Originally posted on 04/17/2023:

any program is only as good as its turning points, no matter what "neural network" it bounces around, the outputs
are only going to be based upon the API sets and subsets. So, if it's merely about scraping lines, I doubt it will be profitable enough to bother with especially considering the watchdogs at books who know how to read those patterns.

thinking outside the box about the nature of data - whether direct players or formations in play, or preferences, or eccentricities like dates, weather, historical trends, and coaching would be more interesting. It would also be undetectable. The machine or bot must be directed to store and pursue/compare perceived relevancies, or testing new possible contingencies in the formation of a new API.

There are deep, hidden facts in the historical data that long-time cappers think about subconsciously or consciously in their minds. Someday perhaps they will be integrated or tested; in the mean time if you're a winning bettor already, there is no reason to bother with this. Your mind is superior. If you're not, then ask yourself which is more likely, becoming a winning bettor using your own "neural network" on your shoulders, or using this stuff to get there? I know my answer.