Funny Looking Rectangles & Useless Drivel

You are welcome and thanks.

To answer your question: no, not really. Just take it all in stride and remember that money is there to serve you, not the other way around.

iDouble,

You have really caught my attention to delta trading and the use of vectors. I have been experimenting the last week with them and I my seventh sense is telling me something is there. However as you well know, there is an endless amount of possibilities!!

Can you please give me some clues as to what to do? I have been looking at the current open compared to previous bars 3,5,8 and taking the vector projection from it. This is too erratic so I am now trying a projection that is a moving average of the 3,5,8 vector over the last 5 bars. Just experimenting.

How many bars back do you take data from to build a vector? (I wont ask you from what points you take the delta's)
How do you filter the projections? (I have thought about using the angle of incline - too high or too low is considered a false signal, and also the number in the sequence the projection was made, or clusters)
The higher the timeframe the better the accuracy, what is the minimum timeframe you will use vectors?

I am long long way off catching up to your system, can you save me an unknown number of years research and throw a dog a bone please?? I'm not afraid to do the work, I just need some guidance.

Thanks.

Inate, as stated previously, I don't discuss the details of how I go about building my edge. However, I've given several concepts and direct statements in this thread that give a small amount of guidance for developing your own. This post here: http://www.forexfactory.com/showthre...49#post7759649 and this post here: http://www.forexfactory.com/showthre...01#post7762701, can help to provide more guidance.

There are Deltas, Trajectories and Vectors yet to be discovered using the basic approach to calculating a Delta, which is very simple Subtraction. All you are doing is subtracting one number from another to derive the Delta and then running that same calculation through history until you have sufficient data to analyze against market price. It really is just that simple.

Hope that helps.

Thanks for the reply. I have read both of those posts and I certainly get what you are on about, but the possibilities are in theory, endless.
Yes the mathematics is surprisingly simple. ​The talent is in how you apply them.

This thread has given me some additional ideas for potentially predicting the location of price across short distances. I will be taking this opportunity to make this my final exit this year. The interesting conversation kept me here long than anticipated, but I really have overstayed my visit. I'm actually going to take some additional time off from trading and work-up the test environment for these new concepts that this thread has given me.

It has been a pretty interesting visit this time around. Have a good rest of your 2014. Trade with confidence and don't let convention get in the way of progress.

Take care and thanks for sharing... It's been fantastic insight..

Hey, thanks again. I have already set upon a new system using a similar set-up to the Bento Box ("Been To Box"?).
It seems to have merit.

I'll keep working on it until next time and throw you all my bones. (Just don't expect any mathematics...just anectdotes).

Our conversations have made me think about this game in a different way.

And don't forget the pending lesson on why Diamonds Are Forever!
Plenty O'Toole.

"Diamonds are forever" is a reference to the statistically most common form of (individual) candlestick pattern (aka the "white body").

iDouble, thanks for your insight. Until next time...

I thought this was cute and interesting. I took a snap-shot of the equity curve produced while doing some additional research on B2i which was inspired by this thread. This looks like it could lead to B2i Version 6.0, but she will have to survive forward testing before I can certify it as a new version. Anyway, this is a non-standard B2i on Equity Curve - just for the fun of it:

Single Indicator Isolation Test Trade Samples - No System Input to Signal: Phase 1/Forward Set Sample 'A'

http://i61.tinypic.com/wsk4rm.jpg

As I continue testing the new additions to B2i's output in isolation (disconnected from the system), I will post the Equity Curve snap-shot to see how this Fall Line worked out in the future.

NOTE: I'm not "back" (won't be doing any follow-up questions) but the reality is that I am always just a click away from any forum in the world - working on the computer so much.

Ciao!

Added Detailed Report (html):

Forgot:

10/14/14 D1 Standard GBPAUD B2i trade with V6 modifications actually broke even today - 2 days later. That was a significant -$1,443.87 hit on the equity curve that did not have to happen had I held the position. Detail Report transaction number: 482900884. I closed it because I had already seen enough of the other trades and their respective recovery periods after missing the B2i Fall Line within 24 hours of entry. Some of these "losses" are actually Gains or BE at worst. I needed to clear the charts, get my available equity back and move on to do other B2i V6 phase testing. So, it looks like the performance of the enhancements was better than previous expected when I account for a proper hold time after a missed approach to the initial target.

FWIW - you can actually build a tight little "system" (method/tactic) around B2i all by itself with the right augmentations on timing of the entry, of course. I'm using another previously unimplemented input from the TCDi7 Indicator to augment (support) B2i and it changes the B2i signal logic which in many cases changes both trade direction, entry location and target level to increase probability. So, far it seems to perform really well - but I need to move through at least a thousand trades to certify the version upgrade. I've got about 900 more trades to go. I'm running in batches of 15-16 depending on available equity and how many are in recovery from the previous day's sample set.

The testing and certification process marches forward on B2i V6. It is tedious work but validation is essential. Can't go to war with unproven weapons and a bad battle plan. If it can't contribute to growing capital consistently then it can't have a place in the system and V5 remains extant.

My work music for today: Relax

All sounds great! Enjoy it I guess I have more work to do as well.

And... I can't thank you enough right now, maybe forever, since I am now able to move my work forward (to v3) with whatever I have learnt from this thread (and still learning and testing). Simple ideas can change the world.
http://i57.tinypic.com/311kis8.png

Funky looking chart

Hello SaverO,

Just wanted to drop a quick note to say thanks for BoxMaker. Not being an MQL developer, it is a TimeSaverO - pun intended. Adding the visual Fall Line was also a TimeSaverO. It will help me speed up the last stage of preparation just before I enter test positions. Since I'm testing 15-16 pairs each day, that saves quite a bit of drawing time.

