Quote from Phenom 300:
I could - but something tells me that you already know about my idea, no?
Trading anything will work at least 50% of the time, in a business where there are only two (2) possible directions to discern.
And, that brings up an interesting thought (at least to my mind). 100% of the charts that you show contain 100% positive results for profit. Yet, only 60% of the distribution is positive for profit. Those numbers don't add up in my mind. I don't know - maybe I'm just an overly attentive fool, but 100-60 = 40. Which means that at least 40% of the charts you post, should show (at minimum) that only break-even was achieved, no?
Which brings up an even more interesting question for an old mathematician like myself:
What is the Density Distribution of the 60%?
What you showed here, was the "frequency" of the distribution. Frequency, is not Density. Frequency is merely a ratio of occurrences within the distribution itself. Density, deals with the cyclical nature of the occurrences themselves - the literal "location" of the clustering of results.
In other words, does the 40% retain the same density characteristics as the rest of the distribution? If so, then yes you can build a predictive model from such data. If not, then all you can do is make an educated guess.
People often times hear the statement that you can be a winner 50% of the time, and still make money trading. Well.... that's only half the truth. The other half of that "truth" is what they never remember to tell you and that is: Only if the density of the losers are proportional in characteristic to the density of the winners. Something they still don't teach in Trading 101, these days (for some odd reason).
Bottom line, if your 40% is not clustered according to the above - you can easily produce a negative equity curve from such a trading method - without ever seeing it turn positive and even with a 60% statistical "chance" of being on the right side of the market.
A simple EA demonstration would prove my theory. Do you have an EA for this method?
Sure, but for the reasons stated above - it would mean absolutely nothing with its counter-part, Density Distribution.
BTW - the Density Distribution of my Indicator is built directly into the Indicator itself. So, you can see the Distribution relative to actual price behavior directly in the pic that I provided.
In other words, the prove is in the price action that follows, not in the frequency of the numbers output in a window, which are not attached to any quantifiable historical price action.
Case in point: Notice the slope of Red Histogram -vs- the slope of the Green Histogram, relative to historical price behavior (or, what people incorrectly call the Trend).
You can clearly see that the extremes of the Red Histogram form what I call a "compression slope" as its recent historical values move closer to zero (0). That's the hidden truth about the pivot that's about to hit the market.
The steeper the compression slope, the higher the probability for a change of direction in price. No compression slope within my Indicator can retain a zero (0) value for very long - and that is how I am able to accurately predict market behavior, before it happens - regardless of the type of market in play - bull, bear or flat.
Statistics are only of value, when they are thoroughly understood.
No, I don't know about your idea. Please elaborate so all may benefit.
No, I don't code EAs.
I understand your point about density distribution.
"In other words, the prove is in the price action that follows, not in the frequency of the numbers output in a window, which are not attached to any quantifiable historical price action."\
I am not sure what you mean by that statement. The frequency distribution IS the price action.
Do you have a thread where you discuss your ideas? We can continue the conversation there.
