According to the PRISMA statement recommendations, a systematic search was undertaken on the electronic databases of PubMed, Web of Knowledge (all databases), and Scopus (20). A search of relevant articles published between January 2011 and July 2023 was conducted using the keywords [“global position system” OR GPS OR “local position system” OR LPS) AND (“team sports*,” Indoor team sports**”) AND (performance* OR “external load” OR “internal load”)]. The publications that were retrieved had to meet the following requirements: (1) should provide relevant statistics and research strategies about elite athletes' performance; (2) athletes' EL and IL; (3) published in English; and (4) should be only about team sports. Exclusion criteria applied if they (1) specifically pertained to the reliability, validity, or precision of global positioning system equipment; (2) pertained to systematic reviews, (3) were published before 2011.

To improve research accuracy, two reviewers (AF and BT) independently screened titles and abstracts to identify articles that would potentially combine the inclusion criteria after noting the features of each study, such as the authors' names, sample, procedure, and results or main outcomes. In the event of disagreement about the article's eligibility, a third reviewer (HS) was brought in to make a final judgment. Following a review of all papers, the study categories were divided into multiple sports depending on the primary research topic.

The quality of the studies was assessed in accordance with what is indicated by Faber et al. (21) and using the criteria for critical review forms recommended by Law et al. (22) (16 items) with the goal of finding studies in which low-quality scores could interfere with the results. Both researchers (AF and BT) separately assessed the quality of each eligible article. The interrater agreement was moderate (k = 0.43) (23). The possible criteria for each item were 1 (meets criteria), 0 (does not meet the criteria), or NA (not applicable). According to the methodology assessment, articles were evaluated with regard to their purpose (item 1), relevance of background literature (item 2), appropriateness of the study design (item 3), sample included (items 4 and 5), informed consent procedure (item 6), outcome measures (item 7), validity of measures (item 8), significance of results (item 9), details of intervention (item 10), analysis (item 11), clinical importance (item 12), description of drop-outs (item 13), conclusion (item 14), practical implications (item 15), and limitation (item 16). According to the guidelines of Faber et al. (21), a final score was determined, allowing the publications to be classified as (1) low methodological quality (50%), (2) good methodology quality (50%–75%), or (3) exceptional methodology quality (>75%). A data extraction sheet (based on the Cochrane Consumers and Communication Review Group's data extraction template) (24) was adjusted to meet the research inclusion criteria for this analysis and then evaluated on 10 randomly chosen studies (pilot test). The data were extracted by one author and validated by another. Disagreements were settled by discussions between these two authors (first and last). To organize the findings, the studies were divided into categories based on the major study issues.

The first search generated 772 titles. These data were then exported to Mendeley (Elsevier, San Francisco, CA, USA), a reference manager software. Duplicates (n = 430) were either automatically or manually removed. The remaining 342 titles were then evaluated for relevance using information from their title and abstract, with 164 being eliminated. The remaining 178 papers were examined further, and 99 were excluded based on the following criteria: (1) articles published in languages other than English (n = 5); (2) articles unrelated to team sports (n = 4); (3) articles unrelated to elite athletes (n = 10); (4) systematic reviews (n = 14); (5) articles unrelated to the analysis of athletes' performance by technology (n = 22); and (6) articles concerning the reliability and validation of tracking technology (n = 44). The review comprised 79 studies in total (Figure 1).

To understand the scope of the development of research methodologies in physical performance evaluation in team sports during the last decade, the publication target years were restricted from January 2011 to July 2023. The articles used for in-depth reading revealed that there had been an increase in GPS and LPS research. In this review, 84.8% of the studies were published in the last 5 years (from 2017 to 2023) and 15.2% between 2011 and 2016. On average, 6.3 studies were published per year.

With regard to the quality of the studies, following the proposal of Faber et al. (21), the average score of the 79 selected studies was 89.2%. Fourteen studies achieved 100%; 52 achieved an average rating of ≥75%; 12 achieved a score between 50% and 75%.

Studies were categorized by the sport in which the studies were carried out: (1) football (FTB) (41.8%), (2) futsal (FTS) (15.2%), (3) rugby (RG) (11.4%), (4) Australian football (AUF) (7.6%), (5) field hockey (FH) (6.3%), (6) Gaelic football (GF) (6.3%), (7) basketball (BK) (3.8%), (8) American football (AMF) (2.5%), (9) ice hockey (IH) (2.5%), and (10) rink hockey (RH) (2.5%).

