Match performance analysis on the Chinese Football Association Super League: A systematic review

Abstract

Founded in the year of 2004, the 20-year-old Chinese Football Association Super League (CSL) serves as a representative of the development of Asian football. Research focused on match performance within the CSL has gained momentum in recent years. This study provides a systematic review of match performance analysis in the CSL, detailing the analytical methods, core topics, major findings, and current research trends. Following the PRISMA 2020 guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analyses), a structured search was conducted in the Web of Science (WOS), PubMed and Scopus databases using targeted keyword combinations, initially identifying 184 studies. After thorough screening, 45 studies were included in the review. These studies were classified into three categories: descriptive analyses, comparative analyses, and predictive analyses. Findings reveal that comparative and predictive analyses are dominant, while descriptive analysis remains essential for data contextualisation. Technical and physical performance analyses are the main research topics to date. Observations suggest the CSL is evolving toward a high-intensity style, with foreign players and substitutes playing pivotal roles. Contextual and environmental factors have been shown to substantially impact match performance. Future research will likely trend towards standardised definitions, speed zones and data, expanding studies on contextual and environmental factors with an emphasis on tactical analysis, goalkeeper performance, set-piece situations, and distinct phases of play.

Keywords

Playing style soccer tactics technical evaluation

Introduction

Football matches are highly complex and dynamic due to the interaction of multiple variables and inherent uncertainties.¹ Coaches often rely on experience and memory during matches; however, given the fleeting nature and volume of information, they can typically recall only a limited portion of events.² This limitation may lead to gaps in post-match analysis and hinder the design of targeted training sessions. In recent years, match performance analysis has played an increasingly critical role in football leagues globally, including the CSL, where its adoption presents both challenges and opportunities. By providing objective, reliable records of match events, match performance analysis allows coaches to evaluate both team and individual performance comprehensively.^3,4 It is now widely considered essential in professional football, supporting training design, match evaluation, and player profiling.^3,5,6

In European and international football leagues, research on match performance analysis is extensive, with numerous studies employing a variety of statistical methods, such as analysis of variance, regression models, and machine learning techniques, to identify key indicators of technical, tactical, and physical performance under varying contextual factors.^3,7,8 Systematic reviews have emphasised the global importance of match performance analysis, identifying trends, key indicators, and best practices.^9,10 However, the CSL, with its distinct developmental context within Asia, has evolved uniquely since its establishment in 2004 by the Chinese Football Association, rising quickly to become one of Asia's premier football leagues, noted for its competitive intensity and strong fan engagement.^11,12 In recent years, the CSL's performance in the AFC Champions League has further boosted its international profile.¹³ Consequently, an in-depth review on the studies of match performance analysis in the CSL is not only crucial for understanding its influence within Asian football but also valuable for revealing the trends that underscore the CSL as a representative of football's growth in Asia.

Research on CSL match performance analysis began relatively late, with the first English-language article on the topic published in 2016 in the International Journal of Performance Analysis in Sport.¹⁴ Over the past few years, the volume of articles on CSL match performance has increased significantly, providing insights into technical, tactical, and physical performance from various perspectives, such as differences in performance between foreign and domestic players, the impact of substitutes, and performance under varying contextual factors.^15–17 Although these studies offer valuable insights, most CSL research remains isolated and fragmented, each addressing a specific topic in isolation. As a result, there has been no comprehensive review that synthesises findings on CSL match performance, which makes it difficult to fully understand the characteristics and developmental trends in CSL performance.

This study aims to provide a systematic review and analysis of research on CSL match performance, addressing the following questions: (1) What are the most common types of analysis and topics in current CSL match performance research? (2) What are the main findings of existing studies? (3) What are the limitations of current research, and what are the directions for future research? Through this review, we aim to not only enhance understanding of CSL performance characteristics but also to provide valuable insights and guidance for future research, team management, and potential improvements within the league.

