Methods of performance analysis in team invasion sports: A systematic review

Integration of machine learning XGBoost and SHAP models for NBA game outcome prediction and quantitative analysis methodology

This study investigated the application of artificial intelligence in real-time prediction of professional basketball games, identifying the variations within performance indicators that are critical in determining the outcomes of the games. Utilizing games data from the NBA seasons 2021 to 2023 as the sample, the study constructed a real-time predictive model for NBA game outcomes, integrating the machine learning XGBoost and SHAP algorithms. The model simulated the prediction of game outcomes at different time of games and effectively quantified the analysis of key factors that influenced game outcomes. The study's results demonstrated that the XGBoost algorithm was highly effective in predicting NBA game outcomes. Key performance indicators such as field goal percentage, defensive rebounds, and turnovers were consistently related to the outcomes at all times during the game. In the first half of the game, assists were a key indicator affecting the outcome of the game. In the second half of the games, offensive rebounds and three-point shooting percentage were key indicators affecting the outcome of the games. The performance of the real-time prediction model for NBA game outcomes, which integrates machine learning XGBoost and SHAP algorithms, is found to be excellent and highly interpretable. By quantifying the factors that determine victory, it is able to provide significant decision support for coaches in arranging tactical strategies on the court. Moreover, the study provides reliable data references for sports bettors, athletes, club managers, and sponsors.

Sequential Mapping of Game Patterns in Men and Women Professional Padel Players

PURPOSE

This study analyzed the sequences of actions in professional men and women padel players to identify common game patterns.

METHODS

The sample comprised 17,557 stroke-by-stroke actions (N = 1640 rallies) of the championship World Padel Tour. Multistep Markov chains were used to calculate the conditional probabilities of occurrence of actions during the rally.

RESULTS

Results revealed that men's and women's padel is mainly defined by 36 patterns constituting 55% and 63% of all actions in the game, respectively, with the 10 most common sequences accounting for 42% to 45% of the game. There were recurrent technical-tactical actions with specific offensive and defensive functions that were constantly reiterated during the rallies. In men, the use of smash, volley, bandeja, direct, back wall, back-wall lobs, and direct lobs followed a foreseeable pattern up to 8 lags, whereas women described predictable interactions for volley, bandeja, direct, lobs, and direct lobs up to 5 lags and for smash and back wall up to 4 lags.

CONCLUSIONS

The ability of padel players to recall these patterns and enhance their anticipation skills may potentially improve their performance. These findings contribute to a better knowledge of professional padel game dynamics while providing coaches and players with useful information to optimize training and decision-making strategies.

Exploration of collective tactical variables in elite netball: An analysis of team and sub-group positioning behaviours

Collective tactical behaviours are aspects of player interactions that are particularly important in netball, due to its unique restrictions on player movement (players unable to move when in possession of the ball and positional spatial restrictions). The aim of this study was to explore variables representing collective tactical behaviours in netball. A local positioning system provided player positions of one team throughout seven elite-level netball matches. The positions were analysed to provide mean, variability (standard deviation) and irregularity (normalised approximate entropy) for each attack and defence possession (470 and 423, respectively) for the team and positional subgroups (forwards, midcourts and defenders) for 10 position-related variables. Correlational analyses showed collective tactical variables could be grouped as lateral and longitudinal dispersion variables. The variables were each analysed after log transformation with a linear mixed model to compare attack and defence and to estimate standardised effects on attack and defence of possession outcome, possession duration, score difference, match time, opposition strength and season time. During attack, the team and all sub-groups adopted greater lateral dispersion between players, while on defence there was generally greater longitudinal dispersion. The team also showed increased longitudinal dispersion when home and opposition possessions ended in a score. Additionally, greater irregularity was observed in active sub-groups (forwards on attack, defenders on defence). Score difference and opposition strength had trivial-small but generally unclear effects. In conclusion, these effects show that analysis of player positions on attack and defence is a promising avenue for coaches and analysts to modify collective tactical behaviours in netball.

