Hi Acko!!
Wow, that was really one great reception in the thread of the banned. Well, from my point of view Effi was clearly inside the limit and the ban is a overreaction. But I really just realized that I knew you and wanted to help.
Gnat
)
)
Was thinking about this during the night (while I was sleeping - I think in my sleep hehe) See chart below. This would allow for example to enter on the first blue dot, then maybe hedge a bit (or not) at the short signal but NOT be taken out of the first trade if x-y-z arent't met, thus enabling to continue the run up and then again either hedge or not at the next retracement (provided it's a retracement) and over and over and over until we've squeezed 99.999999% out the potential heheh?
http://i21.tinypic.com/20a80ti.jpg
- #285
- Oct 2, 2007 1:41pm Oct 2, 2007 1:41pm
- Joined Jun 2007 | Status: Unconventionable Conventionist | 2,768 Posts
That's what I'm talkin about
Now you 2 are beating me big time, when it comes to greediness! But good point there. That would be one possibility.
I have even stranger things in mind... but am unable to draw a coherent picture right now. (It is too loud here!) 'll come to that in a bit!
Ciao
Gnat
- #286
- Oct 2, 2007 1:53pm Oct 2, 2007 1:53pm
- Joined Jun 2007 | Status: Unconventionable Conventionist | 2,768 Posts
What "thread of the banned" WHERE WHERE show me (I feel like I'm 8 years old
Effi Lang, the starter of the thread http://www.forexfactory.com/showthread.php?t=46048
was banned for 2 weeks from FF, because he was promoting his own personal tradinggroup (commercial) and did not obey to the moderators instruction to take the links off and to stop "mentioning" about this other way to make money.
It was on the one hand "half information feeding", as he always lacked the complete information but promised it in the commercial group, but on the other hand he posted pairs, buying limits and targets freely on the thread that mostly worked out well, rendering up to 1000 pips per week to the folks there.
Acko was asking about an entry on the daily chart for a long on GBPCHF and I confirmed according to what Supernova would have to say (as Acko was here quite often).
Well, I should have stuffed that down my everhelping mouth, as I was instantly barked at and bitten, because I did not mention the 25EMA and the T3 was not part of the concept... blablabla.
But Effi will soon be back and the tension will be easing... I hope it for Tim4Kaye. Otherwise he will see the Dark Side of My Moon. hehehe
So far so good!
Gnat
- #287
- Edited 3:42pm Oct 2, 2007 3:14pm | Edited 3:42pm
OK so I'm done with the stop loss and trailing stops tests.
SLs
http://i24.tinypic.com/2e67kfk.jpg
Really trying to squeeze every molecular pip here. The SLs gain 28 pips for eurjpy, 3 for audjpy, 5 for gbpjpy and 0 for nzdjpy, so that’s a total +36 pips.
The tests for usdjpy and cadjpy were already done since they're the same TF as my "longer tf" strategy. See that post for details.
I got better sampling quality today for nzdjpy (no clue why) however it was 85% not 90%.
Trailings
http://i20.tinypic.com/16ge9tv.jpg
These don't change anything in the life of nzdjpy nor audjpy but gross us an extra 49 pips total for the others (I did test much further down for audjpy (not shown), but it still peaked at 189).
I also checked for "interference" between SL and trailings, but everything was ok, the pips were added "for both" without any cancelling out or anything (never know).
And here's the final "optimal" table (for now):
http://i21.tinypic.com/30c3mon.jpg
So, you can either trust me or put the exact same settings as the table below and run the tests from jan 1 to aug 31 2007. Should come out exactly the same. That's except maybe for cadjpy for which lately I've been unable to get good sampling quality, but those numbers were tested when I did have a good % so they should be reliable.
Now, if forwarding testing can be as good as that..........................
:
EDIT: For those of you too (hmmmm lazy? LOL) to search back, here's th Optimal setup for longer TF's (post #193)
http://i22.tinypic.com/zohnc6.jpg
SLs
http://i24.tinypic.com/2e67kfk.jpg
Really trying to squeeze every molecular pip here. The SLs gain 28 pips for eurjpy, 3 for audjpy, 5 for gbpjpy and 0 for nzdjpy, so that’s a total +36 pips.
The tests for usdjpy and cadjpy were already done since they're the same TF as my "longer tf" strategy. See that post for details.
I got better sampling quality today for nzdjpy (no clue why) however it was 85% not 90%.
Trailings
http://i20.tinypic.com/16ge9tv.jpg
These don't change anything in the life of nzdjpy nor audjpy but gross us an extra 49 pips total for the others (I did test much further down for audjpy (not shown), but it still peaked at 189).
I also checked for "interference" between SL and trailings, but everything was ok, the pips were added "for both" without any cancelling out or anything (never know).
And here's the final "optimal" table (for now):
http://i21.tinypic.com/30c3mon.jpg
So, you can either trust me or put the exact same settings as the table below and run the tests from jan 1 to aug 31 2007. Should come out exactly the same. That's except maybe for cadjpy for which lately I've been unable to get good sampling quality, but those numbers were tested when I did have a good % so they should be reliable.
Now, if forwarding testing can be as good as that..........................