Methods

Search strategy

According to the PRISMA 2020 guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analyses), a systematic review of the existing literature on match performance analysis in adult male football within the CSL was conducted. The review began with the earliest available published English literature and involved electronic search in the WOS, PubMed, and Scopus databases, which were chosen for their extensive coverage and relevance to sports science and performance analysis literature, with the search cut-off date being June 24, 2024. To ensure a comprehensive literature search, Boolean operators “AND” and “OR” were used to combine targeted keywords. The keywords were carefully selected to cover the main elements of the study, including the sport (‘football’ or ‘soccer’), the specific league (‘Chinese Football Association Super League’, ‘Chinese Super League’, ‘CSL’, or ‘Chinese football teams’), and key aspects of performance evaluation (‘match analysis’, ‘performance analysis’, ‘notational analysis’, ‘game analysis’, ‘patterns of play’, ‘techni*’, ‘tactic*’, ‘physical’, ‘pass*’, or ‘run*’). This combination was chosen to capture a wide range of literature on technical, tactical, and physical aspects related to CSL match performance, which enhances the comprehensiveness and relevance of the reviewed studies. All documents were managed using EndNote X9.1, which facilitated the removal of duplicate records and the screening of literature. The analysis was carried out by two reviewers (WZY and DRC).

Figure 1.

Literature screening flowchart.

Inclusion and exclusion criteria

Papers were included if they met all the following criteria: (1) The research sample data originated from the CSL; (2) Empirical studies related to match performance analysis, including technical and tactical evaluations, time-motion analysis, and statistical compilation; and (3) Published in peer-reviewed English-language journals.

Studies were excluded if they met any of the following criteria: (1) They were not related to football, such as rugby; (2) They involved match performance analysis not related to players, such as analyses of referees; (3) They pertained to data on female football players; (4) They did not include relevant data; and (5) They were conference abstracts, theses, case reports, or reviews.

Methodological quality assessment

As match performance analysis studies inherently involve data from matches and are thus typical quantitative research, this study utilised the literature quality assessment scale developed by Sarmento et al.¹⁰ specifically for match performance analysis. This scale was adjusted and modified based on the background of match performance analysis after evaluation by five senior researchers with extensive experience in performance analysis. It is designed as a risk-of-bias assessment form specifically for match performance analysis. The scale consists of 16 items, with a score of 1 assigned for meeting the criteria, 0 for not meeting the criteria, and “not applicable” noted where appropriate. Consistent with previous research,^10,18 percentage scores were used as the final literature quality assessment standard. The percentage score is calculated as (score/total score) × 100%, with the following classifications: (1) low methodological quality, with a score < 50%; (2) good methodological quality, with a score between 50% and 75%; and (3) excellent methodological quality, with a score ≥ 75%.

Results

Search results and study selection

Using Boolean search methods as previously described, a total of 184 articles were retrieved from the three selected databases (WOS: 107; PubMed: 30; Scopus: 47). After removing 61 duplicate records, 123 articles were left for screening. Initially, 77 articles were excluded based on a review of the title and abstract. Subsequently, the full texts of the remaining 46 articles were reviewed, resulting in the exclusion of 1 article, leaving of total of 45 articles. Additionally, no articles were obtained through other sources. Therefore, a final total of 45 articles were included in this study. The flowchart is shown in Figure 1.

Study characteristics

The basic characteristics of the literature are presented in Table 1. The articles included in this review on match performance analysis in the CSL were first published in 2016, with research samples drawn from CSL matches between 2012 and 2021, spanning 10 seasons of the league. In terms of analysis types, the classification method of Sarmento et al.⁹ was adopted, categorising match performance analysis into descriptive analysis, comparative analysis, and predictive analysis (Figure 2). The discussion and analysis in this study are also based on these three categories. Among the included studies, 34 articles involved descriptive analysis, 28 involved comparative analysis, and 29 involved predictive analysis. We summarised the research topics in CSL match performance analysis, finding that 32 studies focused on technical performance, 20 on physical performance, and 8 on tactical performance. Additionally, 3 studies examined the impact of rules, technical actions, and behavioural events on match time. Regarding data sources, 14 studies used data from Amisco, 12 from OPTA, and 8 from Champdas, 3 studies combined data from two of these databases, all of which have had their reliability and validity verified.^19–21 Data from Sina Sports, China Sports Media, the CSL official database, or the NetEase Super Data Live Broadcast System was used in 5 articles, though the reliability of these databases was not mentioned. Additionally, 2 studies independently collected data using experienced observers, whose data validity and reliability were verified. Only 1 article did not report its data source.

