Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Thanks. I have a lot of screen time, a lot... Not screen time where I could give it 100% attention and lay money on the line though during those times. Geometry is a newer concept tome, one I thought was too dated, until lately when I actually looked at it.
There's actually a very simple way to reason why this doesn't work.
You've drawn at least 4 non-parallel lines in your diagram, which require 8 model parameters to specify. There's probably some latent (hidden) model in your mind for how many lines you need to draw, so the true number of model parameters you're using is probably in excess of that.
What you want is to have a large probability that your 8 parameter model explains the data i.e. the marginal likelihood. The "ground truth" model (imagine this is the mathematical function that God uses to generate market prices) doesn't necessarily have to be the one that maximizes this quantity, but we want at least a good chance to know that we were right. The closed form way to compute your marginal likelihood requires integrating over all your model parameters, where the larger the number of your parameters, the faster the probability vanishes.
In practice you compensate this effect by gathering more data points, but by experiment design you've also chosen to predict something that has very few data points (number of 'trends'/turning points, i.e. points where the first derivative changes sign). The 'true' number of data points you have is the number of turning points. In the screenshot you attached, there's less than 15 of those starting from where you are fitting those lines.
Another way to think of this is, I want a model to predict @artemiso's location (market price). I can use a model that spreads out the probability mass over a very large area (your 8 parameter model), i.e. I pick a model that predicts artemiso's location by continent.
Your model predicts that I'm located in North America. How do I know you actually have an intelligent model that accurately guesses where I'm located? You ask a BMT moderator to confirm where artemiso logged in at 9 PM (collect data) and he confirms to you that artemiso is located in North America, so your model guessed correctly. Well duh, your model made such a broad guess that you would need to try this thousands of times before anyone is convinced that you have any predictive power.
Compare this to a model that predicts that I'm located on the 40th floor of 233 S Wacker Drive in Chicago. And somehow you find out that I actually am there. All you need is 1 data point to have very strong confidence that your model didn't arrive at its prediction by complete random happenstance.
Sorry, got excited. No one ever made money in the markets from geometry concepts. I need the next whiz bang indicator for $1495, I forgot. Point taken.
Can you send a link? There is a lot of Al Brooks information in that thread. I think Autochartist is a recent site that send out alerts based on geometry.
I think I read all the treads relating to him here. I can’t recall. Go trough all of them, you will find good stuff about it here if you have time...I didn’t realize it before but I thought this topic is discussed in some of Al Brooks treads. Regarding XOM, did you make some sort of correlation with CL or USO or just based on XOM only?
Pulled up a 240M chart on trading view. I drew an up sloping pitchfork from obvious price pivots and expanded the lines out (warning lines). I could see price had been respecting those levels and it was looking like it would break higher as it was approaching the up trending channel below... you would be monitoring on a smaller time frame chart and use the entry method of your choice.
To asses geometric patterns of ANY kind, simple or sophisticated, you MUST have square scaled charts. That is to say that the x and the y axis have to be valued in the same sized (square) increments. Time and price must be equal measures.
A few guys have tried geometry before you. One that you might want to consult is W.D. Gann, a contemporary in this field that I have enormous respect for is Earick Bean. Both of those guys will provide you with mantra and nuance in the field of geometric patterns. When you are well immersed in their work, then consider price action as a psychological response to visual stimuli and you can decide what to develop into your actual trading.
That phrase "actual trading" is also a critical point of focus. Consistent with the spirit that our host @Big Mike intends I want to encourage you without being too harsh. BUT the harsh real world of the last bastion of true capitalism...trading financial markets... will rip your balls off and leave you to die if you insist that you know anything about this vocation based on a theory that you developed back fitting lines to a price chart.
Past that, and yes there is more. You would help yourself if you understood most of the common logical fallacies that trick honest folks into ruin. Google "logical fallacies chart" and run your trading logic through a short course that includes confirmation bias and anecdotal false cause.
All of that said, I do believe that geometry can be a valid and prosperous approach to trading when guided by a skilled hand. To be a skilled hand requires a bit more than jotting some lines at 3am. Learn what you are talking about and then become skilled through thoughtful application of actual trades.
Last, your response to this will indeed, signal a tremendous insight into how respected, thoughtful and experienced traders will form opinion and response to you.
I recommend trading ES between 9 PM CST and 5 AM CST. Also, please use order sizes of prime numbers ending in 7. 
