I suspect that the reason you'd see a stronger "skill" signal in the fantasy results than in the sport itself is that professional players are a hugely skewed s

I suspect that the reason you'd see a stronger "skill" signal in the fantasy results than in the sport itself is that professional players are a hugely skewed sample. When a player starts doing noticeably badly, they get dropped. Face the professional hockey players off against non-professionals and you'd see the signal jump right out. Fantasy sites, on the other hand, aren't filtering out people that play terribly nearly as aggressively. So I'm not sure if this is the right kind of test. It's comparing against a signal-expunging process.