Development of the Talent Identification Questionnaire in Soccer for Outfield Players (TIDQ-OP): Coaches' perceptions of the key attributes for player recruitment

Talent identification in sports is a complex process, whereby the aim is to identify, select and promote players who are considered to have the necessary skills and competencies (e.g., technical, tactical, and physical attributes) to succeed in senior adult competition. To support this process, this study focused on the development of a questionnaire to assist in the identification of talent among South African youth soccer players. This will establish a measure of the attributes coaches consider important when making talent identification decisions for outfield soccer players. The questionnaire was constructed following established instrument development phases. In phase 1, 61 questionnaire items were generated based on a review of the relevant literature. In phase 2, a panel of experts examined the content validity of the preliminary items. In phase 3, the provisional 59-item questionnaire (Talent Identification Questionnaire in Soccer for Outfield Players; TIDQ-OP) was administered to 173 experienced South African soccer coaches (130 males and 43 females) who ranged in age from 18 to 64 years. In phase 3, the factorial structure of the questionnaire was established through principal component analysis. The final questionnaire consisted of 58 items. These preliminary analyses indicate the usefulness of the TIDQ-OP as a tool to assist South African coaches in the identification and selection of talented players.

Investigating the Emergence of Collective States Within Rugby Sevens Gameplay

Rugby sevens is a small-sided variant of rugby union characterised by fast-moving, high-intensity gameplay and is an example of a team invasion sport, where players work together to achieve a shared goal of attacking and defending as a cohesive unit. The dynamics of such sports can be viewed as self-organizing systems, where individual players form collective patterns without a centralized mechanism of control. Inspired by the analysis of collective movement in animals, this novel study investigates the emergent patterns of order and disorder in sub-elite female rugby sevens using order parameters (typically used to analyse particle systems) to characterize the team's collective state during different phases of play. The findings demonstrate that defensive gameplay is more ordered, with more compact formations, compared to attacking play, and there is a correlation between alignment/order in player motion and group speed. The work further suggests that the collective states formed differ between sequences of play with different levels of ground gained by the attacking team. These observations provide a sound justification for player training with a focus on cohesive defensive movements to resist disruptions from opposing attackers, while employing attacking tactics that disrupt the cohesion and order of opposing teams.

The Relationship Between Neck Strength and Sports-Related Concussion in Team Sports: A Systematic Review With Meta-analysis

OBJECTIVE: We aimed to quantify the relationship between neck strength and sports-related concussion (SRC) for athletes participating in team sports. DESIGN: Etiology systematic review with meta-analysis. LITERATURE SEARCH: PubMed, PsycINFO, MEDLINE, CINAHL, CENTRAL, and Scopus were searched on March 17, 2022, and updated on April 18, 2023. STUDY SELECTION CRITERIA: Team sports where an opponent invades the player's territory (eg, football, rugby, basketball) that reported at least 1 measure of neck strength, and 1 measure of SRC incidence, using cohort, case-control, or cross-sectional study designs. The Newcastle-Ottawa scale was used to assess risk of bias; certainty of evidence was assessed using the Grading of Recommendations, Assessment, Development, and Evaluation approach. DATA SYNTHESIS: Studies were summarized qualitatively and quantitatively. To understand the relationship between neck strength and future SRC incidence, random-effects meta-analysis was conducted on prospective longitudinal studies. RESULTS: From a total of 1445 search results, eight studies including 7625 participants met the inclusion criteria. Five studies reported a relationship between greater neck strength or motor control and reduced concussion incidence. Pooled results from 4 studies indicated small (r = 0.08-0.14) nonsignificant effects with substantial heterogeneity (I2>90%). The considerable heterogeneity is likely a result of synthesized studies with vastly different sample characteristics, including participant age, playing level, and sports. CONCLUSIONS: There was very low-certainty evidence suggesting a small, nonsignificant relationship between greater neck strength and a lower risk of sustaining a SRC. J Orthop Sports Phys Ther 2023;53(10):1-9. Epub: 10 July 2023. doi:10.2519/jospt.2023.11727.

