Skip to main content
ABC News
The Raiders Had More Black Players Than Any NFL Team. Then Jon Gruden Took Over The Roster.

Years before Jon Gruden resigned in disgrace as the Las Vegas Raiders’ head coach over a series of racist, misogynistic and homophobic emails, Chip Kelly faced scrutiny over his record on race during his tenure as head coach of the Philadelphia Eagles. 

Running back LeSean McCoy, who was traded from the Eagles to the Buffalo Bills in 2015, told ESPN The Magazine later that year that Kelly was eager to jettison the team’s “good Black players,” adding, “there’s a reason he got rid of all the Black players — the good ones — like that.” McCoy wasn’t the only one to question the racial motives behind Kelly’s roster moves: ESPN’s Stephen A. Smith said after McCoy was traded that Kelly’s personnel decisions “leave a few brothers feeling uncomfortable.” Tra Thomas, a former offensive lineman and assistant coach for the Eagles,1 voiced those same concerns when he asserted that a number of Philadelphia’s players thought there was a “hint of racism” in the locker room under Kelly’s leadership.

The accusations against Kelly — now the coach at UCLA — prompted my own statistical investigation into how race might matter in NFL roster decisions. My analysis of data collected on each player’s racial background from Best Tickets’ Unofficial 2014 NFL Player Censushere. Team rosters list anywhere from 52 to 55 players, so it’s likely limited to active players only.

">2 found that the 10 teams in 2014 who had Black people in the key leadership roles of head coach and/or general manager had significantly more Black players on their rosters than the 22 other NFL teams. No team did more to drive that year’s statistically significant negative correlation3 between whiter team leadership and having fewer Black players on NFL rosters than Kelly’s Eagles. In fact, the significant differences4 between the percentage of Black players on the Eagles (50.9 percent) and the rest of the NFL (68.3 percent) were beyond the statistical threshold that the courts and federal bureaucracy generally recognize as potential discrimination.

Denver Broncos and Baltimore Ravens during week 4

Related: Our 2021 NFL Predictions

Meanwhile, the team most responsible for driving the positive correlation between African American general managers having more Black players on their rosters in 2014 was none other than the Raiders. Under the leadership of the team’s African American general manager, Reggie McKenzie, the Raiders (then playing in Oakland) had a higher share of Black players on their roster (79.2 percent) than any other NFL team in 2014. According to data compiled on the racial composition of each NFL team’s roster by ProFootballLogic,5 the Raiders also had the NFL’s highest percentage of Black players (82.3 percent) in 2016 — the year that McKenzie won executive of the year honors after the team’s impressive 12-win showing. It’s probably not a coincidence, either, that the two teams with the next highest shares of Black players, the Giants and Bills, also had African American GMs. Indeed, the five NFL teams with Black GMs in 2016 had rosters that were, on average, 75.4 percent Black, compared with 67.7 percent for the 27 teams that did not — a statistically significant difference6 in percent of Black players that we can be confident was not simply due to random variation.

The Raiders’ racial composition was virtually identical in 2017, the year before Gruden began his second stint as the team’s head coach. While there’s no publicly available data on the racial composition of NFL rosters after 2016, my admittedly crude codingRaiders’ rosters from and then imputing each player’s race from pictures.

">7 of the team’s roster once again found that 82.0 percent of the Raiders’ players were Black in 2017. But the number of Black players on the Raiders sharply declined soon after Gruden became the Raiders’ “de facto football czar.” By the end of the 2018 season, McKenzie had been fired, and Gruden assumed even more control over the Raiders’ personnel decisions. That included changing the roster’s racial composition: My analysis of the team’s rosters found that the share of Black players on the Raiders declined from 82.0 percent in 2017 to 69.0 percent in 2019 and 67.1 percent in 2020 and 67.2 percent in 2021.

To be sure, those notable differences say nothing about the coach’s racial motives in constructing his roster. Some of the changes to the roster’s demographics, after all, could simply be regression to the mean percentage of Black players in the NFL. But, as former NFL receiver Keyshawn Johnson critically said of his former head coach on his ESPN Radio show with Jay Williams and Max Kellerman, “A team is put together by the makeup of a coach. And a coach wants a certain personality.” It’s pretty clear from his emails the type of personality that Gruden preferred on his team; and it’s easy to look back in hindsight with that knowledge and find a persistent pattern of anti-Black bias in the coach’s decisions.

Regardless of his reasons, though, Gruden’s email scandal is yet another reminder of the importance of diversity in leadership positions in the NFL. The 2014 and 2016 data on racial compositions of NFL rosters suggests that even in professional football — an ostensible meritocracy where performance is much more objective and transparent than it is in other professions — race is still a significant factor in personnel decisions. The power imbalance in a league composed primarily of Black players but largely run by white head coaches and general managers means there will always be the potential for implicit and explicit racial biases to affect roster moves.

What should be the next Summer Olympic sport? | FiveThirtyEight Debate Club


  1. Thomas did not play under Kelly but did coach under him.

  2. The content on has since been removed. But the Unofficial 2014 NFL Census data and my analysis can be found here. Team rosters list anywhere from 52 to 55 players, so it’s likely limited to active players only.

  3. p = .04

  4. p < .01

  5. ProFootballLogic imputed race from player photos, names and other public information. It lists anywhere from 62 to 78 players per team, so its rosters likely include practice squad or other non-active players who were on the team’s roster during the season.

  6. p = .02

  7. My coding began by first downloading spreadsheets of recent Raiders’ rosters from and then imputing each player’s race from pictures.

Michael Tesler is a professor of political science at University of California, Irvine, author of “Post-Racial or Most-Racial? Race and Politics in the Obama Era” and co-author of “Identity Crisis: The 2016 Presidential Campaign and the Battle for the Meaning of America.”