Racial Prejudice in the NFL

Cam Newton. 2015 NFL MVP. Quarterback. Black. Source: Kevin C. Cox (Getty Images)

Author: Sam Mestel (Weinberg ’20)

Introduction

According to a 2018 US Census Bureau report, only 13.4% of the United States population can trace their ancestry to sub-Saharan Africa. So how is it that almost 75% of the players in the NFL are black? How does such a large discrepancy make any sense? Could it be due to statistical prejudice in the NFL? This article aims to answer this question. The best area to look at for this type of analysis would be the draft process, as it is how the influx of players is brought into the league each year.

In order to properly conduct this study, I had to make judgment calls when purporting the race of any individual player. For simplicity, a player who did not fit the profile of either “white” or “black” were excluded from the study. However, some players, who may have been of another ethnic background (say Samoan), but appeared to be either black or white, were included under the corresponding category – since the whole point of this study is to understand how the exterior looks of players affect their draft stock. This study will look at every white and black position player drafted since 2006, which is the beginning of the current Combine era.

Procedure

In order for this study to work, I had to come up with a few new statistics. The first one is “Bust Percentage” (bust%). Bust Percentage is calculated by dividing the number of draft busts on each draft day (explained below) by position, by ethnicity, and by the number of total prospects at that position and of that race on that draft day. The second statistic that I am going to look at is “Sleeper Percentage” (sleeper%). This is essentially the opposite of Bust Percentage – it is calculated the same way but using sleepers instead of busts. For the purpose of this study, busts are characterized by a Career Approximate Value per Game (CarAV/G) of one or more standard deviations below the mean at their position on their draft day (both races are included in this pool). Likewise, sleepers are characterized by having a CarAV/G of one or more standard deviations above the mean. This brings us to the third statistic that I will introduce, which is CarAV/G. CarAV/G is simply a measure of a players production throughout their career, adjusted for the number of games they have played. I have adjusted for games here to account for shorter careers (say a player who was only drafted last year) and injuries (a player could have had a better season but they were only able to appear in a few games, so their CarAV will be lower). CarAV itself is an approximation stat developed by pro-football-reference.com in order to approximate the value of every player in NFL history, weighted with their best seasons counting for the most.

The next statistic that I created to facilitate this study was “qWhite” and “qBlack”. This is calculated by subtracting the bust% from sleeper%. This allows us to gauge the number of players that were overlooked vs over-drafted at each position by race. A higher positive value for these statistics indicates that a higher percentage of players (by race) were taken lower in the draft but ended up being better than players higher in the draft. Thus, in theory, these statistics would show prejudice if the value of the difference between the two is large enough.

The last statistic I created was “Prejudice Quotient”. Prejudice Quotient is calculated by subtracting “qWhite” from “qBlack”. Essentially, a negative Prejudice Quotient allows us to attribute prejudice at the position towards white players and a positive Prejudice Quotient would attribute prejudice at the position to black players.

Lastly, in order to make my calculations, I divided all players into 3 categories when looking at their bust% and sleeper%. These categories were the day of the draft they were picked on (either Day 1, Day 2, or Day 3). Since the skill of players drops off precariously in general after the first and third rounds, it makes a lot of sense to do it this way (I did not do a significance test for all of the positions, but the data shows the vast differences in CarAV/G when looking at the difference in draft day selected).

Results

The table below represents an overview of the results of this study.

Right off the bat, the numbers show a lot of things that are to be expected. Fans constantly banter about how teams are very prejudiced against black players at the quarterback position. It seems that, for some reason, black quarterbacks always begin to slide as the draft approaches. An example would be Dwayne Haskins from this year, who was originally a possibility for the Cardinals’ 1st overall pick early throughout the draft process but fell to the 15th pick. There have also been a lot of black quarterbacks taken later on in the draft go on to become extremely valuable commodities for their team, such as Russell Wilson and Dak Prescott. In fact, the sleeper% for black quarterbacks is 33.3% while only 13.1% for white quarterbacks. That is what makes the really big difference here since they bust at quite similar rates (8.3% and 9.5% for black quarterbacks and white quarterbacks, respectively). This makes sense with the narrative of pundits like ESPN’s Stephen A. Smith, who claims that black quarterbacks are under-drafted on the basis of their race. This study would completely agree with that claim. In fact, it is not even that white quarterbacks are being over-drafted. The real issue is that a large number of capable black quarterbacks are falling into the later rounds of the draft.

Below is an overview of those splits by day when looking at the quarterback position.