The studies were classified based on the criteria used in the study. Most of the team sports research employed tracking technology: (1) combined kinematic and mechanical variables (35.4%); (2) only considered kinematics (21.0%); and (3) included kinematic, mechanical, and metabolic variables (16.5%). On the other hand, the analysis of IL or the relationship between mechanical variables and IL variables was less explored (Figure 2).

Studies were also grouped according to the type of analysis of the studies. The results revealed that the studies were distributed by using the following analyses: (1) characterization and comparison of drills, training, and matches sessions (62.0%), (2) correlation of variables to understand their relationship in training and competition (26.6%), (3) predicting performance, fatigue, and injuries (8.7%), and finally, (4) experimental studies to determine greater training and team management approach (2.7%) (Figure 3).

With regard to the main research topics from the studies in analyses, the following was observed: (1) performance analysis in match demands, (2) performance analysis of training vs. match demands, (3) injuries, and (4) nutrition (Figure 4).

The studies reviewed provide information on the use of tracking systems in various domains of sports sciences, such as the importance of technology in better understanding the physical requirements for coadaptations between match demand information in which players establish interactions with teammates and opponents in order to manipulate training sessions and, as a result, to improve athletes' awareness and physical performance (25, 26). In addition, studies were used to determine whether specialized nutritional intake could improve players' responses to the demands of an indoor team sport (27, 28). Also, in recent years, multiple sports have used various methods to track injuries by understanding the external information that better characterizes sporting activities (7, 29, 30).

With regard to the understanding of training and match demands in the performance and analysis of team sports, several studies have been carried out in order to compare and understand positional (31), tactical, and technical differences (32), intragame information (33, 34), comparisons of the levels of competitions (35), predictors of intensity, fatigue, and match performance (36–39), and more recently, to understand the games’ most high-intensity activities (HIAs) and periods (MIPs) (36, 40) (Figure 4).

With regard to the performance analysis of match demands, there are several studies involving the characterization and understanding of EL metrics in various team sports (32, 41–46) as well as its impact on athletes’ IL (36, 47–51). Some studies describe a moderate to high correlation observed between EL and IL (38, 48). In addition, there are studies using tracking systems in which the main goal is to find correlations between metrics; (1) kinematic vs. mechanical (36, 43, 52); (2) kinematic vs. mechanical and metabolic (37, 41).

A study with 118 elite male ruby league players shows significant increases in the offensive and defensive tackles between senior and junior players and between forwards and backs (53). In the same study, independent of the playing level, running demands (HSR, ACC, DEC) were significantly greater in backs than in forwards (53). There is a high prevalence of studies in which EL metrics are classified based on player level and player position; positioning is often related to an event or to the magnitude of EL metrics (33, 46, 53–56). However, positioning associations are not always found in team sports. In a study of nine elite RH players, there were no significant differences throughout all observed metrics during a competition between players' positions (57). Furthermore, variable load distribution (total PL) based on the axis of movement was not verified based on playing position (58). However, total PL appears suitable for monitoring locomotor demands or accelerometer-derived load in FTB (58).

Finally, and more recently, studies seem to aim for the prediction and proposal of models of fatigue (59) and match performance according to match demand information (44, 60–63) (Table 1).

Based on an analysis of the selected studies, at the same level, studies suggest that it is possible to predict IL such as RPE from GPS training and match data (59), showing that the variation in workload efficiency throughout matches may be explained by the training load variables from the prior days and training cycles (60). In addition, an understanding of load dynamics across training and competition is presented as an important aspect related to performance enhancement in team sports (8, 47, 60, 61), showing, for example, how adjusting specific drills such as large,- medium- and small-sided games may help replicate game demands (86, 87). Correlations between EL and IL in different season periods were applied in order to understand the fluctuation of some metrics such as players’ strength (88), players’ recovery (89), or ACWR, between players’ positions, starters, and non-starters (82, 84, 90, 91). Also, Pizarro et al. (26) demonstrated how different occupational areas may promote different physical variables. The authors state how the manipulation of “floaters” in SSGs according to specific purposes in FTS promotes higher distances and speed variables, accelerations, and decelerations. The same was described by De Jong et al. (81), in which manipulation of the game momentum (with the outfield goalkeeper) promotes changes in team total distances covered (81). Silva et al. (92) demonstrated that teams have the ability to coadapt to performance constraints by manipulating SSGs. In addition, add-ons to SSGs, such as the number of players by drill, player positioning, or touch constraints, are being referenced as potential ways to increase or decrease EL (81, 93) (Table 2).

