With bitcoin, it can. Tea leaves and astrology can too. Anything can fairly accurately predict bitcoin, the only condition is that enough speculators believe it can, and get the same "analysis" (I guess that rules out tea leaves) and therefore act upon the prediction so it will automatically become true. After all, to predict bitcoin exchange rates, you have to predict the actions of other investors, as thats pretty much the only variable. If enough speculators believe the moon phase will have an effect on bitcoin, then it will.
You got the point, yes "Anything can fairly accurately predict bitcoin" in the past because it is easy to overfit the market with most trading theories, generally speaking when you have two or more competing trading theories which make exactly the same predictions, the one that is simpler is the better & more profitable, i say again that complicated theories can fit everything on the past with misleading results (and losses) on unseen data.
I'd like to add two words about "elliott wave theory" that is frequently used here on the forum, i have mixed feelings about this theory. On the one hand it can sometimes enable forecasters to do fantastic market forecasts. On the other,
it offers too much scope for personal prejudice and bias to influence the wave count, for this reason i don't find it useful for short term day to day trading activities. I think that one should
not expect the details of the price action to fit the theory perfectly, instead you should emphasized the importance of looking only at the big picture.
Cheers