In-Season Match Demands Of Men’s Collegiate Soccer: A Comparison By Half, Position, Match Outcome, Match Location, And Competition Phase

The purpose of this study was to quantify athlete external workload by half, position, match outcome, match location, and competition phase (e.g., conference vs non-conference) during match play across a men’s NCAA DIII soccer season. Throughout the competitive season, 16 soccer players wore a GPS device in 17 matches. Workload metrics collected were: total distance (TD), distance per minute (D/min), distance in speed zones (SZ) 1-5, sprint efforts, sprint distance (SD), top speed, accelerations, player load (PL), and player load per minute (PL/min). TD (4164±1235 m), PL (169±52 AU) D/min (116±20 m/min), PL/min (4.7±0.8 AU), SD (80±55 m), accelerations (32±14), decelerations


INTRODUCTION
Quantification of physical work performed during training and competition, commonly referred to as external load, is important to consider when designing and implementing programs for athletes. Global positioning systems (GPS) are a viable tool for monitoring and managing athlete loads in order to minimize injury risk and improve sport performance (Bourdon et al., 2017). Further, monitoring athlete external load can be used to manipulate volumes and intensities, inform coaching decisions, optimize recovery, and guide nutritional interventions (Jagim et al., 2020). Soccer is an intermittent, high-intensity sport, in which players are exposed to high volumes (e.g., total distance, player load) and intensities (e.g., high-speed running, sprints, jumps, change of direction, accelerations, and decelerations). External loads achieved during match play may vary depending upon a multitude of factors including: half, playing position, match outcome, match location, competition phase (e.g., conference vs non-conference), score margin, tactical objectives, pacing strategies, and playing time. National Collegiate Athletic Association (NCAA) soccer rules and seasonality structure differ from other levels of competition, therefore, match loads may differ from those reported at the professional level (Jagim et al., 2020). For example, NCAA soccer rules allow for reentry following substitutions, include two 10-minute overtime periods with a "golden goal" applied, and incorporate clock stoppage for injuries, goals, and card issuance (Andres, 2021). There are also differences in regards to in-season scheduling, where NCAA soccer players can compete in over 25 matches during a 15-week season, with 2 matches per week, compared to European professional soccer players who may play multiple matches per week over a 45-week season (Carling et  Previous research in men's professional soccer has demonstrated that physical performance declines in the second half of matches, specifically total distance and high-speed distance (Mohr et al., 2003(Mohr et al., , 2005. These observed performance decrements may be due to a multitude of physiological changes that occur over the course of a match, including glycogen depletion, increased core temperatures, dehydration (Mohr et al., 2005), pacing strategies, tactical changes, and mental fatigue (Bradley & Noakes, 2013;Paul et al., 2015). The reduction in load across half may also be attributed to score discrepancies and match outcome, such that loads may decrease when teams are winning and increase when teams are losing (Lago-Peñas, 2012). There are also known differences in positional demands as prior research has reported wide defenders and strikers produce the greatest high-speed running, sprinting, and high-intensity acceleration distances, compared to other positions (Abbott et al., 2018;Andrzejewski et al., 2015;Bloomfield et al., 2007;Di Salvo et al., 2007); however, positional workloads may also be affected by tactical formation (Calder & Gabbett, 2022). Match location may also influence external loads, as prior research has shown greater intensity efforts at home matches (Lago-Peñas, 2012;Oliva-Lozano et al., 2021). Lastly, loads may vary across competition phase, with a potential for higher volumes and intensities in conference matches as the level of competition may be greater with more at stake in regard to the outcome. Of importance, match outputs have been observed across a variety of professional levels, and results have shown that lower divisions often covered greater workloads, most likely due to lack of technical and tactical qualities compared to higher divisions . Therefore, it is important to quantify match loads throughout a season across distinct levels of NCAA competition, as a collegiate season may provide a unique distribution and magnitude of external loads. In turn, this can help direct the specific programming and recovery needs for male collegiate soccer athletes.
While previous work has quantified match demands of NCAA Division I men's and women's soccer, Curtis et al., 2018;McFadden et al., 2020), match demands of NCAA Division III (DIII) have not been established despite DIII totaling 410 of the 821 collegiate soccer programs in the United States. Therefore, the purpose of this study was to quantify the athlete external workload by half, position, match outcome, match location, and competition phase (e.g., conference vs nonconference) during match play across a men's NCAA DIII soccer season.