The other position that I want to directly address is Wide Receiver. It seems like every year, the New England Patriots find a white player to successfully fill in the slot receiver position. So why is it that so many white receivers are overlooked throughout the draft process? One explanation that has been made for this time and time again is that white receivers are slower and do not fit the “ideal build” of a receiver in today’s NFL, who teams want to be able to stretch the field. Is this even true? Surprisingly, the average 40 time for white receivers is 4.5 whereas the average 40 time for receivers, in general, is 4.47. While this difference seems minuscule, it is actually significant (up to a 99% confidence level, with a t-stat of 2.508). Ok, so does it make sense that teams are discriminating against white receivers because they are statistically slower? No! The data shows that white receivers are still being severely under-drafted. While black receivers bust at a rate of approximately 8.1%, there have been 0 white receivers who could be classified as busts. Additionally, the black receiver sleeper rate is only 13.6%, whereas it is 20.6% for white receivers. So white receivers drafted high are performing well, while white receivers drafted later are significantly exceeding expectations almost 1/5th of the time. This is a great example of how false stereotypes reinforce an untrue narrative and can lead to inexcusable and self-harming prejudice in the NFL.

Below is an overview of those splits by day when looking at the receiver position.

The overall sample shows significantly more prejudice against white players than against black players; in fact, out of the 11 position groups that I measured, 2 of them were significantly prejudiced against black players. 3 of them had prejudice quotients too small to make a decisive decision, and 6 of them were significantly prejudiced against white players. This is shocking. Why is there so much discrimination against white players when the majority of owners and general managers are white? In my own opinion, I think a lot of this has to do with the media. Whenever pundits talk about white players, they refer to them as “scrappy” and “gym rats”. For example, NFL.com’s Lance Zierlein described Oakland Raiders’ 5th round draft pick Hunter Renfrow (a white receiver) as the following: “He’s a great route runner. He catches everything. He’s quicker than fast. He’s a coach’s son. I think I checked off all the cliche boxes for a white slot receiver, but they are all true. Watch him work against Minkah Fitzpatrick and tell me if you think he can play.” Sure, Zierlein is listing all of Renfrow’s positive attributes, but these are the exact same attributes that are causing the prejudice and demise of white players at every position around the league outside of Quarterback and Offensive Guard.

Discussion/Conclusion

Something else that was quite shocking that I came across when conducting this study was that there were zero white Cornerbacks drafted since 2006, which is the reason I combined Safeties and Cornerbacks for use as a Defensive Backs category. Upon more digging, I found that there has not been a white cornerback drafted since Jason Sehorn in 1994. I truly do not know how to explain this, but looking at defensive back in general, there is a ton of statistical discrimination against white players, which, given this incredible discrepancy, is to be expected.

So now that the problem has been identified, what should be done to address it? I think it is important to take this study with a grain of salt. Sure, there is statistical discrimination at almost every position, but I am also quite sure that owners, GMs, and coaches do not sit around the table in their draft “War Rooms” purposefully moving players down the draft positions because of their race. I do not necessarily know if there is, in fact, a good solution to this issue. That being said, I think it is important to have an open mind. Just because a player is white does not mean he is not athletic enough to become a player in the NFL, and this goes for every position, not only Receiver and Defensive back. On the other hand, I think it is important for talent evaluators to not discredit black Quarterbacks in the same regard. Warren Moon, inducted into the Pro Football Hall of Fame in 2006, was the first black Quarterback to be inducted. Sure, black players seem to choose to play Quarterback at a much lower rate than other positions. However, that does not mean that they are incapable of doing so. In fact, only 3 seasons ago, the NFL’s Most Valuable Player was a black Quarterback in the Carolina Panthers’ Cam Newton. Prejudice is bad regardless of whether it is intentional or subconscious. Moreover, in the long run, teams who are less prejudiced will get closer to the equilibrium amount of busts and sleepers. This may be why Bill Belichick used a 7th round pick on Kent State’s Julian Edelman, who was this year’s Super Bowl MVP. It is because he looks past racial stereotypes and collects talent for what it inherently is.

Works Cited

https://www.nfl.com/prospects/hunter-renfrow?id=32195245-4e35-8499-ff02-b25484b87290

https://www.pro-football-reference.com

https://www.census.gov/quickfacts/fact/table/US/PST045218

(Upon request, the author would be happy to share raw data and spreadsheets used to make all calculations.)

Email samuelmestel2020@u.northwestern.edu.

Be the first to comment on "Racial Prejudice in the NFL"

Leave a comment

Your email address will not be published.


*