Here is a follow up post on the one pattern test surf chatted about:
--------------------------------------------------------------------------------
Quote from tobbe:
How do you programmatically decide that you were in a trend and that it has now ended so that you can start counting (again)?
Thanks, T
--------------------------------------------------------------------------------
Thanks for asking.
There is more to the programming I set up for surf to learn about TA and make money with TA.
In trend monitoring and analysis, the variety of patterns most people ID are actually parts of a foundational pattern.
To either invent or deduce the foundational pattern takes a lot of work. One factor is the nature of market data.
When science was first applied to market data the Dow Theory emerged. Dodd and Granville at that time, also implied and inferred the market's paradigm (or algorithm).
For long and short trends to exist, they also must fail at some point. The paradigm emerges from this trend monitoring and analysis contstruct. The failure has to be in a trend.
Couple this with the market's granularity.
What you get are a correlation with volune peaks and troughs and price extremes in the foundational pattern.
To have a "container of price" the least complex "shape" in two dimensional "space" is the parallelogram. It takes a while to use what you sense with the inference of your mind to be able to percieve. By continually examining the market, your inference gets refined then honed and then it has no noise, no anomalies and no flaws.
At this point the three bar profit making sequence only occurs in three regions of trends: during the trend or just after the peaks that either begin, form or end trends or during the two troughs within trends.
So I did not have to "gamble" in any way with respect to surf and anyone of the quant programmers he believes have skills of any sort.
None of these people would pick the three bar sequence during a trend, They could pick fake trend endings. for either the troughs or the second peak in the foundational pattern parallelogram.
It turns out this three bar sequence also works on these erroneous places, if and when it appears. I had to outwit surf and his quants by using a three bar sequence that worked at every place they could and do make mistakes.
For me, it is okay to use TA that is actually correct or use TA where my adversaries (humorously, surf or quants who perform as surf says) make mistakes.
So that leads to the mistaken early exit. Well, the unique and most difficult part of trends is the place where overlap of trends ends and, naturally, it is concurrent with the mistaken early exit. I do not make that mistake, but the three bar sequence could happen there too. If surf and his quants make the mistake there, then that mistake is the one that makes them the most money after there is no BE ( price zooms past it before surf and his quants do not see what could have been a momentary retrace (which can't happen in reality. See more below).
I had to post this three times in this thread before I realized I had to post the maths so not 100% of the readers would keep ignoring it.
I do have the hope (a humorous condition) that a quant will do the equation and logic. If they do, then they backtest which proves only that money is made. Then MAYBE a forward use would occur. during this use of curiiosity, all the things mentioned above will be seen also for the first time.
this is like looking on a low limb as noticing the essence of the "Cheshire Cat". Any mathematician worth his salt would be reliving the writing of "Alice in Wonderland".
Serendipity had you show up while surf is in the muggy south. and surf will never read this or ask a quant to go through all this discovery.
He actually met, a few years ago, a person at a smokey party. The person told him how this stuff works in the presence of other users. Nothing sank in then either. Later in the same town, a global meeting of the users occurred. and surf didn't show.
If you annotate and monitor trends in volume and price in detail, you get to see the Cheshire Cat show up over and over. Watch how there is no pause to do a BE on the third bar. They might try a limit order at BE but it will become a market order at a higher price due to the effect of granularity under contitions of market "acceleration".
all that had to be induced was surf setting the test conditions and sure enough he did.
LOL..........