With regard to the analysis related to injury prediction, some kinematic variables such as distance covered and mechanical ones such as accelerations/decelerations were considered. For example, in elite female FH players, the authors verified that the risk for in-game knee injury decreased with the increasing distance covered while running at low speed (30). In the same path, Caparrós et al. (29), in a total of 2,613 observations and 246 games from 33 different players of a professional male BK team, described that athletes with less or equal than three decelerations per game and those running less or equal than 1.3 miles per game had a higher risk of injury. Predictive injury models are also proposed by the use of tracking systems (7). In this study with 26 Italian professional male players, the authors propose a multidimensional approach to injury forecasting in professional FTB based on GPS measurements (7). In addition, three studies from contact team sports (AUF and AMF) presented associations between specific levels of kinematic and mechanical variables with the occurrence of injuries (11, 99). Finally, some studies reported that lower workloads during preseason and higher workloads during competition training weeks increased the risk of non-contact injuries (100; Nobari et al., n.d.) (Table 3).

Two studies that analyzed the variation in physical enhancement through nutritional intake were identified. Before and after the players received 140 mL of beetroot juice (BJ) or placebo (PLA) on two separate days, a sample of 10 competitive male BK players was monitored (27). To determine physical improvements, a specific neuromuscular battery of tests followed by a 40 min-simulated BK match was used to measure the players’ running profile, ACC, and DEC; however, acute moderate doses of BJ (12.8 mmol of NO3-) were not effective in improving neuromuscular performance (27). In the other study, the same author verified that the ingestion of moderate doses of caffeine enhanced several physical variables such as jump height, running profile, number of body impacts, ACC, and DEC in FTS professional players (28) (Table 4).

In team sports, coaches and performance personnel invest a lot of time and resources to follow up on their athletes. Physical training can be described in terms of EL (work) and acute responses and long-term adaptations (IL) (104). Through the development of wearable technology for athletes over the past few years, team sports stakeholders have embraced the practice of monitoring their athletes' EL in both training and competition (54). Consequently, technology is used by coaches and exercise physiologists with the intent of enhancing training data and achieving competitive advantage (54). The focus on athlete performance improvement has been driving the need to understand training and match demands. Several researchers have been characterizing and understanding EL metrics (mainly kinematic and mechanical) during specific training or competitive contexts (32, 59). In addition, and as an attempt to enhance knowledge for training improvement, links have been established between kinematic, mechanical, and metabolic variables, and more recently with IL (86, 87). However, and assuming the physiological adaptations of different team sports exigencies, we suggest that stronger steps must be taken based on experimental approaches seeking to better understand the relationship between EL and IL and athletes’ recovery cycles (50, 53). In this regard, we suggest that studying Large-, Medium-, and Small-sided games (LSGs, MSGs, and SSGs) may be helpful in determining which format is most suited to load specific kinematic or mechanical loads in line with a desired IL (86, 87). Positioning, intragames, level of competition, and tactical comparisons have also been pointed out as references for the understanding of how performance may be improved by better understanding EL metrics, particularly in terms of frequency, duration, and intensity. In addition to the difficulties of collecting data at a top-elite level, in order to create a training program that accurately reflects the physiological demands of the game, we suggest that more elite team sports game-specific data are required. In indoor high-intermittent team sports and based on the metrics selected by Ohmuro et al. (75), in order to understand more accurately the demands of games, we suggest a monitoring of EL by effective playing time (in-game). Despite the fact that an agreement between coaches and athletes on overall RPE and sRPE, as well as RPE and sRPE, in moderate and hard categories, has already been recorded (13), there is a lack of information about how demanding is the playing time, by players’ interchange rotation, particularly in indoor intermittent team sports, and how can that information benefit the training structure, and consequently, team performance.

