Racial Discrepancies in FBS Coaching Are a Fact
The intersection of college football and race has ebbed and flowed in the public consciousness for more than a hundred years. From issues of team integration to paying student athletes to demographics of the coaching community, race is never far afield from the preeminent issues that permeate the college football landscape. We would like to present a data-driven look at just how inequitable the current coaching racial distribution is and provide some insight into how this can change.
Off the bat, we know that some of our readers might respond with apathy or even disgust at the very mention of this topic. Isn’t it divisive to bring up race, especially in a context as diverse as that of college football? And to that we would say, it has been racial divisiveness that led to the current landscape, and you don’t fix a problem by ignoring it. We are talking here about positions that can confer generational wealth, or at least put one in line for that kind of position. When the wealth gap in this country persists at astronomical levels between the median Black and White family, it matters greatly if there are racialized discrepancies in whom gets major college football coaching jobs. And we would like to do our part in fighting those racialized discrepancies.
We went to work on this project with a few goals in mind: to use data to analyze just how substantive the problem is, to see if the data could point us toward the root causes of the problem, and to provide some ways forward toward more just outcomes in the college football coach hiring process. We want those in power to know the reality of the situation so that they can change it.
To start, we were privileged to receive a dataset of FBS coaching staffs (including 2000-present) from Geoff Hutchinson. We manually tagged the race of all of the coaches who had served as head coaches or coordinators during this time period. We were also interested in the pool of people who get FBS coaching jobs, so we manually tagged the Head Coaches based on whether or not they played college football. We could then do some analysis to examine the performance of head coaches, offensive coordinators, and defensive coordinators by race over this time period.
What did we learn?
- The racial imbalance in college football coaching is real and persistent.
- This imbalance is not based on differences in coaching performance.
- This imbalance is largely based on coaches’ reliance on their social networks for hiring.
- This imbalance can be mitigated with intentionality from the more influential people in the sport.
Coaching Hires Do Not Reflect the Coaching Candidate Pool
As Ivan Maisel detailed late in 2020, the lack of Black coaches at the highest levels of college football may be the thing that has changed the least about the game in the preceding few decades. Recently, Chris Vannini added another piece of journalistic ammunition aimed at bringing change in this domain.
As we seek to extend this body of work, we will begin by looking at the racial makeup of coaching staffs compared to rosters. The NCAA has kept a database of the self-reported race of all college athletes since at least 2000, so we were able to use that data to show the racial makeup of the college football player pool.
How does that compare to the coaching side of the sport?
Looking back over the past 20 years, the percentage of Black players in college football has continued to grow steadily, while the percentage of Black coaches has failed to keep pace, much less approach the racial distribution of the players themselves. During the mid 2000s, there was an uptick in Black coaches being hired at the highest level. Since then, we have seen stagnation, as there has been only a two percent increase in Black coaches over the past decade.
Why should we compare the percentage of Black players to Black coaches? For one, representation matters in any field, but beyond that, looking at every head coach in the FBS since 2000 reveals that 92.7% of them played college football. That is, the college football player pool provides a substantial majority of the FBS coaching pool. For the above plot to show progress towards more equitable hiring practices, the two lines should be converging. Instead, they maintain a relatively consistent difference of over 25%.
Encompassing all levels of college football, Black players make up 39% of players. Given that 92.7% of FBS coaches come from this pool and given the demographic makeup of the US population to represent the other 7.3%, we can formulate the expected distribution of the racial makeup of college football coaches and compare that to the reality.
It’s a stunning difference. Instead of the expected 37%, Black coaches make up slightly more than 10% of head coaches and offensive coordinators at the FBS level. Although the racial distribution improves to 18% for defensive coordinators, there are still half as many Black defensive coordinators as we would expect to see. For a thorough examination of why there are more Black coaches on the defensive side of the ball, look no further than Richard Johnson’s excellent work from 2017 which describes, among other things, how Black players are funneled into positions thought to require less cognitive and leadership abilities.
As we have shown, the racial distribution shows a weighty imbalance at each of the three most significant coaching positions. Just how much change would be necessary for the actual distribution to match what we would expect?
The actual number of Black head coaches, offensive coordinators, and defensive coordinators is between 24 and 35 fewer than what we expect if those distributions matched the pool of people who receive these jobs.
One of the contributing factors to this imbalance that we found is that Black coaches simply have shorter tenures in their positions.
Average Coaching Tenure
|Average Tenure||Head Coaches||Coordinators|
|Average Tenure||Head Coaches||Coordinators|
|Black||3.4 years||2.3 years|
|White||4.9 years||2.7 years|
Despite a surge of Diversity, Equity, and Inclusion pledges from around the NCAA and an increasing number of influential members of the CFB media who draw attention to these discrepancies, we are still seeing the same trends proliferate. Black coaches get fewer top jobs and less time to do them.
There could be three primary causes for this tenure discrepancy:
- Black coaches could be promoted sooner.
