“When we apply next-generation technology to advance player health and safety, everyone wins,” National Football League (NFL) commissioner Roger Goodell publicly stated in reference to the 2017 partnership between NFL and Amazon Web Services (AWS). The announcement, coupled with cumulative evidence that football is linked to significant injury, including brain trauma, indicated that the league was ready to leverage technology and data in new ways to get in front of the injury problem.
The NFL partnership with AWS expanded Next Gen Stats (NGS), an existing NFL sports data and analytics product by adding in-game, on-field player tracking technology. Now in its seventh year, NGS uses Zebra Technologies, a company that provides radiofrequency identification (RFID) technology in wide use to track inventory and player movement in games. The RFID tags are coin-sized chips that are placed in player's shoulder pads and in the football itself. Paired with 20 to 30 ultrawide band receivers placed throughout a stadium, the sensors are able to obtain individual player data, such as speed, acceleration, and total distance traveled. These measures can be used to calculate indirect measures like player intensity and power, while tracking location within inches at a rate of 10 times per second (1). This novel information can be provided to both coaching staff and fans for tactical insights, including line-up formations, receiver routes, and coverage options.
The promise of the technology and analytics for injury prevention lies in the ability to discover the relationship between individual player loads, play intensity, and number of plays and matchups; this allows for early identification of risk and prevention of injury. There also is the potential to bring a player back from injury using a more accurate method for reintroducing high intensity loads and assessing performance with high-fidelity insights delivered for each individual player.
Injuries among Active NFL Players
As it relates to head injury risk, the NFL has introduced more than 40 rule changes, including those that protect the quarterback, mitigate high-velocity impact (kickoff returns, targeting), ensure playing field safety (elimination of synthetic fields), as well as improvements in helmet technology. Unfortunately, while these rule changes were being implemented, the incidence of concussion reached an all-time high of 281 in the 2017 season. While this number improved significantly to 214 concussions in the 2018 season, concussion rates were again on the rise in the 2019 season at 224 (2). Despite progress, this remains an alarming number, particularly given the association between football related concussions and risk for chronic traumatic encephalopathy (CTE).
Musculoskeletal and soft-tissue injuries impact or will impact nearly every player in the NFL. Lower extremity injuries alone account for about 60% of games missed and these occur roughly 600 times per season (3). In the 2019 season, injuries most often associated with significant games missed, such as anterior cruciate ligament (ACL) and medial collateral ligament tears were reported in 47 and 109 players, respectively (2). When also considering the frequency of significant arm, shoulder, and back injuries, musculoskeletal injury prevention has become a top priority in the league.
The highest incidence of injury occurs with high training loads (4,5). Furthermore, rapid increases in workload from week to week predispose players to higher injury rates (6–8). Several athlete-specific characteristics, such as age, training intensity, injury history, and physical capacity, combined with training loads, are all known factors contributing to predisposition to injury (9,10).
Currently, NFL player injury data are compiled and analyzed by IQVIA (formerly QuintilesIMS), an independent, third-party company. Results of this Injury Surveillance and Analytics service are then shared with the NFL, the NFL Players Association (NFLPA), and the NFL medical and football committees. The analytics seek to identify likelihood of injury in relation to time points in a season, time off between games, and high-risk scenarios, such as kickoffs and special teams play. To the best of our knowledge, there has been no integration of injury data with in-game motion tracking data or with data collected from biometric sensors in practice.
Prior to player tracking technologies, assessment of acute workload and workload over time was limited to crude measurements of internal and external training load (10). While somewhat qualitative, high-intensity internal training load has been shown to be associated with increased injury rates (6). External training load, which is more objective, includes a measure of player distance, strength and conditioning sessions, player-to-player collisions, the number and intensity of sprints, change of direction (acceleration/deceleration), and jumps (9,10). The accuracy and predictive value of these measurements are historically extremely limited.
The 2020 season, due to the COVID-19 pandemic, required a curtailed preseason and restrictions on full-pad practices, leading to a high number of serious, in-game, musculoskeletal injuries. Less conditioned athletes without adequate preseason training time experienced abrupt increases in workload associated with game play that likely contributed to the higher injury rates (11). Player tracking technologies could be purposed on the training field as well as in games to gain a more holistic understanding of individual player loads accumulated prior to and during game play. This could greatly assist in directing training toward incrementally achieving loads comparable to what a player will experience in a game and help in understanding the relationship between load, speed, distance, acceleration power, and actual training or in-game injury.