I would make one other suggestion, however.

You could add even more value to the indicator by coding it to paint the Previous Missed Standard Fall Line and its associated B2i box. Paint the Previous Missed B2i box along with its Fall Line using a Dashed Line. Allow the user to select the color for the box and the fall line. The Dashed Line allows the Trader to quickly identify the missed fall line and its associated box.

How far back should you allow for Previous Missed Standard Fall Lines?

I would say to let that value be an External Variable set by the Trader at indicator load time. So, the Externs could look like:

http://i59.tinypic.com/20qi554.jpg

Where, PreviousMissed_On/Off = 1 (On) or 0 (Off).
Where, PreviusMissed_LookBackRange = User Defined

PreviousMissed_On/Off is simply a switch to toggle the function on or off. PreviousMissed_LookBackRange allows the Trader to set the number of bars included in the PreviousMissed range. Example, if:

MaxBars = 1000
forTimeFrame = D1
drawEvery_X_Bars = 0
PreviousMissed_On/Off = 1
PreviousMissed_LookBackRange = 7

A Standard D1 B2i will be painted with its Fall Line AND all existing Previously Missed Standard D1 B2i will be painted with their associated Fall Lines using a Dashed Line for the Previous seven (7) D1 bars (or, over the most recent 7 day range). Obviously, logically, PreviousMissed_LookBackRange cannot exceed MaxBars (as you have defined MaxBars).

Of course, the "LookBackRange" is always dependent on the B2i range being painted. So, the value of PreviousMissed_LookBackRange is equal to the size of forTimeFrame. Thus, if forTimeFrame is D1, then the value the Trader inputs for PreviousMissed_LookBackRange will literally "look back" X number of D1 bars, regardless of which chart time frame the indicator is being run. If, forTimeFrame is H1, then the value the Trader inputs for PreviousMissed_LookBackRange will literally "look back" X number of H1 bars, etc. Thus, maintaining the time relationship between forTimeFrame and PreviousMissed_LookBackRange.

As usual, you should be able to put the indicator on any lower time frame chart and paint a higher time frame B2i (which is the way it works right now - so don't change that - that works nicely).

Again, this would quickly show the Trader where the missed levels live and help them evaluate price tendency accordingly. In my recent testing, I had a few untouched fall lines that were several days old. I manually highlighted those to standout among the others. Sure enough, price closed in on those missed levels eventually - though I pulled out of some positions before that happened.

I'm not a software developer, obviously. I'm a trade logic developer. Writing software code escapes me as I spent all my time developing trade logic development skills. Some people can do both, but that's a rare bird indeed when you find that person and when the trade logic is as good as the software code can automate it.

Don't know if these additions to BoxMaker are doable but it would certainly be a good idea to create and share.

Ciao!

I thought this was so funny, I just had to post it:

http://i61.tinypic.com/10wtyxc.jpg

She broke through. LOL! I'm sorry, this is just too much fun. Imagine, you create an indicator to trade with. That indicator then becomes an indicator of its own performance. Hilarious! Absolutely, hilarious. We'll see how far this weird, bizarre and strange experimentation goes. The added closed trades that brought Equity up through the Fall Line from below are attached as .doc.

I would classify this as Bermuda Triangle type stuff. The Indicator on the equity data predicts future equity curve location. That's just nuts - but I like it.

Hello Rand,

I wanted to make sure I thanked you as well. I just loaded your B2i indicator into all my test charts and it seems to work just fine. The sliding End and Start lines are a quick way to cover Hybrids as well - so again, thanks. It took me a while before I got around to testing out both yours and the other called BoxMaker.


Hello SaverO,

After further review, it appears as though your BoxMaker already shows previous missed fall lines automatically as it normally functions. I believe I can tweak the code enough to get the line color and style changed myself. So, thanks again!

Hi iDouble,

He is away at the moment, he should be back next week sometime.

MM

Hey, thanks, MM!

To: Rand and SaverO,

What would be a helpful visual is for Saver's Fall Line to be painted on Rand's B2i structures. Saver, your B2i structures are nice but when painting historical structures across many pairs (I am testing 16 pairs right now), MT4 gets sluggish and virtually unusable. Deleting the indicator causes MT4 to cycle at normal speed again. However, your Fall Line projections and the way you stop projecting the line once struck works well.

Rand, your toggle switch for painting all structures within the selected (variable) range of time is really good. However, the indicator does not paint any Fall Line. It would be cool if your structures projected the Fall Line right up to the point where price strikes the line from outside the structure. It would be a quick visual reference while evaluating other indicators before entering a position.

Another other thing I would add is this also, Rand. In addition to projecting the Fall Line until externally struck, when the toggle for "ShowAll" is turned on (value = 1), all previous structures with Fall Lines that have been struck externally (from outside the structure) fade to Dark Gray in color (probably best done at the next tick). This again provides for a quick visual recognition of the remaining structures with their externally untouched Fall Lines.

Lastly, Rand. The indicator seems to repaint or reset itself when you switch from one MT4 profile to another. It does not retain its last state when moving between profiles. I use four profiles currently and with each profile selected, B2iV2 resets itself using the current Time[] settings in the code. That means that not only does it reset itself, but it shift forward whenever the starting line is Time[0]. Of course, it is always seeking the current bar which is always going to shift forward with time. It would be better to anchor the structure where the start and end lines are located/fixed by the user.

If someone else has the coding skill to tackle it - feel free to give it a shot. Rand's multiple structures with Saver's continuous Fall Line projection until struck from outside the structure, would be a best of both worlds solution.

Back to evals...