Curious how they quantify ‘attractiveness’. Is it self-reported, or are they attempting to use some kind of qualitative metric to rate the attractiveness of participants?

If the former, you could just as easily draw the conclusion that more confident people spend less time gaming.