Analysis of contextualized intensity in Men's elite handball using graph-based deep learning

Manual annotation of data in invasion games is a costly task which poses a natural limit on sample sizes and the level of granularity used in match and performance analyses. To overcome this challenge, this work introduces FAUPA-ML, a Framework for Automatic Upscaled Performance Analysis with Machine Learning, which leverages graph neural networks to scale domain-specific expert knowledge to large data sets. Networks were trained using position data of match phases (counter/position attacks), annotated manually by domain experts in 10 matches. The best network was applied to contextualize N = 539 matches of elite handball (2019/20-2021/22 German Men's Handball Bundesliga) with 86% balanced accuracy. Distance covered, speed, metabolic power, and metabolic work were calculated for attackers and defenders and differences between counters and position attacks across seasons analyzed with an ANOVA. Results showed that counter attacks are shorter, less frequent and more intense than position attacks and that attacking is more intense than defending. Findings show that FAUPA-ML generates accurate replications of expert knowledge that can be used to gain insights in performance analysis previously deemed infeasible. Future studies can use FAUPA-ML for large-scale, contextualized analyses that investigate influences of team strength, score-line, or team tactics on performance.

Real time prediction of match outcomes in Australian football

This study aimed to determine whether machine learning models based on technical performance and not score margin could be used to predict end-of-match outcome of Australian football matches in real-time. If efficacious, these models could be used to generate insights about team performance and support the decision-making of coaches during matches. A database of 168 team technical performance indicators from 829 Australian Football League matches played between 2017 and 2021 was used. Two feature sets (data-driven and data-informed) were used to train and evaluate six models (generalised linear model, random forest and adaboost) on match outcome prediction (Win/Loss) over 120 epochs (a representation of normalised time during each match). All models performed well (mean classification accuracy = 73.5-75.8%) in comparison with a benchmark score-based model (mean classification accuracy = 77.4%). Data-informed feature sets performed better than data-driven in most cases. Classification accuracy was low at the start of a match (45.7-48.8%) but increased to a peak near the end of a match (87.2-92.7%). These findings suggest that any of the employed models can be used to formulate in-match decision support. The model which is best in practice will depend on factors such as time-cost trade-off, feasibility and the perceived value of its suggestions.

Methods of performance analysis in women's Australian football: a scoping review

Background

The first women's Australian football (AF) professional competition was established in 2017, resulting in advancement in performance analysis capabilities within the sport. Given the specific constraints of women's AF, it is currently unclear what match-play performance analysis methods and techniques are implemented. Therefore, the aim of this scoping review was to describe and critically appraise physical, technical, and tactical performance analysis methods that have been employed in women's and girls' AF match-play.

Methodology

A systematic search was conducted on the 27th of June 2022 through five databases. Eligibility criteria were derived from the PCC framework with the population (P) of women and girls AF players, of any level of play; concepts (C) of interest which were measures, data, and methods related to the sport's physical, technical, and tactical performance; and the context (C) of methods that analysed any match-play performance. A narrative synthesis was conducted using extracted study characteristic data such as sample size, population, time period, collection standards, evaluation metrics for results, and application of thematic categorisations of previous sports performance reviews. Critical appraisal of eligible studies' methodologies was conducted to investigate research quality and identify methodological issues.

Results

From 183 studies screened, twelve eligible studies were included, which examined match-play through physical (9/12, 75%), technical (4/12, 33%), and tactical analysis (2/12, 17%). Running demands and game actions analysis were the most researched in senior women's AF. Research into junior girls' AF match-play performance has not been investigated. No research has been conducted on non-running physical demands, contact demands, acceleration, and tactical aspects of women's AF. All studies utilised either inferential statistics or basic predictive models. Critical appraisal deemed most studies as low risk of bias (11/12, 92%), with the remaining study having satisfactory risk.

Conclusions

Future research utilising increased longitudinal and greater contextual data is needed to combat the prominent issue of data representativeness to better characterise performance within women's and girls' AF. Additionally, research involving junior and sub-elite AF players across the talent pathways is important to conduct, as it provides greater context and insight regarding development to support the evolving elite women's AF competition. Women's AF has been constrained by its resource environment. As such, suggestions are provided for better utilisation of existing data, as well as for the creation of new data for appropriate future research. Greater data generation enables the use of detailed machine learning predictions, neural networks, and network analysis to better represent the intertwined nature of match-play performance from technical, physical, and tactical data.

