Kept It Simple - Is This Something??

Discussion in 'Index Futures' started by mrpace, Dec 10, 2002.

  1. I like this forum very much and I realize that everyone can post what they feel is OK but this is tread is just too dumb to ignore.
    This guy is asking " what do you all think of my system, but I will not tell you anything about it. Do you think it will keep working in the future ?"
    This is a dumbest question of the year.
    Those who tried to help out despite of unwillingness of person who asked for help to participate , have my admiration.
    #41     Dec 12, 2002
  2. you're so crual :)

    #42     Dec 12, 2002
  3. He gave statistics of its trades but as I said they are not detailed enough. You can conduct some inference if you have all the recorded datas even if you don't know the underlying mecanism. That's what statistical inference cope with fundamentally.

    #43     Dec 12, 2002
    #44     Dec 12, 2002
  5. mrpace


    Whatever....sorry to waste your time.
    #45     Dec 12, 2002
  6. Exactly Walther. "I'm not going to tell you anything about my system but but do you think its going to work??" LOL.
    #46     Dec 12, 2002
  7. GD2KNO


    the secrete is not to move the market - but to let the market tell you where it is about to move.

    Just need to learn to read the map - seems some are quite good at it - others seem to come and go.
    #47     Dec 12, 2002
  8. mrpace


    A select few on here appear to be as arrogant as they come. I post some results from some trades I've taken over the past 10 or so days (well over 100 trades in all), and look for some feedback on the results. How is this such an "out there" request???

    There should be MANY traders out there familiar with statistical analysis of results who could give me a ballpark figure as to how many trades I should have under my belt before I begin to think I have something that might continue into the future. Or they could tell me that it is a positive that my trades in both ES and NQ have similar win percentages....etc...

    Analysis of results (even very simple analysis) does not need the underlying method to be revealed in order to be useful.

    I genuinely thank those who responded in a helpful way to my post. To those who felt the need to clutter this thread with their high and mighty know what you can do.
    #48     Dec 13, 2002
  9. Problem is really that the period of the backtesting is not long enough. Also when back testing there might have been a mistake in the way it is programmed. Since you have done real time trading by now, we might exclude the programming error. A common error here is an open range break out system tested in TS that does have a low "bouncing tick" setting. Otherwise it looks like you make money, but just test it on 5 years of ID data.

    #49     Dec 13, 2002
  10. The statistic you give has a good biais impression but I can't really say nothing else since the detailed trades are not known.

    For example on 10-Dec you made 11 trades for 6 winners:

    10-Dec 6 11 $297.20

    If some of the trades are correlated (for example if you take some positions one after another because you are in a winning position) this biais must be substracted so it's better to sample randomly although it seems to be a rare practice in trading, statistically it would be more significant.

    #50     Dec 13, 2002