Search | User Listing Vidsite | Calendars | Quotes
Home Page ->  Markets & Methods -> Development of New Tools -> View ThreadLogon (or Register, or Join TradeTight)

You are logged in as a limited-access Guest.To join TradeTight, first read the info in the Organization & Content room, then click the link above. 

1 Cor 15:58 (NET) ... Always be outstanding in the work of the Lord, knowing that your labor is not in vain in the Lord.


MZ3 Confluence Enhancement Project
Jump to page : 1
Now viewing page 1 [50 msgs/pg]
Jump to room :
JimDean
Posted 3/2/2015 8:07 PM (#6150)
Subject: MZ3 Confluence Enhancement Project



Owner/Admin

Posts: 3925
2000100050010010010010025
Location:
USA: GA, Lawrenceville
Salim emailed me in response to the (more focused) project about enhancing MoneyZone 2, here:
http://tradetight.org/forums/thread-view.asp?tid=1127

He wants to see a broad-scope tool that examines "confluence" aka support and resistance levels from a variety of different methods, and determines confluence levels from those. He was particularly interested in Mel's "DPL" (defended price levels) ... which certainly are derived using a very different approach ... but other candidates might include Pivot-based levels (a la WaveTrader), Fibonacci levels, consolidation zones, etc.

It seems to me that the proper way to do this is:
1. Get/create indic that finds MZ2 levels with sufficient flexibility
2. Get/create indic that finds DPLevels with sufficient flexibility
3. Get/create indic that finds Pivot levels with sufficient flexibility
4. Get/create indic that finds Fib levels with sufficient flexibility
5. Get/create indic that finds consolidation levels with sufficient flexibility
... and so on for other useful methods of determining S/R levels

THEN
Create an engine that gathers info about levels from those separate tools (could use multiple calls to any or all of them), then determines confluence from them according to user-input weighting and rules.

This could be a "group project" ... but it's got a *lot* broader scope than the MZ2-focused one in the other thread. First, a consensus would be needed as to whether to leverage canned N plugins to the max (MZ2, WT, Fib) ... or alternatively create enhanced versions of them.

Then there would be a need to invent a good method for sorting through the scores of lines created by those various sources, weighting them together based on "aging" somehow, culling them using fuzzy-bands like MZ3 does, and so on.

It's the kind of thing I usually would work on independently then release, since it's so massive and complex. That kind of thing usually takes a back seat to paid contractual devel work. So, in response to Salim's request, I'm starting this thread mainly to find out what interest folks might have in this ... and also to open it up for general discussion.

Respond as you'd like, but please keep the discussion within the "box" I've described above (individual components, plus a "wrapper" eval engine).

Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project

Top of the page Bottom of the page
JeffMcClure
Posted 3/2/2015 8:32 PM (#6152 - in reply to #6150)
Subject: MZ3 Confluence Enhancement Project



Friend

Posts: 28
25
Location:
USA: KS, Overland Park
Some other confluence levels to consider:

1. Opening Range
2. Yesterday's High
3. Yesterday's Low
4. Yesterday's Close
5. Round numbers - 50, 100, 200, etc.
6. Trendline confluence with any of the levels mentioned
7. MACD divergences at any of the levels mentioned
8. VWAP
9. Multi-day pivot ranges i.e. the pivots are calculated using 3 days worth of data for instance rather than just the prior day's data

May want to also consider some of these levels converging in the context of multiple time frames.

Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project

Top of the page Bottom of the page
SalimHira
Posted 3/3/2015 5:55 AM (#6155 - in reply to #6150)
Subject: MZ3 Confluence Enhancement Project



Veteran

Posts: 183
100252525
Location:
USA: MD, Columbia
There is also Frank Ochoa's ADR levels that may be considered as we are on the topic.

Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project

Top of the page Bottom of the page
ThomasHelget
Posted 3/3/2015 8:55 AM (#6156 - in reply to #6155)
Subject: MZ3 Confluence Enhancement Project



Elite

Posts: 473
1001001001002525
Location:
USA: NY, Baldwinsville
Guys:

Yes, I like this.

On a more basic level, I have found that if today's Money Zone High or Low is near a Weekly (or better yet, Monthly) Money Zone High or Low and the two (or three) are penetrated, the stock usually explodes from there. If these levels were confluent with even more levels then better yet.

One should also note that Mel's DPL levels are a bit different in Real Time versus EOD. I have tried to duplicate his effort from his article and had major difficulties in the interpretation of his intent. But since he is a TT member perhaps he might be willing to share his coding for Real Time.

Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project

Top of the page Bottom of the page
ThomasHelget
Posted 3/3/2015 12:33 PM (#6157 - in reply to #6156)
Subject: MZ3 Confluence Enhancement Project



Elite

Posts: 473
1001001001002525
Location:
USA: NY, Baldwinsville
Gentlemen:

Here is an example from today on UNP (and it is just one of four I easily identified about five minutes into the open - the others were GD, STX and XON). The daily Money Zone relationship was an Inside Day which is a bit hard to see because the Lower Money Zone for 3/3/15 (in blue) overlays the Lower Money Zone for the month (in green). Above the Upper Money Zone for 3/3/15 is the Weekly Lower Money Zone (in brown). Once UNP opened below today's Lower Money Zone it kept chugging along racking up an nice profit. A protective stop could be placed above the confluence of today's Lower Money Zone and the Lower Money Zone for the month which never came close to being penetrated.