Operational Insights into Analysing Team and Player Performance in Elite Rugby League: A Narrative Review with Case Examples

In professional team sports, like Rugby League, performance analysis has become an integral part of operational practices. This has helped practitioners gain deeper insight into phenomena like team and athlete behaviour and understanding how such behaviour may be influenced by various contextual factors. This information can then be used by coaches to design representative practice tasks, inform game principles and opposition strategies, and even support team recruitment practices. At the elite level, the constant evolution of sports technology (both hardware and software) has enabled greater access to information, making the role of the performance analyst even more valuable. However, this increase in information can create challenges regarding which variables to use to help guide decision-making, and how to present it in ways that can be utilised by coaches and other support staff. While there are published works exploring aspects of performance analysis in team sports like Rugby League, there is yet to be a perspective that explores the various operational uses of performance analysis in Rugby League, the addition of which could help guide the practices of emerging performance analysts in elite organisations like the Australian National Rugby League and the European Super League. Thus, this narrative review—with accompanying case examples—explores the various ways performance analysis can help address pertinent operational questions commonly encountered when working in high-performance sport.

In the Opinion of Elite Volleyball Coaches, How Do Contextual Variables Influence Individual Volleyball Performance in Competitions?

The main objective was to know the elite coaches’ opinions regarding the relevance, definition, and importance of volleyball contextual variables to measure individual performance in competition. After performing a literature review, an instrument to gather the opinion of the world’s elite volleyball coaches was elaborated by four volleyball specialists. The sample of experts consisted in 20 world’s elite volleyball coaches who met at least three experience years in first division or national teams. The instrument collected experts’ information on the contextual variables in relation to relevance, definition, and importance. Cronbach’s α and Aiken’s V coefficient were used to test the reliability and content validity of the contextual variables, respectively. To compare the importance of the contextual variables U de Mann-Whitney and Kruskal-Wallis tests were used. Results showed that opposition level, set period, score difference, results of the previous set, competitive load variables, high level, final periods ≥20, and high load categories were relevant (Aikens V > 0.70). In addition, high level, final period ≥20 and ≥+10, and high load categories were significantly more important (p < 0.05). We conclude that, according to the elite coaches, the contextual variables should not be analyzed separately. Future studies should consider contextual variables dynamically.

Better Offensive Strategy in Basketball: A Two-Point or a Three-Point Shot?

To better understand and explore the development trend of the offensive strategies of the world's top basketball leagues, this study took NBA shooting data in the regular seasons from 2009/2010 to 2018/2019 as the samples and analysed the relationships between the shooting score ratio and game win probability, and the practical application of offensive strategies in games. The results showed that (1) increasing the number and the percentage of three-point offenses in the game can improve the probability of winning. However, too many two-point shots can affect the team's winning probability to a certain extent. (2) The strong teams in the NBA focused more on the outside offense, while the weak teams focused more on the inside offense. (3) Statistical data further showed that whether a team's opponent is strong or weak, taking the offensive strategy that tends toward the outside shot can lead to a higher game win probability than the offensive strategy that tends toward the inside offense.

Characterizing player's playing styles based on player vectors for each playing position in the Chinese Football Super League

Characterizing playing style is important for football clubs on scouting, monitoring and match preparation. Previous studies considered a player's style as a combination of technical performances, failing to consider the spatial information. Therefore, this study aimed to characterize the playing styles of each playing position in the Chinese Football Super League (CSL) matches, integrating a recently adopted Player Vectors framework. Data of 960 matches from 2016-2019 CSL were used. Match ratings, and 10 types of match events with the corresponding coordinates for all the line-up players whose on-pitch time exceeded 45 minutes were extracted. Players were first clustered into eight positions. A player vector was constructed for each player in each match based on the Player Vectors using Nonnegative Matrix Factorization (NMF). Another NMF process was run on the player vectors to extract different types of playing styles. The resulting player vectors discovered 18 different playing styles in the CSL. Six performance indicators of each style were investigated to observe their contributions. In general, the playing styles of forwards and midfielders are in line with football performance evolution trends, while the styles of defenders should be reconsidered. Multifunctional playing styles were also found in high-rated CSL players.

