I picked this topic because I'm a Caitlin Clark fan, but I didn't want to just confirm what I already believed. For my final project in INFO-I369, I set out to actually test the "Clark effect" instead of taking it at face value. I expected the data to complicate the story - attendance is never about one person. It didn't complicate it much at all. It backed it up.
The stat everyone quotes
League-wide attendance jumped during her rookie season, and games she played in consistently drew bigger crowds, home and away. That part isn't controversial. The question I actually cared about was whether that lift was really about her, or just the WNBA having a good year overall.
The comparison that settled it for me
The clearest evidence came from comparing the same teams' attendance depending on who they were hosting: attendance was consistently higher on nights a team played Indiana than on nights that same team played anyone else. Same city, same fan base, same arena - the only thing changing is whether Clark's on the floor. That's a hard pattern to explain away with rivalry games or season timing, because it shows up across different opponents, not just one or two marquee matchups.
What the regression confirmed
I ran a linear regression against attendance and Google Trends data, and Clark's presence held up as a statistically significant predictor even after controlling for win percentage, rivalry games, and where a team sat in the schedule. That's the part that mattered to me - her effect didn't disappear once I accounted for everything else that normally explains attendance swings. It stayed.
Why this mattered to me
I went into this ready to write the "it's more complicated than the headlines say" version of this project, because that's usually the more responsible take in sports analytics. In this case, the headlines were right. Clark's individual impact on WNBA attendance is real, it's measurable, and it holds up once you actually run the numbers instead of just eyeballing the crowd.