I love these plays as they almost always work out.

(UNP_Chugging_Down_the_Line_.PNG)




Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project


Attachments
Attachments UNP_Chugging_Down_the_Line_.PNG (44KB - 4 downloads)
Top of the page Bottom of the page
JeffMcClure
Posted 3/4/2015 8:45 AM (#6158 - in reply to #6150)
Subject: MZ3 Confluence Enhancement Project



Friend

Posts: 28
25
Location:
USA: KS, Overland Park
Another level not yet mentioned is the Yao Ming High / Low. A characterization used by Gavin at Tradeguider, Yao Ming is very large volume. If you draw the High / Low of the YM bar, you will find that price frequently respects these levels in a reliably tradable way. And when some of the other levels mentioned meet at the YM H / L, the reliability of respect that price has for the level increases quite nicely. And if the respect for these levels is accompanied by indicator divergence, the weight of the evidence increases even more.

(Yao_Ming_High_Low.png)




Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project


Attachments
Attachments Yao_Ming_High_Low.png (207KB - 3 downloads)
Top of the page Bottom of the page
JimDean
Posted 3/4/2015 8:58 AM (#6159 - in reply to #6150)
Subject: MZ3 Confluence Enhancement Project



Owner/Admin

Posts: 3925
2000100050010010010010025
Location:
USA: GA, Lawrenceville
Thanks for the contributions guys ... keep it up. I won't be able to get to this project for a while, but I figured that it would call for plenty of initial discussion.

Here is the big "gotcha" regarding using a plethora of "raw" levels ...

We're looking back in time for past as well as recent levels, set by whatever mechanisms have been invoked. The farther back we look, OR the more "methods" of finding the raw lines that are utilized, the "hairier" the job gets ... code grows geometrically.

Also, if we get a gazillion "raw" level lines, it will become sort of a "cloud" rather than "stripes" ... that's still subject to analysis, but we have to be careful not to "bias" the cloud with lots of low-probability candidates that would dilute the impact of the more "refined" ones. For example, if we included every Daily Open, Close, High and Low while on a *daily* chart, the tool would likely be useless. However, those four values might be very important to consider on a 1-min chart.

So ... as you suggest alternatives, please also attach "qualifiers" such as "Daily OHLC for Xmin bars or faster". That is ... does the method you propose work *better* if it's based on "higher timeframe" bars.

Furthermore, please comment, for each method, on how far back in time (ie number of current-timeframe bars) you think the search should go ... and how much if any "weighting decay" should be applied to the importance/relevancy of each line, as a function of how far back in time it goes.

Part of the confluence routine will include fixed user-input "weighting" factors for each method employed ... 0 weight turns off that method. Also there might be a "time decay" factor input for each method.

Thanks again for your interest and contributions to the discussion.

Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project

Top of the page Bottom of the page
SalimHira
Posted 3/5/2015 8:39 AM (#6160 - in reply to #6150)
Subject: A Confluence Project



Veteran

Posts: 183
100252525
Location:
USA: MD, Columbia
I'd like to clarify or find out the interpretation how folks tend to qualify a "line". I have seen two methods,

1). Nirvana says, when the MZ3 confluence lines are touched often, more powerful;

2). Professional traders find, any untouched lines tend to be more powerful, as "fresh", their saying is, if the lines are touched often, it weakens the support/resistance and easily penetrable, similar to banging on a screen door multiple times.

Jim, your Frontrunner code has something similar to "weighting" combos for RSI, Stoc, CCI, Momentum, etc. which I have found to be very leading if tuned appropriately and "interpreted" correctly with price action (hopefully, frontrunner code too is updated to the latest technologies in near future).... but the confluence project "weighting" would seem to be different, I presume, once some kind of guidelines are established per your request above.

Thank You.


Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project

Top of the page Bottom of the page
ThomasHelget
Posted 3/5/2015 9:06 AM (#6161 - in reply to #6159)
Subject: MZ3 Confluence Enhancement Project



Elite

Posts: 473
1001001001002525
Location:
USA: NY, Baldwinsville
Jim:

Good point.

John Carter has an indicator he calls the "Voodoo Lines". It draws a scadzillion lines on the chart.

Of course the more lines you draw the more likely you are to read into the analysis that these lines actually provide support and/or resistance when all that really is going on is simple chance coincidence.

Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project

Top of the page Bottom of the page
JimDean
Posted 3/5/2015 1:32 PM (#6163 - in reply to #6161)
Subject: MZ3 Confluence Enhancement Project



Owner/Admin

Posts: 3925
2000100050010010010010025
Location:
USA: GA, Lawrenceville
Eric just asked a very relevant question ... how will these improvements help create better trades?

