Secondary Logo

Global Positioning System Monitoring of Selected Physical Demands of NCAA Division I Football Players During Games

Bayliff, Garrett E.; Jacobson, Bert H.; Moghaddam, Masoud; Estrada, Carlos

The Journal of Strength & Conditioning Research: May 2019 - Volume 33 - Issue 5 - p 1185–1191
doi: 10.1519/JSC.0000000000003137
Original Research
Free

Bayliff, GE, Jacobson, BH, Moghaddam, M, and Estrada, C. Global positioning system monitoring of selected physical demands of NCAA Division I football players during games. J Strength Cond Res 33(5): 1185–1191, 2019—Global positioning system (GPS) tracking of athletes in selected sports is a new innovation into obtaining comprehensive data regarding physical output with respect to distance travelled (DT), acceleration, and change of direction. The purpose of this study was to determine selected physical demands of American football players during the course of games and to compare such data by player position. Offensive lineman (OL) (n = 14) and defensive lineman (DL) (n = 9) and offensive wide receivers (WRs) (n = 10) and defensive backs (DBs) (n = 10) were fitted with GPS monitors during games. Collected data included DT, maximum velocity (MV), and acceleration (AC), deceleration (DC) distance at 2 intensities. Results indicated that DBs travelled significantly (p < 0.05) greater distances than OL and WR, but not DL. For MV, DBs and WRs were not significantly different but were significantly different from OL and DL. Also, DL was significantly different than OL. For the most intense acceleration (3–10 m·s−2), WR accelerated significantly further than all other positions and DBs accelerated further than DL and OL. There was not significant difference between DL and OL. For deceleration at the high-intensity measure, significant differences existed among all positions. Underestimation of workload during games could be a factor for the overuse and soft-tissue injuries and more serious injuries. Furthermore, using GPS tracking of similar variables as found in this study may benefit coaches and trainers in many other high-intensity sports.

Neuromuscular Physiology Laboratory School of Kinesiology, Applied Health and Recreation Oklahoma State University, Oklahoma State University, Stillwater, Oklahoma

Address correspondence to Bert H. Jacobson, bert.jacobson@okstate.edu.

Back to Top | Article Outline

Introduction

Tracking technology such as global positioning systems (GPSs) in conjunction with triaxial accelerometers has enabled researchers and coaches to obtain comprehensive data regarding physical output during sport training and competition (1). In addition, earlier studies by Barr et al. (3), Johnson et al. (9), and Wundersitz et al. (14) verified GPS validity and reliability in accurately tracking movement demands in sport. For instance, Boyd et al. (4) conducted a study to calculate accelerations and decelerations in 3 planes and found within-device reliability (CV 0.91–1.05%) and between-device reliability (CV 1.02–1.04%) were both superior to previous studies conducted on different brands of microsensors. Additional important data found from this study was that the devices remained stable over a long period, which in turn kept them from drifting from the baseline measurement.

These tracking devices are capable of accumulating substantial amounts of movement-related data during competition and practice and have been used to collect activity profiles of athletes in Australian football, cricket, hockey, rugby, and soccer (2), but reports of such use in American football is much more rare. For example, a study by Wellman et al. (11) sought to examine movement demands of NCAA Division I football players at various positions through GPS. Another study by the same principal author examined impact profiles by positions using GPS technology (12).

Data collected by GPS tracking systems may be used, in part, to determine distance traveled, velocity, acceleration, and deceleration thereby allowing for the predictability of fatigue (2), need for substitution, need for individualized training to reduce injury risk potential, and to develop game strategies. Indeed, running loads have been correlated with soft-tissue injury (4,7).

Recently, in the sport of American football microsensors in the form of GPSs integrated with accelerometers have been incorporated to gain more accurate insight to the amount of physical activity actually occurring during play and practice. Although coaches may feel they have a good impression of the amount of work and intensities players are incurring, such impressions are subjective and lack finite accuracy. Indeed, it becomes more difficult to subjectively evaluate the physical stress level of players during a competitive game because of the distractions of game strategizing.

