Discussions of race in college sports often rely on faulty statistics. One recent, high-profile example comes from The Atlantic, where sports columnist Jemele Hill offered an intriguing thought experiment involving black student-athletes and historically black colleges and universities (HBCUs). Unfortunately, Hill fell victim to a commonly misunderstood statistic about black, male college athletes: their supposed overrepresentation on university sports rosters at “predominantly white schools.”

To be clear, Hill did not invent these numbers—they match statistics in a media-bombshell of a report on race in college sports, authored by Professor Shaun Harper of USC (“Black Male Student-Athletes and Racial Inequities in NCAA Division-I College Sports,” or “the report”). Because Hill herself would undoubtedly welcome digging into the numbers—she used to co-host a show called Numbers Never Lie on ESPN—let’s dig in: are black athletes overrepresented in college sports?

The common claim goes something like this: black men make up large percentages of major universities’ football and basketball teams (55-56%), but very little of the student bodies (2-3%). This overrepresentation may seem striking, but the numbers have problems, and are at best incomplete.

First, the low number. Hill parrots the report’s claim that, on average, black men make up about 2.4% of students at universities that participate in big-time athletics (or, the universities that are part of the “Power-5” athletic conferences). This number is not incorrect, but it is misleading when compared to the percentage of black men on a men’s basketball roster. Specifically, the number is artificially low because it ignores sex.

The percentage of black men on a men’s basketball roster will always exceed the percentage of black men on a college campus, because the campus also includes women. Even if every man on campus were black, then black men would compose 100% of male athletes, but they would compose only about 50% of students, because the other 50% would be female. Presenting the numbers this way underreports the percentage of black, male students by a factor of two.

A more relevant comparison would be the number of black, male students as a percentage of all male students, not as a percentage of all students generally. Unsurprisingly, this number is about double the previous number, or about 5%. Also unsurprisingly, this number is similar to the percentage of black students, both male and female, on campus.

So, 5% of students at Power-5 schools are black. This number seems low, and it is: about one in eight people in this country, or about 13%, are black. This is the real disparity.

That black people are underrepresented on campus may indeed be due to the fact that many black students face significant hurdles in their educations. Perhaps as an acknowledgment of these hurdles, many people demand to increase representation of black people on campus, as the report does. These demands are seemingly based on the assumption that the number of black students is too low because the number of white students is too high. Of course, compared to the number of black students, the number of white students on these campuses does seem high—in her Atlantic piece, Hill routinely refers to major universities as “predominantly white” schools. But “predominantly” does not translate to “overly.” Yes, there may be too few black students on campus, but it does not follow that white students are necessarily to blame.

According to the same statistics that were used in the report, schools that play top-level athletics are about 67% white, while the most recent census shows that the U.S. is about 64% white. These numbers are statistically indistinguishable, meaning that white people are not, statistically speaking, significantly overrepresented on campus. However, the NCAA’s race category of “other” (i.e. “other” than white, black, or Hispanic), which includes people of Asian descent, is greatly overrepresented on campus. People in this “other” category compose about 8% of the American public, but nearly 16% of students on campuses of Power-5 schools.

Of course, a hypothetical increase in the percentage of one racial group would have to correspond to a decrease of another. To demand “fair” representation of racial groups on college campuses, and to define “fair” as matching the frequency of each race in the population, would result in significant decreases in the numbers not of white students, but of students who are of other minority races, especially people of Asian descent. Even an extreme case of increasing the numbers of black students via hypothetical, race-based quotas, for example, would probably exclude other students of color, even at “predominantly white” institutions.

The other, higher number cited by Hill and the report is the percentage of high-revenue athletes who are black. To be sure, this number is less important than the lower number—access to a college degree is certainly of greater concern than representation on a sports roster. However, this higher number is also problematic.

When the latest version of the report was published, rosters of high-revenue sports at Power-5 schools were about 55% percent black. As we have seen, black people make up only about 5% of students on Power-5 campuses, and only about 13% of people in the population. Compared to these metrics, 55% is indeed a high number. But is it too high?

Interestingly, the report itself explicitly says that it is not (p.6). In fact, one might argue that the number is actually too low. The professional leagues of the two sports in question, the NFL and NBA, have higher percentages of black players than the NCAA. While rosters of football and men’s-basketball teams in Power-5 conferences of the NCAA were 55% and 56% black, respectively, 68% and 74% of players in the NFL and NBA were black. If we suppose that professional teams are interested only in fielding the best players, regardless of race, we might conclude not only that black men are most likely to be successful football and basketball players, but also that they are therefore underrepresented in college athletics. This suggests that the “high” number that Hill references is actually too low.

So, we are presented a disparity between a low number and a high number, and that disparity is deemed problematic. Upon examining the numbers, though, we realize that the low number, even if it had been misreported as too low, still might not increase by much, even in the name of fair representation (i.e. black people compose 5% of students compared to 13% of the population). Also, the high number might itself further increase (i.e. the 55% of black athletes in college could jump to about 70%, the number we see in the professional leagues). In other words, aiming for “accurate” representation of black athletes and students on campus might result in divergence, not convergence, of the two numbers in question. It’s exactly backwards.

Considered this way, the difference between the percentages of black students and black athletes does not represent the real problem. It is a red herring that largely owes to the high number of black athletes. Nonetheless, the disparity seems to feature prominently in our national discussions of race in college sports.

Apparently, this supposed overrepresentation is a core argument for the plight of black, male college athletes. Another core argument is their academic record, which will be the topic of Part 2 and Part 3 of this series.

P. A. Jensen (@RuralityChecker), a former professor and college administrator, is editor of RuralityCheck.com.