Team sports performance is regarded as highly complex and multivariate. Researchers integrate players' technique with EL and IL, while there are studies in which EL and IL are categorized by players’ position, exploring the differences between training and competition (73, 75, 98). This correlational approach seems to translate differences that are assessed for a better individual and collective preparation, as reported by Florian et al. (105),. When a selected EL fits with sports specification, we suggest that this type of method is reliable for better interpretation of competitive microcycles. As proposed by Ispyrlidis et al. (98), in FTB, and verified by Fernández et al. (47), in RH, this information may combine the interchanges of matches with competitive profile training and improve all parameters of players’ physical output. At the same level, research indicates that an understanding of load dynamics during training and competition, such as players' intra- and interweek fluctuations, exercises a substantial influence on improving elite athletes’ performance by properly regulating workload and recovery (82). In this review, it is found that several researchers in their studies have developed a longitudinal understanding of load dynamic (combined kinematic, mechanical, and metabolic metrics) across training and competition, proposing that performance enhancement may be a reflection of a proper adjustment of specific metrics (46, 67).

More recently, performance analysis research is stepping forward on the understanding of HIAs and on predictive studies for fatigue and match performance. The literature suggests that during different phases of a training cycle, such as the baseline or competition phase, training loads must be modulated to either enhance or reduce fatigue levels (89). In the light of this, it is important to understand the match HIA in order to manage fatigue properly, matching both variables to adaptation to training and for achieving competitive performance (10, 40). On the other hand, understanding HIAs and their impacts on IL also improves the knowledge of individual players’ recovery strategies (89). The importance of understanding effort categories (moderate and hard) has already been established in a systematic review, in which agreements between coaches and athletes about total RPE and s_RPE were reported (13). Therefore, in order to effectively adjust the volume, frequency, and intensity of the metrics that are more closely associated with game pick periods, this information can assist in developing novel approaches on tapering training strategies. In this sense, kinematic and mechanical variables are statistically manipulated with indirect IL such as RPE, suggesting that it is possible to predict fatigue from GPS/LPS training and match data. In addition, a dissociation between EL and IL units may reveal the state of fatigue of an athlete; therefore, detecting changes with statistical approaches may provide the necessary confidence for implementing changes (10). Nevertheless, in contact team sports such as AUF, AMF, RG, and GF, the use of mechanical metrics seems to be more significant to understanding players’ fatigue and muscle damage (32, 39, 43, 71).

Strength and conditioning coaches may daily obtain reports with more than 100 variables (106). Thus, it is crucial to choose and identify the key performance indicators according to sports specifications (63). In this review, we identified research focused on proposing models for match performance mainly based on kinematic, mechanical, and metabolic variables (44, 60–62, 99). These studies demonstrate how crucial it is to adopt statistical techniques like PCA in team sports to accurately use significant performance variables by excluding the least significant ones.

Athletes compete fiercely to become the greatest in their respective sports, particularly at the highest level. In team sports, the individual or collective success of this process is often compromised by the occurrence of injuries (30). Therefore, in this review, we observed that the development of injury prevention programs is being supported by the association of tracking positioning systems such as the use of a GPS. Recently, studies tried to understand which (EL metrics) and how (volume, intensity, and frequency) training and game demands have a higher association with the occurrence of specific injuries such as the risk for in-game knee injury (30). In specific intermittent team sports such as FH and BK, low kinematics (low-intensity running) (30) and mechanical (DEC) are associated with a higher risk of injury (29). However these studies do not explain whether this lower EL results from acute or chronic adaptations, such as the one by Li et al. (11), which applied the use of technology to understand the relationship between soft tissue injury and training workload in AMF. For this author, the variation of training stimulus may induce a higher risk of injuries. Once studies have shown that the ACWR is a more accurate predictor of injury than the total workload (18, 101) it is highly advised to use tracking technology to determine an athlete's chronic workload. According to the screened manuscript (7, 11), we suggest that the assessment of an athlete's physical activity based on the intensity, duration, and number of actions of kinematic, mechanical, and metabolic metrics may also predict injuries with acceptable precision. Recently, a higher preseason workload, such as a running performance, was related to a lower risk of injuries (100). On the other hand, higher levels of weekly chronic workload (ACWR) indicate greater chances of sustaining non-contact injuries (84, 91). We suggest that these dichotomic reports may support the need for better longitudinal use of tracking systems in different team sports, in order to establish causal correlations between specific parameters, and consequently, to improve injury prevention measures. The sparse research on the use of tracking technology related to injury forecast is mainly noticed in outdoor team sports; thus, it is recommended that future studies extend to high-intermittent indoor team sports such as FTS, handball (HB), RH, and BK. Finally, there is no injury forecast method that links the most demanding scenarios (MDS) or HIAs in team sports according to specific EL variables and IL information. We suggest that characterizing the dynamic of training workload related to teams' sports MDS may contribute to more reliable specific injury predictions.

