Joe Doaks and ET Develop the Perfect System

Discussion in 'Strategy Building' started by Joe Doaks, Apr 8, 2008.

  1. What makes me ROFLMAO is that long-awaited SCT software that doesn't even give signals. If you can't stoplight it, you don't know how, or even if, it works. My code has exactly 44 SLOCs (logical) to do very close to what SCT tries to do.
     
    #61     Apr 10, 2008
  2. Sorry to send you back to the drawing board, professor, but 6x is possible. The tighter the range, the higher the multiple. On days with larger ranges, the higher multiples become more difficult. However, attempts to move beyond the 6x multiple results in decreased profits - due to increased execution costs and smaller profits per trade. And yes, 1x is more than reasonable - without ever performing SCT.

    - Spydertrader
     
    #62     Apr 10, 2008
  3. But you have to use S/R to do it.
     
    #63     Apr 10, 2008
  4. Jack has done a wonderful job creating the illusion of ephemera, hasn't he? :D

    Nope. :D

    - Spydertrader
     
    #64     Apr 10, 2008
  5. I won't argue with you what I understand less well. ServoTrader is designed to identify and follow rising or falling channels. I use several other codes to handle range-bound behavior. All of them numerical S/R based on market internals.
     
    #65     Apr 10, 2008
  6. I thought this had simplicity? :D Wouldn't simply automating channels have provided the same result - with fewer lines of code?

    - Spydertrader
     
    #66     Apr 11, 2008
  7. My existing peak detection code (bottom pane of the attachment) upon which I would build an automated channel constructor has 87 SLOCs (logical). Just a WAG but I am guessing that the logic to handle all the conditions of drawing, redrawing and testing channels would more than double that count. ServoTrader uses fuzzy logic cybernetic principles to elegantly overcome the ambiguities of channel construction and channel testing. The code I am showing you is NQ 1 minute code not optimized for ES 5 minute, but it makes the point that peaks alone, and therefore channels, are unreliable indicators of future price action. But a properly designed servo can read price's mind from the trail of numbers it leaves behind it, kind of like the pecker tracks you leave on the sheets reveal what you are dreaming about.
     
    #67     Apr 11, 2008
  8. man

    man

    die well ...

    (entertaining thread. most ironic in et history. though i am still
    not sure if all participants were as ironic as i think they were ...)
     
    #68     Apr 11, 2008
  9. Joe: I like your threads, simple in statement and deep in meaning.
    Here is my answer to the characteristics of a bar. I am interested only in bars which are.

    1. Shorter than the bars to the right and left of it, and
    2. Its head (feet) is higher (lower) than the head (feet) of its two neighboring bars.
     
    #69     Apr 11, 2008
  10. hypostomus, congrats on your peak detector, its very hard to make one that is accurate.

    We managed to make one too

    [​IMG]
     
    #70     Apr 11, 2008