Both this MZ3 project re confluence, and the other thread's MZ2 enhancement, envision improvements to the "indicator" feature of the MZ plugins, which are at the heart of the tools.

Those enhanced versions could be substituted into the strat's provided with MZ2/3 (no diff) as a starting point for methods to trade them.

There are of course many other strat possibilities on how to trade, if you "know" the S&R levels. That's really a different kind of project imho ... I've got a some ideas beyond what MZ2/3 offers (but not contrary to theirs). I'm sure other folks, esp people like TomH, might have other ideas as well.

Some of those alternatives might call for a System &/or Stop routine to be created to implement them ... again, a separate project but it would make sense to do. That would be a lot less complex than the two projects already posted.

If anyone has ideas as to how to improve MZ2/3 strategies, using the existing plugin functionalities, it would be appropriate to post them in a new thread in the Money Zone PlugIn Room (not in this Room ;~).

Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project

Top of the page Bottom of the page
JimDean
Posted 3/5/2015 2:41 PM (#6165 - in reply to #6150)
Subject: A Confluence Project



Owner/Admin

Posts: 3925
2000100050010010010010025
Location:
USA: GA, Lawrenceville
I mentioned in my prior post that I believe that the "strategies" which would use the "enhanced confluence" tool would be very similar to strategies using the MZ3 confluence tool ... that is, the logic related to entry/exit decisions vis a vis Support / Resistance levels would be the same, regardless of how you derive the S/R levels. I'm presuming that is the kind of discussion that you are referring to.

Discussions about that specific methodology, and strategy variants of it, would best be conducted either in the MoneyZone Room, or in the General Methodology Room.

Adding entire Rooms is something I usually reserve for something with a bit broader focus than just "how do I best utilize Confluence in a Strategy" kind of thing.

Hoooow-ever ... when/if the tool is created, if there are a meaningful number of recipients, there will be a "support" Room for it that will be limited to its owners (that's S.O.P) ... I suppose that Room could be used for strat-devel discussions.

But, if those discussions really are universally applicable to use with other S/R identification methods (MZ, WT, etc), then I would likely move them into a generally-accessible Room, so that other TT participants could benefit and join in.

The same comments apply to the proposed MZ2 enhancement project.

Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project

Top of the page Bottom of the page
JimDean
Posted 3/5/2015 3:08 PM (#6167 - in reply to #6150)
Subject: MZ3 Confluence Enhancement Project



Owner/Admin

Posts: 3925
2000100050010010010010025
Location:
USA: GA, Lawrenceville
I want to encourage discussions about strat devel, and I also want to keep things "usefully organized" on the forum.

The "general SOP" is mainly in my head ... it boils down to something like:

1. If discussions need to be limited to a group of people who have all "paid for" something (like a video series or a signif custom routine, then I create a limited-access Room for those discussions.

2. If several threads get created that all seem to have significant participation and all have a common general theme, and if it's a theme that is likely to have ongoing significance, I create a Room and move those threads to it.

When I do add a Room, I notify the group so everyone can go in and update their notification preferences to include it. This is sort of a hassle for folks, so I try to wait till I need to do a few new Rooms, then do them all at once.

Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project

Top of the page Bottom of the page
JohnB
Posted 3/5/2015 7:30 PM (#6168 - in reply to #6150)
Subject: MZ3 Confluence Enhancement Project


Friend

Posts: 29
25
Location:
: ,
Hi Folks just thought this might be helpful. My all day fib indicator that I posted in OT some time back. Some say it is invalid because it continually updates the session extremes but I think if you get used to looking at this it does box things up reasonably well and might be helpful in the fib part of the quest.


Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project


Attachments
Attachments hafib.jpg (275KB - 17 downloads)
Attachments JBDIDFib8th.txt (13KB - 13 downloads)
Top of the page Bottom of the page
JimDean
Posted 3/6/2015 12:35 PM (#6172 - in reply to #6150)
Subject: MZ3 Confluence Enhancement Project



Owner/Admin

Posts: 3925
2000100050010010010010025
Location:
USA: GA, Lawrenceville
I've moved this thread to a Room that I previously created for discussions about group-project development discussions.

The core distinctions for this enhancement to the MZ3 Confluence engine include:

1. ability to evaluate S/R lines based on different time-boundaries (ex: day, 3-day, week)

2. ability to evaluate S/R lines based on different identification methods (ex: MZ, WT, Fib)

3. ability to input weights for the various types and timeframes of S/R lines

4. some kind of automated weighting based on aging or strength, etc (diff for diff methods)

Please post other ideas here ... note that the MZ2 native-levels enhancement is a different project.

Thread moved by JimDean on 3/6/2015 12:29 PM from OLang, OScript & OPilot > Writing OLang Plot-Indicators > A Confluence Project

Top of the page Bottom of the page
Jump to page : 1
Now viewing page 1 [50 msgs/pg]
( E-mail a Link | Printer Version | Search Room )

Owner of site: Jim Dean -- Forum content is confidential, and may not be distributed without written permission.