:EDIT: For those of you too (hmmmm lazy? LOL) to search back, here's th Optimal setup for longer TF's (post #193)
http://i22.tinypic.com/zohnc6.jpg
Yeah... there is something for you!
Effi Lang, the starter of the thread http://www.forexfactory.com/showthread.php?t=46048
Gnat
.Something else, all I have left testing wise (for now at least lol) is the TP. Now, I've never really been confortable with TPs, not even sure what figures to test. And not sure there's any worth using them with this system.
But, as a "scientific tester" I'm going to go through with it anyway. We never know what can pop out of a TP testing hat.
$
Great Job!!!!
OK so I'm done with the stop loss and trailing stops tests.
SLs
http://i24.tinypic.com/2e67kfk.jpg
Really trying to squeeze every molecular pip here. The SLs gain 28 pips for eurjpy, 3 for audjpy, 5 for gbpjpy and 0 for nzdjpy, so that’s a total +36 pips.
The tests for usdjpy and cadjpy were already done since they're the same TF as my "longer tf" strategy. See that post for details.
I got better sampling quality today for nzdjpy (no clue why) however it was 85% not 90%.
Trailings
http://i20.tinypic.com/16ge9tv.jpg
These don't change anything in the life of nzdjpy nor audjpy but gross us an extra 49 pips total for the others (I did test much further down for audjpy (not shown), but it still peaked at 189).
I also checked for "interference" between SL and trailings, but everything was ok, the pips were added "for both" without any cancelling out or anything (never know).
And here's the final "optimal" table (for now):
http://i21.tinypic.com/30c3mon.jpg
So, you can either trust me or put the exact same settings as the table below and run the tests from jan 1 to aug 31 2007. Should come out exactly the same. That's except maybe for cadjpy for which lately I've been unable to get good sampling quality, but those numbers were tested when I did have a good % so they should be reliable.
Now, if forwarding testing can be as good as that..........................
- #294
- Edited 4:32pm Oct 2, 2007 4:21pm | Edited 4:32pm
- Joined Jun 2007 | Status: Unconventionable Conventionist | 2,768 Posts
Thanks. I just hopped over there and will be skimming through a bit. But LOL I thought the custom user title "Posting Infractions" was actually cool, didn't notice exactly who had it (I guess many people) I really thought someone had made that up, you know meaning "I'm a tough outlaw when it comes to forex and when it comes to posting, I'm BOLD) or something like that
Something else, all I have left testing wise (for now at least lol) is the TP. Now, I've never really been confortable with TPs, not even sure what figures to test. And not sure there's any worth using them with this system.
But, as a "scientific tester" I'm going to go through with it anyway. We never know what can pop out of a TP testing hat.
I guess you are living in a timebubble and are able just to snip your fingern and then the time freezes... you do your test ... then you post that you start testing...LOL ... and finally you freeze our time again and make the tableau!
So in milliseconds you have run multiple tests on a bin full of pairs and created fancy spreadsheets!
In the same time I have only 2 telephonecons that led nowhere!
Oh yes, oh so productive!Great things happen here!
Gnat
- #295
- Oct 2, 2007 4:31pm Oct 2, 2007 4:31pm
- Joined Jun 2007 | Status: Unconventionable Conventionist | 2,768 Posts
I'm having a look at where my 3 bot systems are currently and pondering on my sandwich bots (bread&butt -> well if you didn't know, ham is pork butt hehe) and thinking I've got a problem with this 40 pips/day thing. See this morning when I flared up the SandwichBots, they took old signals (and also I played around with the parameters as I was getting results from my tests, so today doesn't really count).
Anyway, this "closing off" after 40 pips, then re-starting the next morning isn't working, may well work if some bugs are fixed (maybe they're unfixable).
But here's my point. I just realized that with a potential of 1,396 pips/mth on the shorter time frames (which is what these bots are setup for) when I calculate (I should nominate myself for the Nobel in math for this one):
1396/4 = 349 pips/week
349/5 =69.8 pips/day which is just slightly over my objective, so just why not leave the ham&cheeze bots roll away, since on average I'll get the 40 I'm after, sometimes a little bit more (yay) sometimes a little bit less (snif) but the potential 40 is there without my intervention (I love not intervening with bots, I didn't like zapping them nor reviving them).
So, from now on I'll call these the autosandwichbots (looks like a german word heheheh
)
Anyway, this "closing off" after 40 pips, then re-starting the next morning isn't working, may well work if some bugs are fixed (maybe they're unfixable).
But here's my point. I just realized that with a potential of 1,396 pips/mth on the shorter time frames (which is what these bots are setup for) when I calculate (I should nominate myself for the Nobel in math for this one):
1396/4 = 349 pips/week
349/5 =69.8 pips/day which is just slightly over my objective, so just why not leave the ham&cheeze bots roll away, since on average I'll get the 40 I'm after, sometimes a little bit more (yay) sometimes a little bit less (snif) but the potential 40 is there without my intervention (I love not intervening with bots, I didn't like zapping them nor reviving them).
So, from now on I'll call these the autosandwichbots (looks like a german word heheheh
)
- #297
- Edited 2:54pm Oct 3, 2007 10:51am | Edited 2:54pm
- Joined Jun 2007 | Status: Unconventionable Conventionist | 2,768 Posts
