[lionhearted]

I wonder if the causality might run the other way — like, if you imagine a set of "all possible commitment levels of sports fans" you'd have hardcore committed... moderately committed... middle of the road... somewhat fairweather... complete bandwagonner.

Something like that, eh?

Losing might strip away the fairweather/bandwagoner fans, leaving only the people with a natural predilection to commitment. Winning might attract more bandwagon/fairweather fans.

You could probably test this by finding multiple teams that had around a 50% win percentage and finding fans that adopt them in early adulthood, like someone who goes to college in Indiana starting to root for the Indiana Pacers (basketball), etc.

If you build a sample of around 1000 people that adopted different middle-of-the-road teams, then followed them longitudinally, you could see whether they increase in commitment, drop out, or stay about the same with rising and falling fortunes of the team.

A lot of work, though. No doubt certain experiences shape perception of fandom, but I imagine it's different populations — not just the same type of starting population being effected by different events.