fortes fortuna juvat

[llIHeroic]

@stepan7 - I may very well have nothing practical for you; I don't clearly understand what you specifically are trying to do and how far you've already gotten. If you're trying to fit together some components of this method into working automation you will have to test and discover what works as Spyder and Jack traded this manually and never gave instructions on how to do that. I could possibly weigh in on individual components if you have specific questions which is what svrs asked me to do with EEs.

@river : The TL thread early on has Spyder leading a lateral differentiation drill. The discussion gets confusing and goes off track at points but ultimately he confirms the answer to the drill which is laterals that start with a sym pennant and touch the lateral boundary exactly before it's pierced at all are distinct from other laterals.

@Sprout ; This is a complicated question and basically more than half of the TL thread revolves around maintaining fractal integrity. Spyder states he found himself jumping fractals in his earlier work up to 2009 I think and redefined the criteria of what makes a Tape, a Traverse, and a Channel. He posted bar start and end times spanning multiple days which constituted a "Channel" and talked a lot with Breakeven in the TL thread toward the end of his posting on how to complete one of the drills he posted.

As far as I know the was the biggest difference from jack who never paid this much attention to objective fractal integrity and instead just focused on what was visible and never seemed to teach such detailed rules for building specific fractals.

 

[stepan7]

@stepan7 - I may very well have nothing practical for you; I don't clearly understand what you specifically are trying to do and how far you've already gotten. If you're trying to fit together some components of this method into working automation you will have to test and discover what works as Spyder and Jack traded this manually and never gave instructions on how to do that. I could possibly weigh in on individual components if you have specific questions which is what svrs asked me to do with EEs.

I have one practical question - Do you have any ideas or algorithms how to automate Gaussian drawing?

 

[Sprout]

Stats says that 95% will not make it.

But here is the trick, little understood - As soon as trader deeply and honestly accept he is the part of the number, money will come in this business.

Understand?

Enough to know that opening a retail brokerage and promoting day trading makes better money than Vegas - just by taking the opposite side of your customers trades.

No empathy = Dom reptilian brain /non-dom mammalian brain.

 

[Sprout]

My view of the day.

First attempt at understanding how to annotate volume bar by bar.

 

[stepan7]

My view of the day.

First attempt at understanding how to annotate volume bar by bar.

View attachment 173478

Looks good.
Next time mark your trades (real or paper doesn't matter) in real time, but not in hindsight.
That's what develops the sports memory Jack talk about.

Are you on West Coast?

 

[llIHeroic]

Automating the fastest level of visible gaussians divorced from context is probably doable without much effort. But qualifying the fractal of the gaussians with geometry let alone pace considerations seems like it would be almost impossibly complex; if we're talking about replicating the procedures Spyder last taught on TL before going off the forums.

Starting from a potential P1; the traditional way I learned to draw fast gaussians is mechanical. You need an increasing volume bar and the cross of the previous RTL in order to form the x2x. Then you need to wait for a non-dom price movement; FBP/Hitch/Lateral or non-dom translation to assign the 2y [i.e. doesn't violate the geometry of the beginning of the sequence (P1)]. Then you simply need more increasing volume on dominant translation.

After the completed sequence though an x2x in the opposite direction doesn't always follow. The sequences can repeat many times over before sentiment changes. Type D Trend if you will in RDBMS nomenclature. So every increasing volume bar against the current gaussian can either materialize into a seq in the opposite direction or simply another 2y. You will be able to tell definitively if you get an immediate return to dominance which makes a higher high / lower-low or not. Or likewise if you violate the P1 then the previous seq. must have ended already. See image for example of this process and some notes. Again, when pace is very high this stuff is a lot easier to apply but the pace drops and low volatility complicates things 100x.

You can automate a basic gaussian tracker but to get meaningful information out of it requires a lot of geometrical context and look-backs to confirm which data set prevailed when more than one possibility is on the table. At least going by the final rule-set Spyder seemed to be using and teaching before he stopped posting. Seems like this is one of the main reasons automation never seemed to be looked into or talked about a lot. Spyder coded his own set of indicators like laterals; symbol for IBGS, formations, peak volume, etc. to make annotations easier over time. Seems like if there was a way to format all this data into a fully mechanical procedure he wouldn't have stopped there but that's just my 2c.

 

[stepan7]

Automating the fastest level of visible gaussians divorced from context is probably doable without much effort. But qualifying the fractal of the gaussians with geometry let alone pace considerations seems like it would be almost impossibly complex; if we're talking about replicating the procedures Spyder last taught on TL before going off the forums.

Starting from a potential P1; the traditional way I learned to draw fast gaussians is mechanical. You need an increasing volume bar and the cross of the previous RTL in order to form the x2x. Then you need to wait for a non-dom price movement; FBP/Hitch/Lateral or non-dom translation to assign the 2y [i.e. doesn't violate the geometry of the beginning of the sequence (P1)]. Then you simply need more increasing volume on dominant translation.

