Learning patterns of the past stock price changes helps to predict the future stock price?

Not really. There are 3 folds-thought about this:

(1) Not only you but others are trying to predict using the past pattern data. The effect of crowd behaviors is enormous in dynamically changing stock price. “Compares current price vs the 50 day moving average” ? Yes, people already have done this. All others are analyzing the past data for their own benefits. And, stock professionals hire more machine-generated prediction — read this article from CNBC (“https://www.cnbc.com/2017/06/16/ai-assault-on-stock-market-ibms-watson-is-getting-into-etf-business.html”). So, what do we expect by using the same strategy as others in the stock market field? Answer is equal or less earning by using the same information.

(2) The past patterns will not happen again. More accurately, the exactly same patterns will not happen again in the stock market. The price of stock is like the change of human mind. Human changes their mind in unstructured, unpredictable ways. Further, the same past event will not happen again in the future. For example, the oil shock in 70s differs from any oil-related event in 2020 — the size of market, people’s education, the number of companies related are different from the past. So, what’s the point learning the past pattern to predict the future stock market? just for the pease of mind?

(3) The past patterns were created by human, but the future patterns will be dependent on human and machine. The machine gets evolved rapidly than human. The machine will have no mercy, no room for inefficiency, and no biological fatigue. The future patterns will be as optimized as the best patterns in the past. So are you creative enough to break the machine-generated patterns in the stock market? or are you bold enough to stand against the machine-powered stock price adaptation? I don’t think so.