- Black coaches could be justifiably fired sooner due to poor performance.
- Black coaches could be unjustifiably fired sooner.
The first cause would result in shorter tenures for coordinators and longer tenures for head coaches. However, the average discrepancy in the table above is even larger among head coaches than it is among coordinators. Ruling out the first option, we will turn to an analysis of coaching performance to see if this tenure discrepancy is justifiable or not based on on-field results.
Coaching Performance Does Not Justify Discrepancy in Hiring Practices
The thorny justification for the discrepancy we outlined above would be that Black coaches simply perform worse than their White peers. Is that what is happening here?
We built a dataset that includes advanced stats by team for every season starting in 2005. We matched that data to our database of coaches and were able to compare data during a coach’s tenure with data from the preceding years to provide a baseline impact on team performance for each coach. We could then calculate the net impact of a coach’s tenure: overall net change in team performance for head coaches and net change on a particular side of the ball for coordinators.
We also looked specifically at coordinators who were hired to be head coaches, as we might expect differences in performance as coordinators to significantly impact who is hired as head coaches.
Overall, we looked at seven specific ways of grouping and evaluating the performance of the coaches:
- Net change in team performance for each head coach, grouped by the race of the head coach
- Net change in offensive performance for each offensive coordinator, grouped by the race of the offensive coordinator
- Net change in defensive performance for each defensive coordinator, grouped by the race of the defensive coordinator
- Net change in defensive performance prior to getting an HC job for defensive coordinators who became head coaches, grouped by race
- Net change in offensive performance prior to getting an HC job for offensive coordinators who became head coaches, grouped by race
- Net change in team performance after getting an HC job for defensive coordinators who became head coaches, grouped by race
- Net change in team performance after getting an HC job for offensive coordinators who became head coaches, grouped by race
Below, you can see a representative graph that includes more than 200 head coaches and their impact on the team’s net success rate, as well as PPA. PPA is an expected points metric developed by www.collegefootballdata.com that gives us a very predictive measure of a team’s overall ability.
White coaches simply aren’t better than Black coaches at improving team performance. A simple regression using race as the predictor variable and Net PPA as the outcome variable yields p-values around 0.3, far outside of the statistically significant range, and has an R-Squared value of 0.0093, meaning that race accounts for less than 1% of the variance that we see in head-coaching performance. Focusing in on defensive coordinators who became head coaches, we have similar p-values and an R-squared value of 0.013 when using race as the predictor and Net Defensive PPA as the outcome variable. The other results we found were similar; we won’t show them all here to avoid redundancy.
In all seven cases, the data shows that there is no statistical significance for race in determining improvement of on-field performance. What does that tell us? Performance is not the underlying reason for the racial discrepancy of coaching hires and tenure lengths at the FBS level. Black coaches are being fired unjustifiably sooner than their White counterparts.
Athletic directors, current head coaches, and other administrators and boosters have strong influence in the hiring practices for the most significant coaching jobs within the FBS. If these groups of people were using the coaching pool we identified above and simply making performance-based hires, we would see a drastic change in the racial distribution of current coaches. Instead, we see a perpetuation of the racial imbalance that has long plagued the ranks of college football coaching.
What else might be going on here? Perhaps head coaches and athletic directors tend to hire people they know or have personal connections with.
The Coaching Social Network Is Strong
If power brokers are not hiring a representative distribution of former players but also aren’t hiring better-performing White coaches, what drives the ongoing imbalance of hiring practices? The Coaching Social Network. This network is intricately tied to the inequitable racial distribution of college football coaches.
The racial imbalance in college coaching intensifies when moving from the assistant to coordinator level. These networks help us explain who it is that actually gets to make that jump. Using a Louvian analysis to group interconnected communities in the college football coaching fraternity, we can visualize how the networks operate. This analysis dives into the working relationships between head coaches and coordinators for the past 22 years. Below is an example of the kind of network developed by a longstanding head coach.
The Bill Snyder coaching network is mostly White, but there’s one distinctive branch. Turner Gill is three connections removed from Snyder, but once he got his first heading coach job, the racial distribution of the overall network began to change considerably. Gill himself is responsible for 6 of the 12 Black coordinators hired in the entire Bill Snyder coaching community.
It’s evident visually that the race of the head coach impacts the race of whom gets hired as coordinators. Our research shows that Non-White head coaches employ Non-White coordinators 30% more frequently than do their White counterparts. We posit that a coach’s race largely determines his social network within coaching circles, and that social network largely determines who receives coordinator and head coaching jobs.
Below you can see which coaches have the best and worst track records of hiring Non-White coordinators, starting from the 2000 season.
All 20 of the coaches who hired the fewest minority coordinators are White. Without intentionally changing, we expect these same patterns to continue. But the current state of things does not have to be fait accompli.