In addition to Zebra in-game player tracking, there are a number of validated technologies in use that can provide tactical information related to in-game scenarios, such as offensive and defensive formations, receiver routes, and coverage, as well as having high fidelity for assessing individual levels of player performance like speed, acceleration, power, and distance. These digital devices, much smaller than a cell phone, are worn on shoulder pads or tucked under jerseys during training and use inertial sensors embedded within the device to provide real-time data on nonlocomotive activities including jumps and collisions (12). Many of the wearable devices can be used in training as well as in play, and when equipped with gyroscopes, accelerometers, and magnetometers offer a more comprehensive assessment of player load. Beyond measuring workload for health maintenance purposes, these devices can be used to ensure training intensity in practice matches the level seen in games and are purposed to predict player matchup success based off of individualized historical pace of play. The Table lists the technologies that are currently being used during NFL games (Zebra) or are approved for use in practice (13,14).
List of wearable devices approved for use in NFL games (Zebra) or practice.
||Primary Product Functionality
||Zebra MotionWorks Sport
||Device chips in uniforms, pads, and footballs connected to sensors and receivers throughout stadiums.
used to quantify player and football movement and distance profiles.
||Wearable device in upper body garment
antenna tracking of sport-specific movement, player load, orientation, and heart rate.
||Wearable device in upper body garment; Wearable wrist sensor for contact tracing
designed to record player position, motion, orientation, and heart rate; UWB measuring physical distance and length of time between users.
||Wearable device in upper body garment
||Quadruple GLONASS aimed for player tracking and inertial sensors quantifying performance block.
||Polar Team Pro
||Wearable device in upper body garment or around trunk
||GPS and MEMS
designed to measure performance metrics and heart rate.
|STATSports Group Limited
||Wearable device in upper body garment
||GPS designed to monitor player load and performance, collision analysis, and heart rate.
aA wireless system comprised of two components: tags and readers. The reader is a device that has one or more antennas that emit radio waves and receive signals back from the RFID tag. Tags, which use radio waves to communicate their identity and other information to nearby readers, can be passive or active.
bA navigational system using satellite signals to fix the location of a radio receiver on or above the earth’s surface.
cA satellite-based radio navigation system run by the Russian Ministry of Defense. It uses 21 medium earth orbit (MEO) satellites and three spares. Similar to the GPS in the United States, GLONASS enables 3D positioning anywhere on earth within 100 m to 150 m for the public and 10 m to 20 m for the military.
dA radio-based communication technology for short-range use for the fast and stable transmission of data indoors and outdoors.
eTechnology defined as miniaturized mechanical and electro-mechanical elements (i.e., devices and structures) that are made using the techniques of microfabrication. Microsensors are as transducers, which are devices that convert energy from one form to another. In the case of microsensors, the device typically converts a measured mechanical signal into an electrical signal for almost any sensing modality.
Wearable Digital Biosensor Studies in NFL Athletes
Many of the wearable sensors listed have position sensors, which triangulate signal transmission from multiple Global Positioning System (GPS) satellites orbiting the earth and can accurately determine the velocity and position of an athlete (13). Several studies have used wearable GPS-based technology to evaluate risk of injury to NFL players. One of the first studies to use GPS technology in the NFL used the Catapult device and assessed athletes on a single team from the 2014 through the 2016 season. Of the 101 soft-tissue injuries with complete GPS and clinical data, injured players saw a 111% increase in workload, whereas uninjured players saw just a 73% increase in workload during the week of injury. Furthermore, athletes with acute to chronic workload ratio higher than 1.6 were 1.5 times more likely to sustain an injury (11). These findings were particularly notable during the preseason, where workload changes are generally higher. A subsequent 2018 study of 101 NFL athletes over 24 wk found that very high player loads substantially increased the risk of injury on a given training day (15).
Digital sensors are used not only to quantify workload to prevent injury but also used to assess recovery from injury. After Carson Wentz tore his left ACL and lateral collateral ligament in 2017, the Philadelphia Eagles used the Zebra RFID chip to track his acceleration during common drop-backs, plants, cuts, and throws (16). In addition to a tag on Wentz, data were analyzed from a chip inside the football capable of assessing ball velocity and spin rate after every Wentz throw. Because the Eagles were one of the first teams in the NFL to adopt and use this technology in 2014, they had previously established baseline mechanics for Wentz while healthy. These cumulative data of acceleration, force, and ball velocity allowed for throw-by-throw comparison and were invaluable in assessing Wentz's progression to recovery to his normal baseline. Future analysis of lower body movements using accelerometers like Plantiga could be useful in assessing asymmetry, ground contact, and gravitational forces over time.
