Disliked{quote} giovanni4000, As I said, building the database is based on the ZZ. I'm not moving it an inch. I count the bars, register their OHLC, count the fractals in between and their spaces, and add one fractal at the top (for Long move) or bottom (for Short move). That's it. I'm not moving anything or changing positions. So, to answer your question based on your chart: the ZZ's are ABDEG. AB - one fractal (top) BD - 7 fractals DE - 5 fractals EG - 3 fractals G.Ignored
Great, thank you very much. Can you explain this :
does it mean, we need to put the candle up/down information also beside the fractal it self ?
Please see the image below. Does it mean, from A to B, we must put to database the fractals + spaces like this ( but without brackets "{}" ) :
{0} 0 {1} 1 {00} 1 {00} 0
Another question:
As we want to do a "fractal match" calculation, and at that time there is no fractal exist, so we must compare the current
bars with the "spaces" (exactly "1" / "0" bars) that exist in front of the first fractal (compare to zz leg that has only exactly 1 fractal in the zz leg), from the database ?
Example:
*) We now at B-C zigzag leg.
At the first time, there is no fractal (bar number 1), and it must be a fractal at the end of current zigzag leg. Because we are at "up" going zigzag leg, so we are waiting for a "1" fractal at the end of B-C zigzag leg.
*) We are now at second bar, and we have bars : 0 in current zigzag leg, so we compare to database with {0} bar in zz leg that has only 1 fractal. But there is no such pattern in database, we are waiting.
*) Now we are at 3rd bar, and have bars : 01 , then we compare to database with {01} bars in the zz leg that has only 1 fractal. But from database, we get {010} in zz leg with only 1 fractal has more probability. So, we wait again.
*) And then we get bars : 011 , and still have no fractals. So, we compare to database again, looking for bars {011} in the zz leg that has only 1 fractal, so the bars pattern {010} is a wrong prediction. And from database, we get the more probability is {0110}. We are waiting again ...
Is it like that ? Or .... how do you calculate the "fractal match" probability if there is no fractal at all in current zigzag leg ?
*) And the same thing, if we get 1 fractal already, we compare to database with zz leg that has 1 or 2 fractals in it. In other words: comparing current fractal to "+1" / "+0". Example: we have "0" fractal now. We must compare to database, looking for more probability of zz leg that has "0" with "00" and "01" fractals.
Thanks in advance.
Gio
Attached Image