The use of nutritional supplements and ergogenic aids to improve performance is becoming more popular among intermittent sports (107), but few have demonstrated benefits (108) such as creatine, caffeine, sodium bicarbonate, and beta-alanine. Indoor team sports (BK, FTS, HB, RH, or netball) are becoming more intense, with explosive demands (37, 47) stronger athletes, and repeated combinations of short and long intermittent efforts (97). This review noticed almost no research on the use of tracking technology to measure nutritional impact in kinematic, mechanical, or metabolic metrics. In this review, only two studies were screened. With regard to the experimental approach developed to understand whether the nutritional supplement of beetroot juice had some impact on the monitored EL variables (27), no significant results were found. Furthermore, the effects of supplementation on physical performance and match running load or true exposure during a competition (total distance covered, speed achieved, number of accelerations and decelerations) in team sports are either sparse or unknown (27). In their recent work, López-Samanes et al. (28) demonstrated how the application of this technology could advance the experimental approach to athletes' nutritional development. The authors showed that 3 mg/kg of caffeine enhanced the physical variables associated with FTS, such as jump and sprint performance, and also improved kinematics (high-speed running) and mechanics (ACC/DEC) during a simulated FTS match. This finding can be interpreted as an opportunity for researchers not only to provide experimental proof of performance enhancement, but also to better understand the role of supplementary nutrition in an athlete's recovery process. To the best of our knowledge, there are no experimental studies aimed to understand athletes' recovery cycles with the influence of supplementary nutrition and monitored by tracking technology.

The fact that this review only included papers published in English in the electronic databases of PubMed, Web of Knowledge, and Scopus may have led to the exclusion of relevant publications in other languages. Furthermore, the exclusion of all systematic reviews can be seen as a constraint because some categories related to the aims of this scoping review may not have been identified because of this exclusion.

Currently, research on the use of tracking systems is mainly focused on the physical performance analysis on matches, training vs. matches, followed by injuries, and nutrition. Both kinematic and mechanical variables combined, and then only kinematics, appear to be the primary observed metrics. In addition, we verified that in some studies, IL information is combined solely with kinematic or mechanical metrics. Few studies combined both metrics. These manipulated strategies used for integrating IL information with EL may have been established according to team sports specifications. Still, we suggest that tracking and studying kinematic, mechanical, and IL parameters together in training and matches would provide more accurate data for the purpose of enhancing our understanding of athletes’ performance and, consequently, to better manipulate the structure of drills, for example, the manipulation of LSGs, MLGs, and SSGs, according to specific physiological responses.

Based on the main results screened in this review, further research and practice should consider how frequency, volume, and intensity may be adjusted according to the optimal (1) training structure; (2) to promote injury prevention; (3) to consider nutritional supplements; and (4) to promote physical performance analysis. The integration of these four dimensions has not yet been proposed. Consequently, how can performance analysis lead to the process of enhancing athletes' performance? This integrative vision proposal illustrates the significance of the domains explored in this review and how they may be oriented in accordance with specified research methodologies for both outdoor and indoor team sports (Figure 5). Despite the existing studies, we consider that the combined information of training and match demands remains the main source of the physical performance analyses. However, the integrative and longitudinal contribution of the presented domains related to the proposed researched strategies may help produce more reliable and accurate information, leading to an enhancement of the training structure. The development of more correlational analyses between metrics or prediction studies will improve the understanding of athletes’ performance enhancement. Also, experimental and prediction studies, particularly in the domain of injury prevention and nutrition, as well as in the manipulation of the metrics measured and compared between training and match sessions, are highly recommended. Therefore, because of its impact on high levels of muscle damage, as well as neuromuscular and perceptual fatigue, longitudinal tracking of EL variables, including kinematic, mechanical, and metabolic metrics, when combined with IL, can provide insights into injury prediction and prevention measures. In addition, the application of tracking technology to nutrition in team sports is relatively unexplored, with limited research investigating the impact of nutritional supplements on such metrics. This emphasizes the current trends in research, the need for more integrative models combining EL and IL data, and the importance of experimental and prediction studies to advance the understanding of team sports' physical performance, injury prevention, and nutritional impact. Ultimately, the integration of these domains into a coherent framework can guide both applied sports science research and practical interventions, leading to an improvement in athlete preparation and decision-making based on reliable data, which may promote the implementation of new drills, training, or competitive designs.