Current Use of Tactical and Individual Performance Technology in the NFL
The NFL and AWS have worked to create a “Digital Athlete” platform, described as “a computer simulation model of an NFL player that will be able to be used to model infinite scenarios within the game environment without any risk to the athletes” (17). Similarly, Amazon recently inked a deal with the Seattle Seahawks to provide the company's hometown team with cloud, machine learning, and artificial intelligence services (18). The technology will help the Seahawks better track player performance in both games and practice by collecting a data set of player and ball movement within inches. Once the data are acquired and analyzed, AWS can build predictive models for thousands of in-game scenarios including how best to rush individual quarterbacks, route detection, lane integrity during special-teams plays, formation line-ups and matchup success, expected yards after catch, and other crucial aspects of the game. These tactical data should allow for a competitive advantage for the team, a more engaged experience for the fans, and most importantly, has the potential to gain deeper understanding into the relationship between individual workload and injury.
To date, it seems that the most used aspect of this technology has been purposed for fan engagement and sports media. For instance, during media time outs, fans are able to see exactly how fast a quarterback like Lamar Jackson ran on a quarter back (QB) sneak, or what the catch probability was and how much separation wide receiver DK Metcalf got from the defensive back on a pass play from Russell Wilson. Furthermore, fans can look up novel statistics, such as QB pressure-evaded rate, expected player rushing yards, and predicted performances, to ensure fantasy football success and bragging rights for the year.
The NFL also is using RFID technology to better manage COVID-19 risk. The league recently inked a deal with Kinexon that provides players and staff with a half-ounce tag worn as a wristband or embedded into equipment; the tag lights up red and alarms when two sensors are within 6 ft of each other for more than 5 s. Each sensor is registered as a unique ID and can track contacts.
Digital Biosensors, Health Data, and Players’ Rights
The recently approved 2020 to 2030 NFLPA Collective Bargaining Agreement (CBA) details terms and conditions governing the use of digital sensors and technologies in practice and in game play. The agreement allows the NFL to require all athletes to wear sensors during games to track player movement and may use these data commercially as long as the NFLPA is given advanced notice. Implications on the release of these data are vast, and regulations on the privacy of this information “shall comply with all federal and state laws” (14). The CBA also defines all digital biometric data collected from an NFL athlete as health data and states that each player has ownership of their personal data. While team staff may have access to the data, it cannot be referenced or used in contract negotiations. For practice and training purposes, it is up to the individual players to approve the use of worn technology themselves. The CBA also states that a Joint Sensors Committee will be created to review and formally approve the use of current and future player worn sensors (14). The purposing of biometric and tactical data, when applied to player assessment, can be viewed as a dynamic and continuous health and human performance medical record. In traditional medical settings, patient health data are protected by privacy legislation, but, other than the CBA, the policies and laws that govern data collected in elite human performance is nonexistent. The issue is complicated by the fact that the same set of data is currently used and approved for enhancing the fan experience in products such as NGS. An additional challenge will be related to the use of player data for sports wagering. In-game Zebra data could be leveraged to include a human performance or readiness health score on each player, would be attractive to a sports wagerer or fantasy user. How the CBA governs this issue could determine whether the dynamic health information of an NFL player will be truly protected.
Players are clearly concerned about protecting themselves from the risk of injury and the strongest objections to the 2020 CBA related to workload. Marquee players, such as Aaron Rodgers, Russell Wilson, and JJ Watt, have all voiced concerns about the CBA, and the addition of one more regular season game to the NFL season that Rodgers stated “was never something to be negotiated” (19). This additional game poses additional risk of injury, which many players argue far outweighs the benefit to the salary increase. Concessions, however, were made by the NFL in reducing the preseason from four to three games, limiting the total number and consecutive days of full-padded practices, and capping the length of practice to 2.5 h during training camp (14). Whether these mitigations will increase or decrease injury risk is unknown.
Currently available technologies, used in practice and games, have the potential to revolutionize and inform training, injury prevention, and recovery, if seriously purposed toward improving human performance and health outcomes of players. These data can serve as a dynamic and individualized health record for NFL players. Players will be able to develop a sense of their loads over a preseason, season, and career, and this knowledge can help inform all sorts of decisions from player valuation to player health longevity.
The issues surrounding player tracking and other novel data collection and analysis are complex. The NFL-Amazon 2017 announcement promised that the new data sets would be purposed in the service of injury. Yet, other than delivering data back to teams, this does not appear to be a very centralized or organized effort. Player ownership and privacy issues need to be further defined so that player protections are in place. This type of dynamic assessment of player health from week to week can easily be weaponized against a player, particularly if made available as data that can influence sports wagering.
From a tactical perspective, digital sensors and player tracking technology can provide accurate individualized information to better inform players, coaches, physicians, trainers, and physical therapists to fine tune player workload.
It also will take time, perhaps more than one season, to make sense of continuous player and team tactical data that have not been seen before. Time is needed to bench mark these new data to things that are known, like injuries or other player and team characteristics prior to developing plans or policies that are transparent and in the best health interest of the player. Players must trust that the data will extend their health and performance, especially if it results in resting a player when they are not yet injured. This will take time, policy, organization, and unprecedented collaboration. It will take more than an NFL-Amazon press release.