Different Sampling Frequencies to Calculate Collective Tactical Variables during Competition: A Case of an Official Female's Soccer Match

The objective of the study was to assess the impact of the sampling frequency on the outcomes of collective tactical variables during an official women’s soccer match. To do this, the first half (lasting 46 min) of an official league match of a semi-professional soccer team belonging to the Women’s Second Division of Spain (Reto Iberdrola) was analysed. The collective variables recorded were classified into three main groups: point-related variable (i.e., change in geometrical centre position (cGCp)), distance-related variables (i.e., width, length, height, distance from the goalkeeper to the near defender and mean distance between players), and area-related variables (i.e., surface area). Each variable was measured using eight different sampling frequencies: data every 100 (10 Hz), 200 (5 Hz), 250 (4 Hz), 400 (2.5 Hz), 500 (2 Hz), 1000 (1 Hz), 2000 (0.5 Hz), and 4000 ms (0.25 Hz). With the exception of cGCp, the outcomes of the collective tactical variables did not vary depending on the sampling frequency used (p > 0.05; Effect Size < 0.001). The results suggest that a sampling frequency of 0.5 Hz would be sufficient to measure the collective tactical variables that assess distance and area during an official soccer match.

Capture, analyse, visualise: An exemplar of performance analysis in practice in field hockey

The goal of performance analysis is to capture the multitude of factors that affect sports strategy, and present them in an informative, interpretable, and accessible format. The aim of this study was to outline a performance analysis process in field hockey that captures, analyses and visualises strategy in layers of detail culminating in the creation of an RStudio Shiny application. Computerised notational analysis systems were developed to capture in-game events and ball tracking data of 74 matches from the Women’s Pro League 2019. Game styles were developed using k-means cluster analysis to reduce detailed in-game events into practical profiles to identify the attack types, game actions and tempo of a team’s strategy. Ball movement profiles were developed to identify the predictability (entropy) and direction (progression rates) of ball movements, and consequent distribution of possession in different attacking zones. The Shiny application, an interactive web-platform, links the information from simple game profiles with detailed game variables to understand each teams’ holistic game plan, how they are different, and how to exploit these differences. The process outlined can be applied to any team invasion sport to understand, develop and communicate successful strategies under different match situations.

The Impact of a Swimming Training Season on Anthropometrics, Maturation, and Kinematics in 12-Year-Old and Under Age-Group Swimmers: A Network Analysis

Understanding fluctuations and associations between swimming performance-related variables provide strategic insights into a swimmer's preparation program. Through network analysis, we verified the relationships between anthropometrics, maturation, and kinematics changes (Δ) in 25-m breaststroke (BREAST) and butterfly (FLY) swimming performance, before and after a 47-week swimming training season. Twenty age-group swimmers (n =11 girls: 10.0 ± 1.3 years and n = 9 boys: 10.5 ± 0.9 years) performed a 25-m all-out swim test (T25) in BREAST and FLY techniques, before and after 47 weeks. Three measures of centrality, transformed into a z-score, were generated: betweenness, closeness, and strength. Data were compared (t-test) and effect sizes were identified with Hedges' g. Large effect sizes were observed for swimming performance improvements in BREAST (32.0 ± 7.5 to 24.5 ± 3.8 s; g = 1.26; Δ = −21.9 %) and FLY (30.3 ± 7.0 to 21.8 ± 3.6 s; g = 1.52; Δ = −26.5 %). Small to moderate effect sizes were observed for anthropometric changes. Moderate effect size was observed for maturity offset changes (−2.0 ± 0.9 to −1.3 ± 1.0; g = 0.73; Δ = 50.9 ± 281 %). Changes in maturity offset, stroke rate (SR), and stroke length for both BREAST and FLY swimming speeds were highlighted by the weight matrix. For betweenness, closeness, and strength, changes in arm span (AS) (BREAST) and stroke length (FLY) were remarkable. The dynamic process of athletic development and the perception of complexity of fluctuations and associations between performance-related variables were underpinned, particularly for simultaneous swimming techniques in age-group swimmers.

Identifying and analysing game styles and factors influencing a team's strategy in field hockey

Characterising a team's game style is a performance analysis approach that captures game events, and groups them into profiles using clustering techniques to identify the consistent (and winning) strategies a team implements. The aim of this study was to identify the game styles of international hockey teams. Video footage from the 2019 Pro League tournament (n = 74 female and n = 57 male matches) were analysed retrospectively using a notational analysis system in SportsCode™. Variables were arranged into six game style categories (established attack game actions, counter attack game actions, established attack success, counter attack success, set pieces, tempo) and two game style types identified per category using a k-means clustering algorithm. Decision trees were used to identify the influence of extrinsic and intrinsic match factors on the probability of a team playing a particular game style. Opposition and other reference team game style categories were shown to be more important in predicting a game style category than contextual factors. Examination of team profiles highlights how different strategies are successful for different teams such as high-intensity attack or absorbing pressure and counter attacking. This performance analysis process provides practical insights into the holistic performance of international hockey teams.