Participants
NCAA DIII men's soccer players (n = 16, age range: 18-21 years) classified as "starters", participated. Starters were defined as players who maintained a minimum playing time of 45 minutes per match. Goalkeepers and non-starters were excluded due to relatively low total distances travelled. Soccer athletes were under the direction of the same Certified Strength and Conditioning Specialist ® and were following a similar training regimen. All athletes completed a medical history form and were cleared for intercollegiate athletic participation. Risks and benefits were explained to athletes, and an institutionally approved written informed consent form was signed before participation. All procedures involving human subjects were conducted in accordance with the requirements of the Declaration of Helsinki and approved by the Springfield College Institutional Review Board for Human Subjects (IRB #3182021).

Procedures
Athlete external loads were collected over 10 weeks during the 2021 NCAA men's soccer season from "starters." External loads were collected during all in-season matches (n = 17). Information pertaining to match location, outcome and conference status was also recorded and used for later analysis.

External Load
External load was quantified during all matches using 10  External load metrics collected were: total distance (TD) (m), distance per minute (D/min) (m/min), distance in speed zones 1 (SZ1: 0-30% max speed), 2 (SZ2: 30-50% max speed), 3 (SZ3: 50-75% max speed), 4 (SZ4: 75-90% max speed,) and 5 (SZ5: > 90% max speed ), sprint distance (SD) (> 5 m/s1), top speed (m/s1), acceleration efforts (> 3 m/s2), deceleration efforts ( # > -3 m/s2) player load (PL) (AU) which is calculated as ∑√(instantaneous rate of change in acceleration in all 3 orthogonal planes)), and player load per minute (PL/min). Player load has been shown to be a valid and reliable measurement of total volume accrued during soccer training (Barrett et al., 2014). Additionally, maximal speed was determined for each player during preseason fitness tests and was continuously adjusted throughout the season if a player achieved a new higher speed. The use of individualized speed zones has been shown to provide more useful information regarding player velocity, especially when comparing different playing positions (Sánchez et al., 2017). Additionally, the use of individualized speed zones may be more useful when comparing higher speed zones (SZ4 and SZ5) across playing levels to modify zones based on physical abilites (Bradley & Vescovi, 2015).

Statistical Analysis
SPSS version 25.0 (IBM, Armonk, NY) was used for summary statistics. All values are presented as means ± SDs. Normality was assessed and nonnormally distributed variables were log transformed for subsequent analyses. A multivariate analysis of variance (MANOVA) assessed differences in external load measures across halves, sportposition, match outcome, match location, and conference status (e.g., in-conference opponent, out-of-conference opponent) (p < 0.05). Bonferroni post hoc comparisons were calculated when a significant effect was identified. Partial eta 2 (η 2 ) effect sizes were calculated and interpreted as follows: small: 0.01-0.06; moderate: 0.06-0.14; and large: > 0.14.

RESULTS
A summary of external loads by half, sport-position, match outcome, match location, and conference status are presented in Table 1.

Match Outcome
Differences in workload based on match outcomes are displayed in Table 1 0.053), SZ4 (p = 0.006, partial η 2 = 0.049) were higher in matches that resulted in wins. However, SZ5 (p < 0.001, partial η 2 = 0.117) was higher in matches that resulted in losses. No significant differences based on match outcome existed in TD, SD, accelerations, decelerations, PP, SZ1, and SZ2.