Currently, little is known of the volume of work with respect to distance covered, velocities, and similar physical demands that are placed upon American collegiate football players at selected positions. An earlier study by involving NCAA Division I football players determined that starters obtained higher velocities than nonstarters, and that nonlineman covered more distance than linemen during practice, concluding that linemen engage in more isometric work than nonlinemen (5). In a more recent study by Wellman et al. (11), involving NCAA Division I college football players, the researchers examined the physical physiological movement demands during games using GPS technology. The authors found that for both offensive and defensive positions teams, significant (p ≤ 0.05) physical performance differences existed between football positions. The results of the study suggested that wide receivers (WRs) and defensive backs (DBs) completed significantly (p ≤ 0.05) greater total distance, high-intensity running, sprint distance, and high-intensity acceleration and deceleration efforts than players holding other offensive and defensive positions. However, the authors did not indicate the type of offense or defense that was analyzed. Indeed, the newer spread offense used largely in the Big XII conference contrast sharply with the run-oriented teams of many of the other major conferences, in that the spread offense is much faster and is pass oriented.

Thus, the goal of this study was to compare GPS tracking of distance traveled, top speed, acceleration, and deceleration variables among corresponding offensive and defensive positions to gain a better understanding of the demands placed on athletes during games, for example, offensive linemen (OL) vs. defensive linemen (DL) and receivers vs. defensive secondary. With a thorough understanding of the physical workload placed on players at various positions, strength/conditioning coaches may be able to construct better preparation methods for competitive play by implementing specific practice and conditioning protocols for each position, and athletic trainers may gain better insight to injury occurrences because of the demands placed on the players.

Back to Top | Article Outline

Methods

Experimental Approach to the Problem

NCAA Division I American football players were the participants in this study. Data were collected after the season based on position and playing time. Participants were grouped by their playing positions on both offense and defense. The groups consisted of the positions that most closely resembled each other on the 2 sides of the ball (offense and defense), which were OL and DL, and WRs and defensive secondary players. For analysis, only data for those players logged a 66% on-the-field presence used. In addition, all individual data were collapsed into an aggregate mean and SD for comparisons.

Back to Top | Article Outline

Subjects

NCAA Division I American collegiate football players 18–24 years of age (mean ± SD; 19.9 ± 1.5 years) participated in this study after reading and signing a University IRB-approved consent form. This study was approved by Oklahoma State University. Wearing the GPS tracking device presented not additional risk or harm to the participants. Participants included 14 OL, 10 WRs, 9 DL, and 10 DBs. All players' physical characteristics are included in Table 1. Participants were chosen based on equitable percentage of total snaps, which, similarly to another study (11), required all subjects participating in this study to have accumulated a minimum of 75% of total snaps (

= 59.1 ± 4.7) in at least 2/3rd of the 9 conference games. Characteristics of the participants are illustrated in Table 1.

Table 1

Table 1

Back to Top | Article Outline

Procedures

To track physical game activity, GPS tracking devices (Optimeye S5; Catapult Innovations of Australia, Melbourne, Australia) capable of a real-time 250-m wireless range using triaxial accelerometers with up to 1,000-Hz sample rate and 100-Hz default, and gyroscopes measuring at 100 Hz were placed on each player. Each device was capable of recording frequency and magnitude of acceleration and deceleration (m·s−2) in 3 dimensions. Computer software was used as an interface for managing the data after following the download. Microsoft Excel was used to place the data in a table and graph formats for easy access.

Each monitor was fully charged when placed between the shoulder blades on the posterior side of the shoulder pads. For the monitors to be secured throughout the entire game, they were placed in custom-designed mesh pouches with a Velcro strap sewn on for increased stability. In no way did the monitor restrict the athlete's movement during athletic play. Conference games were aggregated into one data set. Players' on-the-field time was standardized similarly to Wellman et al. (13), to accurately compare each variable recorded. Upon completion of each game, the monitors were immediately removed, docked in their respective cases, plugged into a laptop, and downloaded. Once all the data were downloaded and in the operating system, it was manually time-stamped into quarters for each game. The data needed to be time-stamped to cut out unwanted “noise” in the data, which included rest times (time-outs), down times between quarters of a game, and halftime. Time-stamped, data were dropped into premade, custom tables and graphs that were designed manually before that particular session. For ease of use, the tables were saved and downloaded onto a Microsoft Excel document.

