Introduction:
As I’ve grown up, I’ve witnessed the palpable rise in popularity of women’s basketball, soccer, and hockey—especially on the collegiate level. In tandem with this rise, my mother, the most influential figure in my life, has encouraged me to document and engage with it. Last year, Caitlyn Clark, Angel Reese, and others had an immense effect on the attendance and viewership of the Women’s NCAA March Madness tournament. Last summer, I attended the Wharton Moneyball Summer program, where I engaged with sports analytics. We looked through studies across sports: diving, baseball, hockey, fencing, and more. I conducted a final project: Will Ovechkin break the goal record? I left Wharton hungry to complete more studies in sports analytics. I decided to delve into the raw attendance numbers to document the growth in Women’s collegiate sports. I know social media has been giving increased attention to these stars. For example, players like Caitlyn Clark and Paige Bueckers have blown up on TikTok. Naturally, these should increase attendance, so I wanted to see numerically how much growth there would be. Alongside this study, I am conducting an interview series for women athletes across all sports to spotlight their lived experiences and generate opportunities and positive discourse for women in sports. Numbers can’t tell the whole story, so I also want to capture the emotion around the rise. Women are still paid 83 cents on the dollar (National Women's Law Center). But I believe that the rise of women’s sports is vital as it can be a catalyst for equal pay across all professions. Sports can spearhead positive discourse around gender equality that can spread into other facets of society. They can be that first domino to fall and start the movement toward equity in all sectors of life.
Research Questions:
After my reading, I decided to focus on the following questions:
-
Has there been a general increase in attendance at women’s college basketball games over the last 10 years?
-
How does the increase vary between colleges from different conferences, specifically from colleges from “big” and “small” conferences?
-
What are some potential factors for this increase or decrease?
Data:
For my analysis, I chose 19 of the 350 Division 1 women’s basketball schools. I used convenience sampling; whatever schools had readily available data, I added to the study. I wanted to get a holistic view across different-sized colleges. So, I chose a couple of schools from the big conferences (SEC, BIG 10, BIG 12), some mid-majors, and some smaller conferences. I separated the schools into Big and Mid-Major, grouping the small schools into the Mid-Major category. The final list of schools. Big: UConn, Iowa State, University of Tennessee, Iowa, Auburn, Texas, Texas A&M, Oklahoma, Illinois, and Penn State. Mid Major: South Dakota State, Dayton, Akron, Colorado State, Bowling Green, Furman, and Stony Brook. And finally, small: Loyola Maryland and Penn. I think these schools are a good full representation of big, small, and medium-sized colleges.
One of the biggest challenges was the lack of centralized data: there was no single PDF containing each school’s data, so I had to collect my game-by-game data by searching for the given college’s schedule from the years (2013-14 or 2023-24) and then visiting the box score for each and every home game before noting the attendance. For a sport like the MLB, there are datasets containing all the attendance numbers in one spreadsheet. This wasn’t available in my case. I had a Numbers document, and when I chose to add a new school, I would add 15 slots for 2013-14 and 15 slots for 2023-24, and then just add the attendance from each box score into the data set. So I ended up with a sample of 15 home games for each school (usually that would capture all the games). I didn’t use any data from away games, as the attendance number depends too much on the opponent, causing variance. My original plan was to also include 2003-04 data to see a 20-year progression; however, that data was not publicly available.
Mean vs. Median:
Figure 1: Texas game-by-game attendance with the median and mean plotted for 2023-24 and 2013-14.
Figure 1 shows what typical game-by-game data looks like for a school, and shows how I decided to measure the center or summary of each school. A mean is the mathematical average of the data, and a median is the middle number of the data (when arranged lowest to highest). The red diamond shows the mean, and the blue the median. Means are skewed by extreme values, causing them to be further away from the center. The median is untouched by these values. So, the median is the better measure of the center and will be used for the study.
Data Analysis:
The graphs of the data show that ten of the nineteen schools, the University of Iowa, Bowling Green, South Dakota State, University of Connecticut, Akron, Illinois, Colorado State, Auburn, Texas, and Stony Brook, all had rises in attendance. Five of the schools: Iowa State, Penn, Dayton, Loyola Maryland, and Furman all stayed relatively constant. And the last 4: Tennessee, Penn State, Texas A&M, and Oklahoma’s attendances all dropped.
Figure 2: Game-by-game data from 2013-14 and 2023-24 for South Dakota State and Texas with the medians connected.
Figure 2 graphs the attendance for Texas and South Dakota State. An interesting visualization actually showed how the games were scattered among some of the schools, one mid-major and one big school. The medians are connected for each school to see a summary of that change. Clearly, in this case, Texas had exhibited greater growth than South Dakota State. The following graph will show the percent change of all the medians.