DISCUSSION
This is the first study to examine match loads in NCAA DIII men's soccer throughout an entire competitive season. The goals of the current study were to provide descriptive and quantifiable information about the physical match loads experienced by NCAA DIII men's soccer players and how they may differ by half, playing position, match location, match outcome, and competition phase. The main findings were that external loads differed for all of the aforementioned parameters. In the current study, TD (+11.5%), PL (+10.7%), D/min (+16.4%), PL/min (+10.6%), accelerations (+22.1%), decelerations (+20%), and distances in SZ2 (+17.2%) and SZ3 (+12%) were higher in the first half of match play (See Table 1 Different activity profiles were evident among playing-position. Forwards (9.5 ± 2 m/s) recorded higher top speeds than midfielders (8.0 ± 1.1 m/s) and defenders (7.7 ± 1.2 m/s) but no differences were observed in the number of sprint efforts or sprint distance across position (See Table 1). These findings differ from previous research in professional soccer players, which reported that forwards are commonly involved in more high-speed running and sprinting activities during match play than other positions The distinction in positional demands may be attributable to the fact that forwards are required to produce higher speeds in their attempts to win balls and create distance between themselves and the defenders. Additionally, differences in external loads per position may fluctuate depending upon decisions made by players, team dynamics, playing time, or tactical strategies and formations employed by the coach. For instance, if a player is out of position, other players may have to work harder until that player recovers (Dalen et al., 2020). Tactical formation has also been shown to alter positional workloads, as workloads may change based on the configuration of defenders, midfielders, and forwards due to space allotted for each player to cover while attacking, defending, and transitioning (Calder & Gabbett, 2022). Another important consideration is that tactical assignments may vary from player to player, even within the same position from match to match . While positional data at the collegiate level remains limited, one study investigating five NCAA DI men's soccer teams (n=107) examined positional differences across a full-season and reported no differences in TD or high-speed running (Curtis et al., 2020). Interestingly, in a women's DIII soccer team, forwards covered some of the lowest volumes and intensities when compared to other positions (Jagim et al., 2020). Therefore, it is important to continue to examine differences in match demands across each level of play and between the men's and women's divisions in order to characterize the demands of each position. Additionally, positional classification (i.e., central vs wide players) has also shown to influence the workloads observed between positions (Abbott et al., 2018;Schuth et al., 2016). For example, it has been reported in elite men players that wide defenders tend to cover more TD (11,410  708 m) and sprint distance (402165 m) than central defenders (TD: 10,627  893 m; sprint distance: 215100 m (Di Salvo et al., 2007). However, due to the smaller sample size used in the current study, these positional classifications could not be analyzed, but should be considered for future research. Classifying players to these positional groupings may pose a challenge at the collegiate level, as players may re-enter the match into a different position after being substituted (Altmann et al., 2021). Establishing position-specific competition demands will further allow coaches to specialize training sessions to fulfill players' physical needs based on their match demands.
The results from the current study demonstrate that PL (309. 45 Table 1). Limited attention has been paid to quantifying differences in loads by match outcome, but preliminary findings have demonstrated that professional soccer players perform significantly fewer high-intensity movement patterns during wins compared to losses (Bloomfield et al., 2005Lago et al. 2010). This phenomenon suggests that players may assume a ball retention strategy when winning, resulting in a slower pace of play with attenuated speeds (Bloomfield et al., 2005Lago et al. 2010). On the other hand, when losing, players may try to increase their physical outputs to gain ball possession to improve the likelihood of scoring. Other research found higher external loads in wins, but higher TD and high-speeds running distances in the second half of losses. (Nobari et al., 2021). This may be attributed to a closer score, which may be associated with various motivational implications, amongst other variables.
Match location appeared to influence differences in external loads as current results indicated PL  Table 1). Although data in regard to match location and external loads remain limited, current findings are in support of a prior study showing professional men's soccer players covered greater TD at home matches. (Lago et al., 2009;Zubillaga et al., 2007). Home advantage in soccer is well-known and players may have taken advantage of the familiar crowd, playing surface, absence of travel, pride, and other psychological factors that may result in greater effort and more movement (Pollard, 2008 Table 1). These findings align with previous research in NCAA DI women's soccer players, where non-conference matches elicited greater training loads, TDs, and energy expenditures relative to conference play . Of important note, the first four matches of the season were played against non-conference opponents; therefore, it is likely players produced greater workloads because they had not experienced much accumulated fatigue from the upcoming season (Gualtieri et al., 2020). Another explanation for the greater workloads observed in non-conference play man be attributed to the higher frequency of player substitutions as players are competing for a starting spot on the roster . This aggressive play coupled with the potential for higher quality opponents, could potentially lead to the increased external loads observed in non-conference matches. Further exploration of such differences is warranted to ensure athletes are balancing progressive overload and recovery during accumulating in-season demands and conference play .
This study does not come without limitations. First, data was collected from one NCAA Division III men's team in the northeast region and therefore may not be comparable to teams in other divisions or regions. Further, the small sample size (n=16) prevented additional positional classifications (i.e., central vs wide defenders). It's also important to note that tactical decisions including formation and substitutions may have also influenced match demands. Lastly, the current study used Playertek GPS systems, which may differ from other GPS systems in its satellite recruitment and filtering threshold. This is the first examination of external workload match demands in DIII men's soccer players over the course of an entire season. Our results indicate that external loads were affected by half, position, match outcome, match location, and competition phase. Further research exploring interactions between contextual factors affecting external loads in competitive soccer is warranted. Such contextual variables, which result in changes in workload and performance, may be considered for more appropriate load prescriptions and improved programming for load management and periodization strategies during congested match schedules. Additionally, future research should consider examining accumulated fatigue and fluctuations in workload during training leading up to matches to optimize training without hindering match performance. Coaches may use this information to identify key performance indicators and tailor practice activities that will maximize their technical and physical performance for matches.