Back to Top | Article Outline

Data Acquisition

Seven independent variables were used to measure game play stress and intensity. Distance travelled (DT) was measured in meters (m) and was used to determine the total distance traveled by each athlete during the game. Maximum velocity (MV) was measured in kilometers per hour (km·h−1) and was defined as the maximum speed that the athlete attained during game play. A parameter developed by the manufacturer termed total inertial movement analysis (Total IMA) was used to measure the total number of times the athlete accelerated, decelerated, or changed direction at high intensities (>3.5 m·s−2). Based on the manufacturer's description, IMA allows for tracking of performance through the use of inertial sensors in the form of 100-Hz triaxial accelerometers for recording change in motion and triaxial gyroscopes for recording rotation without the use of the GPS capabilities. Thus, IMA assesses anything the GPS will miss because the movement is too small by converting the combination of the data into movement of the body, not the movement of the unit. Therefore, IMA records accelerations, deceleration, change in direction, jumps, and free running at low, medium, and high efforts.

Acceleration and deceleration distances travelled were categorized into predetermined bands to track intensity of DT. Deceleration bands ranged from 1 to 8 eight with bands 1 through 4 being decelerations and bands 5 through 8 being accelerations. For this study, only moderate- and high-intensity deceleration and acceleration distances were analyzed, thus deceleration bands 1–2 and acceleration bands 7–8 were used for analysis. Band 1 consisted of high-intensity decelerations (>3–10 m·s−2), band 2 consisted of moderate decelerations (2–3 m·s−2), band 7 consisted of moderate acceleration (2–3 m·s−2), and band 8 consisted of high accelerations (>3 m·s−2) (Table 2).

Table 2

Table 2

Back to Top | Article Outline

Statistical Analyses

A one-way analysis of variance (ANOVA) was used to determine positional group effects. For significant main effects indicted by the one-way ANOVA, Newman-Keul post hoc tests were performed to determine the specific location of significance among the groups. Alpha (α) levels for all statistical testing were set at p ≤ 0.05 as the acceptable level of significance. Statistical analyses were performed using IBM Statistical Package for the Social Sciences (SPSS for Windows, version 24; SPSS, Inc., Chicago, IL, USA).

Back to Top | Article Outline

Results

Anthropometrics

Comparative analysis yielded no significant difference in age. However, OL were significantly (p < 0.01) taller than WR, DB, and DL. In addition, OL were significantly heavier than WR (p < 0.01) and DB (p < 0.01) and DL significantly heavier than WR (p < 0.01) and DB (p < 0.01).

Back to Top | Article Outline

Distance Traveled

For distance traveled by player's position, group mean values and SDs are illustrated in Figure 1. Defensive backs travelled a significantly greater distance than WR (p = 0.04) and OL (p = 0.01), but not further than DL (p = 0.05). There were no significant differences in DT among WR, OL and DL.

Figure 1

Figure 1

Back to Top | Article Outline

Maximum Velocity

For MV by position (Figure 2), DBs and WRs had significantly greater MV (km·h−1) than OL (p < 0.01 and p < 0.01, respectively) and DL (p < 0.01 and p < 0.01, respectively). In addition, a DL had significantly greater (p < 0.01) velocity than OL. There was no significant difference between DB and WR in MV.

Figure 2

Figure 2

Back to Top | Article Outline

Total Inertial Movement Analysis

Group means for Total IMA are illustrated in Figure 3. Offensive lineman and DL accelerated, decelerated, or changed direction significantly more times than DBs (p < 0.01 and p < 0.01, respectively) and WRs (p < 0.01 and p < 0.01, respectively). There were no significant differences between OL and DL (p = 0.66) nor between WR and DB (p = 0.12).