The network analysis can be extended to help us understand the coaches who have been best positioned to change the landscape. We can measure the number of direct connections for every head coach in our dataset, and we can measure their influence over the coaching landscape.
Connections measure the number of coordinators the coach has employed or the number of times they have been employed by someone else on the list. Influence (also known as Closeness Centrality in the Louvain method) measures how close a coach’s network is to all of the other networks in the community. We will take a look at the 20 most connected and 20 most influential coaches from our dataset to see how they have done at hiring Non-White Coordinators.
There’s a couple of fascinating insights we can glean here. 18 of the 20 most connected coaches are White; the 19 most influential coaches are White. Only 4 of the 20 most connected and 5 of the 20 most influential coaches (4 of whom aren’t actively coaching anymore) have been above average in their rate of hiring Non-White coordinators. On average, White coaches are 43% more connected than Black coaches.
Here we can begin to see why the racial imbalance in college coaching persists. College coach hiring practices are fundamentally intertwined into a social network; the most connected and influential coaches are mostly White and mostly hire White coordinators. Black coaches get less time as head coaches, so they have less opportunity to make hires. The cycle perpetuates itself because those with the most ability to do something differently in their hiring practices largely have not.
We can think of at least four different reasons why these social networks continue to operate as they do: random social connections, bonds created by shared affinities, implicit bias, and explicit bias. It is beyond our scope of knowledge to try and identify which of these factors is operating in any given network. We do believe it is wise for the reader and for decision makers to consider how each of these factors may be at play and to consider appropriate measures of remediation.
We sincerely hope to see these hiring patterns change. The connected and influential coaches who are still active - coaches like Nick Saban, Gus Malzahn, Mack Brown, Mike Gundy, Bret Bielema, Kyle Whittingham, Lane Kiffin, and Brian Kelly – are the ones best positioned to help make the racial distribution of college coaching more equitable. Those who look to be mainstays in the upper echelons of college football coaching - Dabo Swinney, Kirby Smart, Lincoln Riley, Luke Fickell, Ryan Day, Dave Aranda - are all positioned to bring change for years to come.
It won’t happen by continuing to dial the Favorites on their contact lists, but it’s quite possible to broaden their horizons and identify excellent Black coaches who simply haven’t gotten a shot at the top jobs. Hire Black coordinators this year, and it won’t be long until the head coach distribution significantly changes as well.
We reached out to representatives from the 14 Athletic Departments where the above-listed coaches are employed to share our findings and ask for comment. As of publishing, Illinois, Oklahoma State, Utah, and Ole Miss had declined to comment, and we did not receive a response from the other 10.
Top Black Head Coaching Prospects
To close, we want to highlight five coordinators who should be strongly considered for FBS head coaching jobs in the coming cycle. These would be five more Black head coaches and five more coaches who are more likely to hire Non-White coordinators. Beyond that, the impact these men have had as coordinators indicates that these are excellent football coaches.
These five coaches have all produced net gains in PPA and success rate on their respective sides of the ball that far outpace a Coordinator’s average impact. Who are they?
Brain Norwood is currently the Assistant Head Coach at UCLA and has past stints as Defensive Coordinator at Navy, Kansas State, Tulsa, and Baylor. Josh Gattis is the new Offensive Coordinator at Miami and has already excelled as Offensive Coordinator at Michigan and Alabama. Newland Isaac is the Co-Offensive Coordinator at Coastal, and he has helped to oversee the offensive explosion for the Chanticleers after serving as Offensive Coordinator at Albany State. Alex Atkins is the newly promoted Offensive Coordinator at Florida State, and he has past impressive tenures as OL coach at FSU and OC at Charlotte. Finally, Maurice Harris is the OC for Liberty and has previously served as Tight Ends coach at both Ole Miss and Arkansas State.
All five of these coaches have overseen significant improvement on their side of the ball and should be front and center whenever the next round of college football hiring kicks off.
We have shown that the racial imbalance in college football coaching is real and persistent. 35 more Black head coaches are needed in the FBS for the coaching ranks to match the hiring pool. The gap between the percentage of Black players and Black coaches is higher than it was in 2000. Black head coaches get 1.5 fewer years than their White counterparts. Why?
The imbalance is not based on differences in performance. We have shown that White and Black coaches perform at a level that is not significantly different. We have also shown that the hiring of head coaches and coordinators is largely based on social networks. White coaches are 43% more connected than Black coaches. Non-White coaches hire 30% more Non-White coordinators, but most of the opportunities to hire go to White coaches. The 20 head coaches who have employed the smallest percentage of Non-White Coordinators are White; the 19 most influential head coaches are White, but only 4 of them are above average in hiring Non-White coordinators. White head coaches have more connections, and they hire more White coaches.
But this imbalance does not have to be a permanent fixture. It can be mitigated with intentionality. Coaches with significant influence can do their part. One new Black head coach can have ripple effects for years at multiple schools. There are wonderful candidates who could and should get hired in the next cycle. It’s time to act.