>It's not luck if you make consistent returns over 4 years.
Sure, if your sampling was unbiased. But here, it's not - so that user might well have been lucky. How many people have failed to make those returns, or any kind of return?
There's very obvious evidence that somebody can outperform the market, even to a massive degree. From what I know, there's very little evidence that a particular person can deliberately outperform the market - and daytraders, especially, cannot do so consistently.