Figure 3

Figure 3

Back to Top | Article Outline

Acceleration Band 7 (2–3 m·s−2) and Band 8 (3–10 m·s−2)

Group means for acceleration bands 7 and 8 are illustrated in Figure 4. For moderate acceleration (2–3 m·s−2) distance covered in band 7, DBs had significantly greater acceleration distance than DL (p < 0.01) and OL (p < 0.01). Similarly, WRs had significantly greater acceleration distance than DL (p < 0.01) and OL (p < 0.01). Furthermore, DL had significantly greater (p < 0.01) acceleration distance than OL. There was no significant difference between DBs and WRs.

Figure 4

Figure 4

For acceleration in band 8 (>3 m·s−2), WRs generated significantly (p < 0.01) greater distance than all other positions. While considerably less than WRs, DBs covered significantly (p < 0.01) more distance than OL and DL. There was no significant (p = 0.08) difference between OL and DL in distance covered.

Back to Top | Article Outline

Deceleration Band 1 (2–3 m·s−2) and Band 2 (3–10 m·s−2)

Group means of deceleration band 1 and band 2 are illustrated in Figure 5. Defensive backs and WRs recorded the greatest distance of moderate deceleration as defined by band 1 and DL and OL the least distance. Significant differences in band 1 distance were found between DBs when compared with DL (p < 0.01) and OL (p < 0.01). Similarly, WRs recorded significantly greater deceleration distances than DL (p < 0.01) and OL (p < 0.01). There were no significant differences between DBs and WRs and between DL and OL.

Figure 5

Figure 5

For high-deceleration distance covered in band 2, DBs posted significantly greater deceleration distance than DL (p < 0.01) and OL (p < 0.01). Wide receivers also posted significantly greater deceleration distances than DL (p < 0.01) and OL (p < 0.01). There was no significant difference in deceleration distance between WRs and DBs (p = 0.17). However, there was a significant difference between DL and OL (p = 0.03).

Back to Top | Article Outline

Discussion

One goal of GPS tracking is to identify the amount of physical stress in the form of distance covered and in the form of moderate- and high-performance intensities that are placed on selected football position during actual games. Largely anecdotally, it has been suggested that if the defense is on the field for an extended amount of time, they will tire more quickly than the offense. Although this has not been thoroughly measured, one isolated article indicates that the defense is no more at jeopardy of incurring fatigue than the offense (10). These data propose that defensive player logs greater distances than offensive players, and that DB and WR reach higher velocities, and greater acceleration and deceleration than linemen.

Data involving factual physical exertion such as distances traveled and moderate- and high-intensity movement may be useful to possibly develop better exercise protocols and practice protocols that are more position-specific to the demands physical output. In doing so, coaches, strength and conditioning professionals, and exercise scientists can put their athletes in a much better position to succeed during competitive play. Much of football today is replete of many high-powered offenses that move at a fast rate. Some of which are no-huddle thereby allowing little time for recovery. By measuring selected variables that contribute to energy usage during play by certain positions, a better understanding of exerted effort may be had.

Results suggest that the DB traveled significantly further than OL, DL, and WR. While not significant (p > 0.05), WR traveled 4 and 9.7% further than DL and OL, respectively. These results are not in full agreement with others (5,11,13), in that the results suggest that nonlineman (i.e., DB and WR) cover significantly greater distances than lineman during practice and games. Logically, much depends on the style of offense and defense played. For instance, a passing-oriented team would require WR to accumulate greater distances in contrast to a run-oriented offense. It may also be reasonable to assume that DBs travel further than WR because while the WRs have defined routes for each passing play and some blocking assignments during running plays, DLs are not only covering a receiver or a zone but constantly pursuing the ball from sideline to sideline regardless of what type of play has been called.