Leslie Saxon serves on the NFL/NFLPA Joint Sensors Committee as one of the NFLPA representatives. Otherwise, the authors declare no conflict of interest and do not have any financial disclosures.
1. NFL Next Gen Stats. NFL Football Operations
. 2020. [cited 2021 July 1]. Available from: www.operations.nfl.com/the-game/technology/nfl-next-gen-stats/
2. Injury Data. NFL Play Smart, Play Safe
. 2020. [cited 2021 July 1]. Available from: www.playsmartplaysafe.com/newsroom/reports/injury-data/
3. Battista J. NFL Reveals 2019 Injury Data, Hopeful Rule Changes Are Working. NFL.com
. [cited 2021 January 23]. Available from: www.nfl.com/news/nfl-reveals-2019-injury-data-hopeful-rule-changes-are-working-0ap3000001098679
4. Foster C. Monitoring training in athletes with reference to overtraining syndrome. Med. Sci. Sports Exerc
. 1998; 30:1164–8.
5. Colby MJ, Dawson B, Heasman J, et al. Accelerometer and GPS-derived running loads and injury risk in elite Australian footballers. J. Strength Cond. Res
. 2014; 28:2244–52.
6. Cross MJ, Williams S, Trewartha G, et al. The influence of in-season training loads on injury risk in professional rugby union. Int. J. Sports Physiol. Perform
. 2016; 11:350–5.
7. Bowen L, Gross AS, Gimpel M, et al. Accumulated workloads and the acute: chronic workload ratio relate to injury risk in elite youth football players. Br. J. Sports Med
. 2017; 51:452–9.
8. Piggott B, Newton MJ, McGuigan MR. The relationship between training load and incidence of injury and illness over a pre-season at an Australian Football League club. J. Aust. Strength Cond
. 2009; 17:4–17.
9. Impellizzeri FM, Rampinini E, Marcora SM. Physiological assessment of aerobic training in soccer. J. Sports Sci
. 2005; 23:583–92.
10. Gabbett TJ. The training—injury prevention paradox: should athletes be training smarter and
harder? Br. J. Sports Med
. 2016; 50:273–80.
11. Li TR, Rambhia S, Sheehan J, Salata MJ, Voos JE. Does overexertion correlate with increased injury? The relationship between player load and soft tissue injury in professional American football players utilizing wearable technology. 2017 ISAKOS Congress Abstract. [cited 2021 July 1]. Available from: https://www.isakos.com/meetings/2017congress/AbstractView?EventID=8912&Date=&courseid=
12. Chambers R, Gabbett TJ, Cole MH, et al. The use of wearable microsensors to quantify sport-specific movements. Sports Med
. 2015; 45:1065–81.
13. Seshadri DR, Li RT, Voos JE, et al. Wearable sensors for monitoring the internal and external workload of the athlete. NPJ Digit Med
. 2019; 2:71.
14. National Football League. Collective Bargaining Agreement. 2020. [cited 2021 July 1]. Available from: https://nflpaweb.blob.core.windows.net/media/Default/NFLPA/CBA2020/NFL-NFLPA_CBA_March_5_2020.pdf
15. Ward P, Tankovich M, Ramsden JS, et al. Volume and intensity are important training related factors in injury incidence in American Football Athletes. In: MIT Sloan Sports Analytics Conference
. 2018. [cited 2021 July 1]. Available from: http://www.lukebornn.com/papers/ward_ssac_2018.pdf
. How Nickel-sized Tech Helps Eagles Track Wentz’s Recovery. [cited 2021 July 1]. Available from: http://www.espn.com/blog/philadelphia-eagles/post/_/id/25698
17. National Football League and Amazon Web Services Team Up to Transform Player Health and Safety Using Cloud Computing and Artificial Intelligence. Business Wire
. [cited 2021 July 1]. Available from: www.businesswire.com/news/home/20191205005673/en/National-Football-League-Amazon-Web-Services-Team
18. Levy N. Seattle Seahawks Partner with Amazon Web Services for Cloud Deal to 'Fuel Future Championships'. GeekWire
. [cited 2021 July 1]. Available from: www.geekwire.com/2019/seattle-seahawks-turn-amazon-web-services-new-cloud-deal-fuel-future-championships/
19. Florio M. Aaron Rodgers Explains His Opposition to Proposed CBA. ProFootballTalk
. [cited 2020 February 27]. Available from: profootballtalk.nbcsports.com/2020/02/26/aaron-rodgers-explains-his-opposition-to-proposed-cba/