Evaluation of cognitive load in team sports: literature review

In team sports, load management has become one of the most common areas of investigation, given that effective control over load is the key to being able to optimize performance and avoid injuries. Despite the constant evolution and innovation in the latest theories, we can see a clear tendency in load management that focuses on physiological and mechanical aspects and neglects its cognitive character, which generates the variability inherent in the performance of athletes in a changing environment. Indicators of response that inform methods of control over cognitive load can include cognitive, physiological and behavioral indicators. However, limited investigations exist to support the reliability of each indicator regarding cognitive load. For this reason, the objective of this literature review is to present strategies used to manage cognitive load in team sports, as well as the indicators utilized for such a proposition and their relationships in specific contexts.

Analysis of the Research Methodology in Spanish Doctoral Theses on Handball. A Systematic Review

The objective of this investigation was to analyze scientific production assessed by indexed doctoral theses in the Ordered Spanish Theses (TESEO) database, on the topic of the sport of handball in Spain. Productivity was analyzed on the basis of variables grouped by contextual information, methodologies and procedures. Seventy-two indexed theses from between 1976 and 2021 were analyzed. A progressive increase was identified in scientific production based on these theses during this period. The scientific disciplines that presented the highest number of theses were Sport Sciences (n = 33) and Sport Pedagogy (n = 19). The main results show that quantitative approaches are the most common (n = 49), doctoral theses using descriptive studies based on systematic observation represent the majority (73%), the predominant type of data collection was the use of cross-sectional studies (70.8%) versus longitudinal studies (26.4%), and the most used sampling method was that of convenience (n = 65). The results make it possible to ascertain the reality of this research topic, the methodological positioning and research tendencies, and to draw the basic lines for development.

The porous high-press? An experimental approach investigating tactical behaviours from two pressing strategies in football

The aim of this study was to analyse footballers' tactical behaviours from their position data, as an effect of two contrasting pressing strategies, high-press defending and deep-defending, using a trial-based experimental approach. Sixty-nine youth footballers participated in this 11 versus 11 study, performing 72 trials of attack versus defence, in a counterbalanced crossover study design. Players' position data were captured using a local positioning system, and processed to calculate measures of inter-team distance, trial duration, distance to nearest opponent, dispersion, team length, team width, team shape, space control gain, inter-line distance, and individual area. This was augmented by the notational analyses of passes. The findings showed that using a high-press defending strategy leads to: closer inter-team distance; larger dispersion, due to a longer team length; and larger inter-line distances between defenders, midfielders, and forwards. The resulting effects on the attacking team include reduced ball possession time; larger individual areas for attacking midfielders and forwards; longer team length; and more penetrative passes performed. Some differences in marking behaviour were also observed. Consequently, the study recommends that high-press defending be used sparingly due to these trade-offs.

Is Perceived Autonomy Support Provided by a Coach Related to the Intention of Injury Preventative Behavior Among National and International Level Golfers?

The successful implementation of injury prevention programs is reliant on athletes and coaches accepting, adopting, and complying with behaviors that reduce injury risk. Exploring factors, such as motivation and planned behavior, that might increase the frequency of these behaviors warrants investigation. The aim of the study was to investigate the complex interaction between perceived autonomy support, self-determined motivation, planned behavior, and how this relates to golfers self-reported intention injury preventative behavior. A total of 60 golfers completed questions on psychological measures of perceived autonomy support from coaches, autonomous motivation, and intentions of injury preventative behavior. A neural network model analysis was performed to investigate the strength of connection between covariates and construct a network structure. Analysis of results was performed by assessing edge strengths and node centrality to guide inference of the network topology. The most central node was autonomous regulation and the results showed one cluster comprising positive interactions between perceived autonomy support, effort of injury preventative behavior, and frequency of injury preventative behavior. When aiming to encourage injury preventative behavior, coaches should consider giving feedback that supports autonomous motivation since it is positively associated with effort and frequency of injury preventative behavior among high-level golfers. Injury prevention programs should include strategies to improve the athlete's autonomous motivation to carry out preventive activities.