I think that even if we found that children who play difficult games tend to be more intelligent, that still makes it hard to separate correlation from causatio

I think that even if we found that children who play difficult games tend to be more intelligent, that still makes it hard to separate correlation from causation. Do children become smarter because they play difficult games? Or do smart children enjoy playing difficult games? Or both?