Someone can win at roulette and make more money than the average player over some measurement period, but nobody can be good at roulette (when properly implemented and stuff). Stocks are somewhat possible to be good at but results are mostly random and the fee you'd pay is usually way too much.
How would you know it is or is not luck?
> roulette
Has no winning strategy - it's very different.
The winning strategy with stocks is understanding the underlying businesses better than the average investor. Peter Lynch's Magellan fund did consistently better than others because Lynch had insights others didn't. When others figured it out, Magellan's returns retreated to market levels.
I.e. investors can do better than average if they have insight others don't have and stay below the radar.
It's hard to know in the moment, but almost every promising fund has subpar long term results. Whether they lost their touch or were lucky in the first place, it means that seeking out promising funds is a very bad way to find a place to put your money.
The number of funds with significant valuable insights is low, and the number where those insights are bigger than the fees is lower.
Anyway my point was just that a big spread of outcomes doesn't prove that significantly different skill levels exist.