I just thought of this right now:
It seems we should have enough data to be able to analyze stock market trading and find out how often people are correct and incorrect in their betting. I mean individuals’ bets as assessed from individual positions. We could get into strategies later.
Anyone have an idea why no one has thought of this?