Thanks Userque...but I think we all know that.
I was wondering if you knew your system well enough to identify the specific pattern that took it from a sell to buy in one day.
My bad. I have no way of knowing, generally, what folks do and don't know. There's a broad spectrum of knowledge here from novice to expert.

The spectrum is likely wider with machine learning trading as not many are discussing such things here.
The system does not label patterns in a way one could say, "yes, yesterday matched pattern #487." The system looks at (so far) ~60 price/volume/date metrics. For example, percentage change for: O,H,L,C,O2, H2,L2,C2, etc. Also 5EMA crossover, ..., etc. Also New 52 week highs, ... etc. Also Advancers/Unchanged/Decliners, etc. Matching month?, etc. Etc.
It tries combinations of one or more metrics, and matches recent metrics to historical metrics--back-testing to see how well each combination forecasts; and finally settling on the 'best' combination. I simplified my earlier discussion by referring to 'all-this' as 'a-pattern.'
So there are really two questions you really
mean to ask, imo:
What metrics did the system match to which day(s) in the past?
I could've answered that question before I overrode that data.