This study's findings concerning MV are similar to previous studies conducted (11–14). Not unexpectedly, skill positions (DB and WR) attained significantly greater maximum velocities (km·h−1) when compared with non-skill positions (DL and OL). Wellman et al. (11,13) and DeMartini et al. (5) both recorded similar findings during games and practices. What may have been somewhat surprising is that DL attained significantly higher maximum velocities than the OL group (15.6 vs. 11.3 km·h−1). Presumably, ball pursuit warrants greater velocities by defenders than offenders. In addition, OL have scripted step counts or blocking schemes that limit their opportunities to get into open-field sprinting.

Inertial movement analysis is the number of times the athlete accelerated, decelerated, and changed directions at high intensities (>3.5 m·s−2). Defensive lineman and OL had significantly more such activity than DL and OL than DBs or WRs, measuring an average of 67.8 episodes in comparison with 37.7 episodes registered by DBs and WRs. It is reasonable to suggest that the discrepancy between DL and skill players is because DL positions require more frequent changes of direction because of attacking the opponent in pass rush situations, maneuvering around a blocker, pursuing the football carrier, or a scrambling quarterback. Similarly, OL during running plays drive forward or at an angle to reach the opponent and subsequently attempt to block downfield. Furthermore, during pass protection, the OL must mirror the defender lending to frequent changes in direction. Wide receivers, however, have preset routes consisting of little directional change while DBs react to cover the WR by following the same, largely vertical patterns.

For acceleration band 7 (2–3 m·s−2), DB and WR recorded significantly greater distances than DL and OL. There was no significant difference in distance traveled in band 7 between DB and WR (p = 0.07), but a significant difference was found between DL and OL with DL traveling 43% further than OL (111.6 m vs. 63.4 m). For band 8, WR accelerated significantly further than all the other positions and DB accelerated significantly further than DL and OL, and there was no significant difference between OL and DL (p = 0.08). There was a drop-off of total distance in acceleration in band 8 in comparison with B and 7. This is an indication that the athletes cover more distance in accelerating at moderate intensities than high intensities. The significantly greater distance generated by WR in high-intensity acceleration may be due to the fact that for each play, receivers are taught to “explode” off the line, so the defender cannot read pass or run in the initial seconds. Certainly if the DB is playing “off” the WR, no initial high acceleration is warranted.

Deceleration and acceleration were measured as the total distance that was accrued in each band (moderate and high). Deceleration band 1 (2–3 m·s−2) recordings showed that WR and DB traveled significantly further when compared with OL and DL. Similarly, deceleration band 2 (>3 m·s−2) indicated that WR and DB traveled further than OL and DL. These results agree with others (11,13), in that skill players decelerate at greater distances in both moderate and high intensities than linemen. These data verify that skill players' physical demand in deceleration is greater than in linemen presumably because they generate greater velocities that mandate more forceful deceleration to change direction and to anticipate offensive movement.

From these data, it is clear that all positions throughout a game spend more time accelerating than decelerating. Specifically, the OL and DL groups showed substantial increases from deceleration to acceleration at moderate intensities, with OL still recording the least amount of yardage. It is plausible that the distance generated by the DL could be that when the ball is put in play, the goal of this position is to accelerate at a high rate of speed out of a 3-point stance to create angles and beat the OL to get to the quarterback or running back. In doing this, they accumulate more accelerating distance than OL. By contrast, the job of the OL is to stop or decelerate the DL from penetrating and getting to the ball, which may explain why the OL accelerated less than the DL.

Back to Top | Article Outline

Practical Applications

Findings from this study could potentially determine how strength and conditioning professionals, coaches, and trainers develop training, conditioning, and practice protocols rather than adhering to a “one size fits all” program. Also, coaches may use this information to develop game substitution guidelines to rest players in those positions that are involved in greater intensities and longer running distances. Indeed, awareness of the players' volume of exertion through the GPS data may be a means to become more aware of players' fatigue. With such information, decisions can be made for substitution, thus eliminating one factor related to injury (6).

