed and no doubt more will be needed with the 1.5 progression.

Thank you for your interest and contribution to this thread

John

Attached Images

	Direction Change.GIF (55.9 KB, 146 views)
	Direction change1.64.GIF (41.2 KB, 134 views)

John,

I would like to apologize for my neglect with my revised EA from 10point3. Things have just been absolutely crazy for me at work lately but I'm trying to stay on top of it. I was going to look for the latest version of the Goblin EA so that we could take advantage of the indicator(s) they use to determine direction. Everything is set on my revised except for picking a direction to begin with. I have been running the test EA with long positions only to see if everything works as it should and it all looks good. I am hoping everyone will find it useful but at least for my sake I feel that the code is better organized, from a programmers standpoint, and hopefully everyone will benefit from this. Your devotion to the 10point3 thread has been most inspiring and I feel that you have done a great job in keeping it alive and useful for everyone. My testing took a huge hit about a week ago using only long positions but has since rebounded like you wouldn't believe. I know that's not the best way to go but it just shows, in my opinion, that the theory behind this EA is a good thing even when your in the wrong direction as long as you have margin to cover it. Could anyone tell me which page the latest Goblin EA can be found?

I have also included My test results since running this revised version and would like everyone to keep in mind that it is not 100% complete and not suitable to run on a live account as it stands. All that is needed is the indicator(s) to point the direction. Also, the results may be weird in the beginning due to some manual manipulation while I was tweaking. This is just to show that I have actually been working on it.

Attached Images

10Point3 Dynamic Stop v2 - 2006-12-04.gif (5.5 KB, 143 views)

Attached Files

	10Point3 Dynamic Stop v2 - 2006-12-04.htm (157.1 KB, 68 views)
	10Points3_Dynamic_Stop_v2.ex4 (7.8 KB, 49 views)
	10Points3_Dynamic_Stop_v2.mq4 (11.3 KB, 108 views)

Hopefully my comments in the new EA's code can be easily understood but if I did not explain something well enough then any questions regarding the code are most welcome. Also, for right now anyways, I have dropped the idea of closing trades that may be called hoepless due to some complications I have seen. The EA starts with a 1.6 progression because 1.5 is not quite big enough to maintain a profit if multiple orders are opened when looking for the TP and my testing has been run on the EUR/USD, GBP/USD, USD/CHF, USD/JPY with an H1 TF. The EA has been opened up so that it can be run on any currency pair without any modifications and it will work as it should.

Hi m8,

I didn't realize we were pursuing dual objectives with improving 10Point3 until I saw your posts here to that effect. Perhaps we should do a "Vulcan Mind Meld" on your efforts and my development of Goblin....if you've made some progress with your order handling logic, I'd love to take a peek at it.

Not a problem. My apologies. The latest version of my revised EA is in the post above. You will find a function in there where I would like to set OrderDirection to 1 if long, 2 if short. Anything else will be ignored and nothing will be done. The function is called "CheckForDirection()". That is where the indicator(s) would be placed and the variable "OrderDirection" would be set appropriately. I had planed on giving credit to the goblin EA as well in the header section.

Thanks again,
Matt

Oh, hey...no apologies necessary. I think it's cool that someone else was looking at this too. I think we all see a glimmer of possibilities in 10Point3. I look forward to taking a look at your fresh approach. Cheers!

Could you direct me to the latest download version of the Goblin EA and any support/indicator files I may need so that I can have a look?

Matt,

bluto has updated the first post on page 1. I believe that is the latest version.

File Type: zip Goblin.zip (17.5 KB, 411 views)
Last edited by bluto : 11-28-2006 at 06:32 AM.

Tross discusses his Fibo Goblin on page 39 post # 390

Very pleased to hear from you and hope that a dialogue with bluto will bring out the best for us all.

John