Methodology behind USA TODAY's analysis of college athletic rosters

·8 min read

In a comprehensive data analysis, USA TODAY found widespread use of roster manipulation across many of the nation’s largest and best-known colleges and universities.

USA TODAY downloaded the athletic participation numbers schools report annually to the U.S. Department of Education under the Equity in Athletics Disclosure Act from the department’s website and compared them against four different data sources:

• NCAA revenue-and-expense reports that break down expenses and athletic participation by team and gender;

• team rosters on athletic department websites;

• internal athletic department rosters called "squad lists;"

• and reams of competition results.

The news organization’s analysis centered on 107 public schools in the Football Bowl Subdivision – the highest level in Division I – during the 2018-19 school year, the last full year before the pandemic upended the college sports landscape. USA TODAY filed hundreds of public records requests for the schools’ squad lists and NCAA reports, wrote computer programs to collect online rosters and stats, and interviewed 51 Title IX experts and attorneys, lawmakers, athletes, coaches and athletic department administrators.

USA TODAY’s analysis found the vast majority of those schools counted athletes in a way that inflated their reported numbers of women participants. Sixty-six did so by at least 20 athletes. Altogether, the schools added more than 3,600 participation “opportunities” for female athletes without adding a single new women’s team.

The graphics featured in USA TODAY's series of stories on roster manipulation display the data for 86 schools that added a total of at least eight women's roster spots through any combination of the three roster manipulation methods described below. We chose eight roster spots as the threshold to display the data for a school because eight women is sufficient to sustain a viable, NCAA-recognized women's team in an emerging sport, such as triathlon or bowling.

Data was not displayed in the charts for six schools that added a total of seven roster spots or fewer through any combination of the three methods. For those schools, the graphics show "no noted roster manipulation" or zero roster spots added.

Duplicate counting

To calculate the number of women's roster spots each school gained through double- and triple-counting, USA TODAY compared the schools' "duplicated" counts of male and female participants with their "unduplicated" counts, both of which it reports to the U.S. Department of Education and NCAA.

Whereas the "duplicated" counts represent the total numbers of roster spots the athletes filled, the "unduplicated" counts are raw head counts. USA TODAY subtracted each school's unduplicated counts from their duplicated counts to find the number of total male and female roster spots gained through duplicate counting. It then subtracted the number of men's roster spots gained from the number of women's roster spots gained to determine the total number of women's roster spots netted.

Of the 107 schools analyzed, 75 showed a net increase of women's roster spots through duplicate counting. They collectively added 2,252 women's roster spots through double- and triple-counting. Schools that showed no net gain for women net or a net gain for men are not displayed in the graphics on duplicate counting.

Unnecessary rowers

To calculate the number of women who filled unnecessary roster spots on women's rowing teams, USA TODAY compared the numbers of female rowers the schools reported to the U.S. Department of Education with the maximum number of female rowers that schools' respective conference championships accommodate, plus extra for substitutes/reserves.

Whereas NCAA championships accommodate up to 23 women – two eight-person boats and one four-person boat, each with a coxswain – conference championships often allow for additional boats. The Atlantic Coast Conference, for example, permits a third eight-person boat and a second four-person boat, plus a coxswain for each, for up to 37 women. Big Ten Conference championships allow up to 51 women, the most of any conference.

To determine the maximum number of reserves a women’s rowing team needs, USA TODAY relied on benchmarks set in recent Title IX lawsuits against the University of Iowa and University of Connecticut. In settlements of the lawsuits, each school agreed to cap its women's rowing roster. Iowa, which competes in the Big Ten, capped its roster at 75 – 47% more than Big Ten championships allow. Connecticut, which in 2018-19 competed in the ACC, capped its roster at 45 – 41% more than ACC conference championships allow.

Using these caps as a guide to determine the maximum number of rowers a school needs on its roster, USA TODAY set a benchmark of the maximum number its conference championships permit, plus 50% of that for reserves. For schools with odd numbers necessary for conference championships, we rounded down to the nearest whole number. For instance, the maximum number ACC schools would need on its roster is 55 (37 + 18.5, rounded down to 55 stay under the 50% cap).

Based on those benchmarks, 27 of the 30 schools that sponsored women's rowing teams rostered more athletes than necessary. Of the 2,347 total female rowers reported, the analysis found, 838 of them filled unnecessary roster spots.

The University of Wisconsin competes in the Big Ten, which accommodates up to 76 rowers. But because it also sponsors a varsity lightweight team, in addition to its openweight team, USA TODAY calculated it can accommodate up to 100. The lightweight team competes in the Intercollegiate Rowing Association championships, which allows up to 16 women in lightweight boats. Using the 50% benchmark, eight more could serve as lightweight reserves.

Male practice players

To determine the number of male practice players schools counted toward their women's basketball teams, USA TODAY compared the per-team participant counts they reported to the NCAA with those they reported to the U.S. Department of Education.

The U.S. Department of Education specifically instructs schools to count male practice players as participants on women’s teams.

When the department was asked why it instructs schools to report numbers this way, a spokesperson pointed back to the regulations, which do not provide an explanation.

The NCAA, however, instructs schools not to count male practice players as female participants. The existence of male practice players in the U.S. Department of Education data is the only significant distinction between two datasets and the reporting instructions.

USA TODAY flagged likely practice players by subtracting each school's NCAA numbers from its U.S. Department of Education numbers. Fifty-two of the 107 schools reported apparent practice players on women's basketball teams, the analysis found. Some schools appeared to count male practice players toward lacrosse, soccer and volleyball teams. A few schools also appeared to count practice players toward men's football and baseball teams. Women’s basketball, however, was the only sport where practice players consistently appeared.

When schools report numbers to the U.S. Department of Education, they don't include the names of the players they are counting. They simply provide the total of the number of participants on a given team's roster. It is therefore possible some of the practice players flagged in the analysis matched the gender of their teams toward which they were counted. For instance, the men counted as practice players on football teams were likely, indeed, men. Nevertheless, 90% of the apparent practice players flagged in the analysis were on women’s teams.

After flagging likely practice players, USA TODAY cross-referenced the reported numbers with the names that appear on the schools’ internal squad lists, which the news organization obtained through public records requests. These squad lists often, though not always, list the names of male practice players. Google, LexisNexis and social media searches showed many of the people identified as, or appeared to be, men. Many of the names in the squad lists were common male names, such as Michael, Brandon, and Robert.

Eight schools erroneously included male practice players in their counts to the NCAA, USA TODAY confirmed by contacting officials at those schools, who described their overcounts as clerical errors. Because subtracting the NCAA numbers from the U.S. Department of Education numbers was not an option for these schools, USA TODAY calculated the count of practice players in one of two ways: 1) counting the names of apparent men on the women's team squad list, or 2) subtracting the number that appeared on their online roster plus three from the number it reported to the federal government. The extra three account for the fact that online rosters are snapshots in time and do not always represent a team’s full squad for the whole year.

In total, USA TODAY’s analysis found at least 601 of the 1,317 women's basketball participants reported by the 52 schools were likely male practice players. They represented more than a quarter of the 2,131 total women's basketball participants reported across all 107 schools in the analysis.

This article originally appeared on USA TODAY: Methodology behind USA TODAY's analysis of college athletic rosters

Our goal is to create a safe and engaging place for users to connect over interests and passions. In order to improve our community experience, we are temporarily suspending article commenting