Intentional Curve Fitting

Discussion in 'Strategy Building' started by Here2learn, Dec 18, 2012.

  1. e. interlocking fractal ratio.

    This invention was done with a deductive proof. Think of geometry or even quantum mechanics in theoretical physics.

    I prefer the volume proof but that involves too many other inventions; so lets stick to the lagging indicator of markets called price.

    You look at market patterns.

    By thinking, you discover how geometry was built.

    You use a simple theorem and its deductive proof. I will assume you can prove that two points determine a line.

    Step up to triple A league for a moment. Use Keynes's theory of algorithms. HS's must be two things: in kind, and complete as a set.

    Making money in markets is done with only one variable; it is the lagging indicator, price.

    We use significant points on price to do a geometric proof.

    Potential traders have a very low bar to pass to enter trading. They have to have parents that will set up accounts for them to trade using the parents money and a PC.

    This makes geometry proofs something to be learned in the future for some of these traders.

    Amazingly, only two adjacent bars are required. Some work for the proof others do not. You must go through this discovery by looking at a graphic already prepared by fifth graders in a one hour class during the first 15 minutes. Assume you did this.

    From this you discover that the Keynesian Paremetric Measure is a vector and not a scalor.

    Wow.

    We know know that connecting extremes of two bars give us the significant side of the market price container. I have seen published books on trading have authors and readers NOT know this. Publishers know less.

    On that sheet the fifth graders did, they invented turning adjacent bars that did not "qualify" into single bars that set the scene for adding the next bar to see if the two bars did qualify.

    They named the process "quishng". You can learn when and how to squish.

    To continue to prove (like be a smart algorithm), you do draw the line on the left side of the wo bars using Their extreme and ge vector result. This "defines" the sentiment of a lagging indicator, price.

    In geometry there is another theorem that you pick out and prove. The one for parallel lines. Boink! there it is. yu are not very far into the book you used which explained 20% of the proofs of geometry in the cheap skate school you when to.

    Okay clone or duplicate the line you drew and create the highest volatility container you can by chosing an extremest point on one of the two adjacent bars.

    My goodness what have you done. You ar- tick- u - lated the "simplest" and most foundational container of a granulated container. Now, simplify it more by taking granulation to account.

    You now have two one tick bars that are offset by a vector amount in algorthmic vector measure.

    In trends, volume advances your thinking. Trends have an even number of ends.

    Lets climb out of trend theory a little farther. if yo do the minimalist approach on this parallelogram by watching the futre move into the Present, you see more ticking going on.

    You know the sentiment and that could be one of two characters: dominant or non-dominant. The vector of the Right Trend Line (RTL) dictates the dominance.

    A person like a fifth grader can learn that trading is a left/right winning strategy. As a person gets olde he is inflexble and becomes stupider and stupider. He sesides in the up down world where he can no longer prove all trends have three moves.

    I assume here that I can tell you this fact above stated and you like all fifth graders can do the "ticking of adjacent bars just like they learned Arithmetic the Montisorri way. But I notice ET members are mostlky shitheads who do not know Arithmetic as yet. You are NOT because you want to learn to learn to be able to be an expert trader.

    All trends have three moves constructucted from the fastest fractal to the slowest fractal by interlocking the relative moves fractal to fractal.

    Use line drawings to prove this to yourself. Or google all this stuff that has been posted before and after the invention of the PC.
     
    #21     Dec 22, 2012
  2. f. overlap is proven and is there always trend to trend.

    there are soo many dumbies going through learning failure.

    Some of the laziest fuckers that can't trade have over 12,000 statements of their personal ignorance.

    Okay. This proof is more fun since it make obsene amounts of money possible.

    If you have a partner who wants to move out of Hoboken, DOMA violation or not, take it to a tattoo parlor.

    Have "to" tattoed three inches above the sternum.

    Then you, daily for a year use a washable sharpie to letter FTT on both sides of "to" Avoid the whole nipples. Work up higher.

    The world on the slowest fractal is still in its original trend since the three moves have not taken place. But they did take place first on a somewhat slower fractal.

    This proof is based upon the cognitive process of iterative refinement and not just "feeling" it is true.

    Look at the last move of these faster fractals. they all have an end of the thrid move.

    Lets name it. You see the move is inside the container and the point on the parallelogram the price moved away from was. unbelievably .... the third point of the construction which you used to draw the RTL and slip in in hindsight the LTL.

    Price is working its way to the LTL from the RTL. We sophisitacted thinkers do not do up or down. We just make money. We take profits on a trend ............. when .......... The trend FAILS. FAILS to do what?

    Get a chicken and go mid block down your street. Why can't the chicken get across the street. The leading indicator of operatioun of markets WILL NOT LET THE CHICKEN CROSS THE ROAD.

    We sak this golden chicken the answer. He says/she says .... "I just couldn't do another traverse after pt 3 of my parallelogram".

    Go to your scheduled meeting with your psychologist. As usual you discuss failure. Failure to .......... TRAVERSE because your psychologist know he/she knows this is the END EFFECT of every trend. I have the list and there are 35 names on it. Hey Zues!!!!

    Look under End Effects in the new 5 verson of MD's. They are now listed.......

    By deduction, since we know the identity of exits and entries. This FTT is where the new trend begins and the old trend ends.

    But the RTL has not been reached. We still have to go from a vlume peak (P1) to a volume trough (T!) to end the oveerlap just at the RTL BO on the way into the REGION in which pt 2 of the new parallelogarm may occur without having ....... an FBO.

    We always know we know. Casncel your appointment and get version 5.

    Ovelap begins on the FTT and ends on the RTL BO. We named names and reasoned why.

    Moral of the story: Humping right to left is just as much fun as up and down, and you can afford children.
     
    #22     Dec 22, 2012
  3. g. I snuck g in on you already.

    remember to use the washable sharpie.

    Number the children, too whereever they come from.

    VN has a boy!!!!!!! Urban renewal is alive.

    Check you local Improv for my next appearance.
     
    #23     Dec 22, 2012
  4. heech

    heech

    I would be interested in a reference / paper explaining the vulnerability of different strategies to curve-fitting...if you have a link to share!
     
    #24     Dec 22, 2012
  5. Back in the days of $200 round trips?
     
    #25     Dec 22, 2012
  6. SORRY!!

    See red correction above.

    This is the point French Fry was making about my posts.
     
    #26     Dec 27, 2012
  7. That is interesting,and I was wondering about the same long time ago...

    But thing about this.Tha market has only two conditions-trend or chop.Ive created two templates which are for trend and chop,and run them in parallel.

    Good luck!
     
    #27     Dec 28, 2012
  8. Heeh,
    I found the book "Evidence-based Technical Analysis: Applying the Scientific Method and Statistical Inference to Trading Signals" from Aronson to be the most interesting regarding curve fitting issues.
    It addresses exactly what you are looking for : In a controlled experiment, the amount of curve fitting is measured versus each property of the trading strategy (number of trades, average return and so on).
     
    #28     Mar 15, 2013
  9. Murray Ruggiero

    Murray Ruggiero Sponsor

    When developing a system the premise is what important. Why should the system work. You need a good theory and do experiments to prove the theory.

    Optimization can be valuable, but not in the way you think. The distribution of results over the optimize space tells you a lot about how the system should perform in the future.
     
    #29     Mar 15, 2013
  10. zordan

    zordan

    in my experience anything the is curve fit will fail. especially in so little trades and such a small time period.

    i am very wary of any developer that recommends optimizing the strategy every 1-3months. they never made money in my experience.

    my only good experience has been with robust strategies that can be tuned every year and work across various market conditions.
     
    #30     Mar 18, 2013