After the completed sequence though an x2x in the opposite direction doesn't always follow. The sequences can repeat many times over before sentiment changes. Type D Trend if you will in RDBMS nomenclature. So every increasing volume bar against the current gaussian can either materialize into a seq in the opposite direction or simply another 2y. You will be able to tell definitively if you get an immediate return to dominance which makes a higher high / lower-low or not. Or likewise if you violate the P1 then the previous seq. must have ended already. See image for example of this process and some notes. Again, when pace is very high this stuff is a lot easier to apply but the pace drops and low volatility complicates things 100x.

You can automate a basic gaussian tracker but to get meaningful information out of it requires a lot of geometrical context and look-backs to confirm which data set prevailed when more than one possibility is on the table. At least going by the final rule-set Spyder seemed to be using and teaching before he stopped posting. Seems like this is one of the main reasons automation never seemed to be looked into or talked about a lot. Spyder coded his own set of indicators like laterals; symbol for IBGS, formations, peak volume, etc. to make annotations easier over time. Seems like if there was a way to format all this data into a fully mechanical procedure he wouldn't have stopped there but that's just my 2c.

We have different point of view on what the word "algorithm" means.
But I agree with you in one thing - "it would be almost impossibly complex" to automate Gaussian drawing.

 

[Sprout]

Looks good.
Next time mark your trades (real or paper doesn't matter) in real time, but not in hindsight.
That's what develops the sports memory Jack talk about.

Are you on West Coast?

Salmon Nation !!

Here's a couple of debriefs,...

 

[Sprout]

Automating the fastest level of visible gaussians divorced from context is probably doable without much effort. But qualifying the fractal of the gaussians with geometry let alone pace considerations seems like it would be almost impossibly complex; if we're talking about replicating the procedures Spyder last taught on TL before going off the forums.

Starting from a potential P1; the traditional way I learned to draw fast gaussians is mechanical. You need an increasing volume bar and the cross of the previous RTL in order to form the x2x. Then you need to wait for a non-dom price movement; FBP/Hitch/Lateral or non-dom translation to assign the 2y [i.e. doesn't violate the geometry of the beginning of the sequence (P1)]. Then you simply need more increasing volume on dominant translation.

After the completed sequence though an x2x in the opposite direction doesn't always follow. The sequences can repeat many times over before sentiment changes. Type D Trend if you will in RDBMS nomenclature. So every increasing volume bar against the current gaussian can either materialize into a seq in the opposite direction or simply another 2y. You will be able to tell definitively if you get an immediate return to dominance which makes a higher high / lower-low or not. Or likewise if you violate the P1 then the previous seq. must have ended already. See image for example of this process and some notes. Again, when pace is very high this stuff is a lot easier to apply but the pace drops and low volatility complicates things 100x.

You can automate a basic gaussian tracker but to get meaningful information out of it requires a lot of geometrical context and look-backs to confirm which data set prevailed when more than one possibility is on the table. At least going by the final rule-set Spyder seemed to be using and teaching before he stopped posting. Seems like this is one of the main reasons automation never seemed to be looked into or talked about a lot. Spyder coded his own set of indicators like laterals; symbol for IBGS, formations, peak volume, etc. to make annotations easier over time. Seems like if there was a way to format all this data into a fully mechanical procedure he wouldn't have stopped there but that's just my 2c.

Thanks for your comments, the context is supportive of the effort.

An initial strategy would be to scan for very paced trends. The angle of the rtl > 45 deg. I haven't come across the rocket threads yet, but I'd guess that would be the focus. I understand the qualifier of liquidity being a requirement for the market system of operation to work. The rocket threads from my understanding was based on creating the fundamental skill in a trader of using increasing pace to one's benefit - as well as and perhaps more importantly when it would be prudent to exit/reverse when pace is decreasing.

 

[stepan7]

Thanks for your comments, the context is supportive of the effort.

An initial strategy would be to scan for very paced trends. The angle of the rtl > 45 deg. I haven't come across the rocket threads yet, but I'd guess that would be the focus. I understand the qualifier of liquidity being a requirement for the market system of operation to work. The rocket threads from my understanding was based on creating the fundamental skill in a trader of using increasing pace to one's benefit - as well as and perhaps more importantly when it would be prudent to exit/reverse when pace is decreasing.

"I discovered something else, and that is that suckers differ among themselves according to the degree of experience.

The tyro knows nothing, and everybody, including himself, knows it. But the next, or second, grade thinks he knows a great deal and makes others feel that way too. He is the experienced sucker, who has studied not the market itself but a few remarks about the market made by a still higher grade of suckers. The second-grade sucker knows how to keep from losing his money in some of the ways that get the raw beginner. It is this semisucker rather than the 100 per cent article who is the real all-the-year-round support of the commission houses. He lasts about three and a half years on an average, as compared with a single season of from three to thirty weeks, which is the usual Wall Street life of a first offender. It is naturally the semisucker who is always quoting the famous trading aphorisms and the various rules of the game. He knows all the don'ts that ever fell from the oracular lips of the old stagers excepting the principal one, which is:

Don't be a sucker!"​