Fooled by Randomness talks a good bit about this, and argues it’s true - except you don’t know what to bet on. Hence the outlier successes will be from extreme risk takers, and for each such person there will be 1,000s of other gamblers that bet on the wrong thing.

Taleb does the math as well IIRC, assuming there are x hundred thousand extreme risk takers, and outlier “correct bets” are y% chance, then you will have a surprisingly high number of people with a long series of “correct” bets behind them looking like business geniuses, from pure chance & basic statistics.