Figure 3: All the colleges and their median percent changes over the last 10 years.
Figure 3 shows the percent changes of the medians from 2013-14 to 2023-24 for all the schools. The percent change was computed by subtracting the median from 2013-14 from the median from 2023-24, dividing by the 2023-24 value, and multiplying by 100.
Risers are the schools that gained attendance; Constants are the schools that stayed constant; and the Dips are the schools that lost attendance.
Smaller/Mid-Major schools:
-
Risers:
It seems the mid-major schools were the most consistent risers: South Dakota State, Akron, StonyBrook, Colorado State, and Bowling Green all convincingly gained attendance. The increased national attendance towards women’s basketball benefited these schools as they didn’t have the national base that bigger schools already contained.
-
Constants:
The other two mid-majors—Dayton and Furman—-stayed pretty constant. Penn was a small school that also didn’t see growth. And, Penn didn’t have any breakout success; however, it exhibited moderate median and lower quartile growth in attendance, mirroring larger national trends.
-
The Dips:
Loyola Maryland’s squad lacked any major marketing push and supported a weak 9-21 record, contributing to the only dip in small or mid-major schools’ attendance.
Bigger Schools:
-
Risers:
UConn, Texas, and Iowa have all emerged as powerhouses in women’s basketball. Iowa and UConn both have immensely popular superstars, Caitlyn Clark and Paige Buekers, respectively, contributing to their rise. Auburn had a similar record, so I believe their rise may be attributed to simply the increasing popularity of Women’s Basketball. Lastly, Illinois’ attendance rose as the team improved immensely in the early 2020s under the leadership of coach Shauna Greene.
-
Constants:
Iowa State stayed constant largely due to a stable, well-respected coach and no major-star-induced spikes. They have consistently been at the top of women’s basketball attendance, so their supporter base has been steadily present, rather than growing from national attention.
-
The Dips:
Tennessee, Penn State, Texas A&M, and Oklahoma. In large part, the dips in attendance probably came from dips in performance. Tennessee was one of the best teams in the NCAAW during the 2013-14 season, going 29-6 and having a one seed in March Madness. In 2023-24, however, they were a mere 20-13 and a six seed in March Madness. For Penn State, they were 25-8 throughout the 2013-14 season and a three seed in March Madness. Whereas in 2023-24, again a much worse 22-13, and did not qualify for March Madness. Similarly, Texas A&M was 27-9 and a three seed in March Madness, before dropping off to a 19-13 record and an 11 seed in the 2023-24 season. Lastly, Oklahoma’s attendance drop can be attributed to the retirement of longtime head coach Sherri Coale.
Change Analysis:
Figure 4: Percent change in median over the last 10 years for big and mid-major schools.
Figure 4 is a more careful comparison of the differences in changes between mid-major and big schools. Figure 4 is a graph of the change in median with mid-majors separated from big schools. It shows that the Small and medium-sized schools have had their attendances grow more consistently in the last ten years than big schools (only one lost more than a marginal amount, and six grew. The majority of schools grew, and the ones that didn’t only had slight drop-offs. Bigger schools, with the exception of Iowa and Illinois, seem spread evenly around that zero line. To me, this shows that bigger schools had their supporter base, and increased attention to women’s sports wouldn’t cause growth; however, smaller schools have benefitted immensely, as they previously lacked that fanbase.
Table 1: Summaries of the changes in the 10-year period for the two groups of schools. N being the total number of schools. P being the total number that grew. PCT_PLS being the percent that grew. Ndchange being the median percent change.
The raw numbers strengthen this claim: 66.7 % of mid-majors grew with a median growth of 16.3 percent. And, 50 % of Big schools grew with a median growth of 12 percent.
Wrap-up:
Overall, this project has deepened my interest in sports analytics, and I hope to continue using data to document change. The data shows that mid-majors rose most consistently, showing that the popularity of women’s basketball has penetrated through the outer shell of top schools and into the middle. A lot of potential factors correlate with this rise in attendance. Star players correlate to attendance success, i.e, Caitlyn Clark and Paige Bueckers. There has been significant institutional investment in the sport, creating a better product. The explosion of WNBA popularity and expanded broadcasting rights all potentially contribute to the rise in attendance. In the future, I hope to look deeper into the factors of the rise and take further actionable steps to continue the drive upward.
To critique my study, I’d say the most challenging part was actually getting the data. It took me weeks and weeks of going through box scores and different teams to actually compile all the data. Once it was there, I enjoyed playing around with graphs in R, seeing which graphs are more telling. If I did this again, I would probably just try to incorporate more schools to expand the study. But, overall, I enjoyed it, and I think the product is telling about that upward trajectory in women’s sports.