Results indicated that skill (DB and WR) position groups travel further, average higher top speeds, and accrue more high-intensity acceleration and deceleration throughout a game. From a strength and conditioning perspective, a special focus should be placed upon developing a specific program for these select athletes. Taking into account, their overall workload and amount of high-intensity efforts throughout a competitive game and season, specialized plyometric, sprint training, and aerobic conditioning can be used to better prepare them for play. For linemen (DL and OL), distance, velocity, and acceleration and deceleration were lower than skill position; however, the body mass of linemen is significantly greater than that of skill players. A study conducted by Jacobson (8) traced Division I football players' anthropometric values for over 50 years found that, in 2010, OL averaged 133.5 kg and DL averaged 131.6 kg in body mass. With such large individuals, distance covered at rapid velocities in addition to acceleration and deceleration requires much more effort than for lighter players. Underestimation of workload during games could be a factor for the overuse and soft-tissue injuries (9) and more serious injuries. Therefore, athletic trainers may benefit from exertion data to install preventive measures.

Back to Top | Article Outline

Acknowledgments

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this manuscript.

Back to Top | Article Outline

References

1. Austin DJ, Kelly SJ. Professional rugby league positional match-play analysis through the use of global positioning system. J Strength Cond Res 28: 187–193, 2013.
2. Aughey RJ. Applications of GPS technologies in the field. Int J Sports Physiol Perform 6: 295–310, 2011.
3. Barr M, Beaver T, Turczyn D, Cornish S. The validity and reliability of 15 Hz global positioning system unites for assessing the activity profiles of university football players. J Strength Cond Res 2017. Epub ahead of print.
4. Boyd LJ, Ball K, Aughey RJ. Quantifying external load in Australian football matches and training using accelerometers. Int J Sports Physiol Perform 8: 44–51, 2013.
5. DeMartini JK, Martschinske JL, Casa DJ, Lopez RM, Ganio MS, Walz SM, et al. Physical demands of National Collegiate Athletic Association Division I football players during preseason training in the heat. J Strength Cond Res 25: 2935–2943, 2011.
6. Fuller C, Taylor A, Raftery M. Should player fatigue be the focus of injury prevention strategies for international rugby sevens tournaments? Br J Sports Med 50: 682–687, 2016.
7. Gabbett TJ, Ullah S. Relationship between running loads and soft-tissue injury in elite team sport athletes. J Strength Cond Res 26: 953–960, 2012.
8. Jacobson BH. Anthropometric cross-sectional comparisons of college football players and potential health implications. J Strength Cond Res 26: 3358–3364, 2012.
9. Johnston RJ, Watsford ML, Kelly SJ, Pine MJ, Spurrs RW. Validity and interunit reliability of 10 Hz and 15 Hz GPS units for assessing athlete movement demands. J Strength Cond Res 28: 1649–1655, 2014.
10. Rock AS. PhD Football: An Ivory Tower Take on NFL Analytics. Available at: http://phdfootball.blogspot.com/2014/02/do-defenses-get-tired.html. Accessed: February 3, 2014.
11. Wellman AD, Sam CC, Grant CG, McLellan CP. Quantification of competitive game demands of NCAA Division I college football players using global positioning systems. J Strength Cond Res 30: 11–19, 2016.
12. Wellman A, Coad S, McLellean C, Goulet G. Quantification of accelerometer derived impacts associated with competitive games in national collegiate athletic association Division I college football players. J Strength Cond Res 31: 330–338, 2017.
13. Wellman AD, Coad SC, Flynn PJ, Climstein M, McLellan CP. Movement demands and perceived wellness associated with preseason training camp in NCAA Division I college football players. J Strength Cond Res 31: 2704–2718, 2017.
14. Wundersitz DWT, Netto KJ, Aisbett B, Gastin PB. Validity of an upper body-mounted accelerometer to measure peak vertical and resultant force during running and change-of-direction tasks. Sports Biomech 12: 403–412, 2013.
Keywords:

tracking; acceleration; performance; intensity; competition

Copyright © 2019 by the National Strength & Conditioning Association.