Figure 2.

Scopes of match analysis (adapted from Sarmento et al.⁹).

Table 1.

Basic characteristics of the included literature.

Study	Sample	Topics	Analysis types	Data resource	Quality score
Mao et al.¹⁴	480 CSL matches (2014–2015)	TePA	D, P	OPTA	94%
Gai et al.²²	240 CSL matches (2014)	TePA, PPA	D, C, P	Amisco	94%
Zhou et al.²³	1056 CSL matches (2012–2017)	TePA, PPA	D, C, P	Amisco	88%
Gong et al.²⁴	237 CSL matches (2019)	TePA, PPA	D, P	Amisco	100%
Li et al.²⁵	1200 CSL matches (2014–2018)	TePA	D, C, P	Champdas	94%
Gai et al.¹⁶	240 CSL matches (2014)	TePA, PPA	D, C	OPTA	94%
Liu et al.²⁶	720 CSL matches (2014–2016)	TePA, PPA	D, C, P	OPTA	94%
Han et al.²⁷	480 CSL matches (2017–2018)	TePA, PT	D, C, P	OPTA	94%
Zhou et al.¹³	240 CSL matches (2015)	TePA, PPA	D, C, P	Amisco	100%
Gai et al.²⁸	240 CSL matches (2014)	TePA, PPA	D, C, P	OPTA, Amisco	94%
Gai et al.¹⁵	240 CSL matches (2016)	TePA	D, C	OPTA	94%
Zhou et al.²⁹	1429 CSL matches (2012–2017)	TePA, PPA	D, C, P	Amisco	94%
Wu and Swartz³⁰	CSL matches (2019)	PPA	P	OPTA	75%
Yu et al.³¹	240 CSL matches (2017)	TaPA	D, C	Champdas	94%
Li et al.³²	1899 CSL matches (2012–2019)	TePA, PPA	C	Amisco	100%
Chen et al.³³	397 CSL matches (2019–2020)	TePA, PPA	D, C, P	Amisco	100%
Gong et al.³⁴	1200 CSL matches (2014–2018)	TaPA	C	Champdas	88%
Wei et al.³⁵	CSL matches (2014–2018)	TePA, PPA	D, P	Amisco	81%
Li et al.¹¹	1440 CSL matches (2014–2019)	TePA	D, P	Sina Sports	75%
Yao et al.³⁶	1820 CSL matches (2012–2019)	TePA	P	China Sports Media	88%
Han et al.³⁷	400 CSL matches (2019–2020)	TePA	C, P	Champdas	88%
Ma et al.³⁸	2400 CSL matches (2010–2019)	TePA	D, P	CSL database	88%
Zhou et al.¹⁷	1429 CSL matches (2012–2017)	TePA, PPA	D, P	Amisco	88%
Guan and Swartz³⁹	237 CSL matches (2019)	PPA	P	—	81%
Lago-Peñas et al.⁴⁰	240 CSL matches (2016)	TaPA	D, C	OPTA	94%
Zhao and Liu⁴¹	50 CSL and Premier League, respectively (2019)	PT	D, C, P	VAEO	100%
Ruan et al.⁴²	1120 CSL matches (2016–2020)	TaPA	P	OPTA	94%
Feng et al.⁴³	1899 CSL matches (2012–2019)	TePA, PPA	D, C	Amisco	100%
Ruan et al.⁴⁴	240 CSL matches (2018)	TaPA	D, C	OPTA	88%
Yuan et al.¹²	632 CSL matches (2014–2016）	TePA, PPA	D, P	NetEase	81%
Zhou et al.⁴⁵	1429 CSL matches (2012–2017)	TaPA	C	Amisco	94%
Pan et al.⁴⁶	240 CSL matches (2018)	TePA, PPA	D, P	Amisco, Champdas	100%
Zhao and Zhang⁴⁷	1680 CSL matches (2012–2018)	PT	P	Champdas	88%
Zhou et al.⁴⁸	1427 CSL matches (2012–2018)	TePA, PPA	D, C	Amisco	94%
Liu et al.⁴⁹	240 CSL matches (2018)	TePA, PPA	D, C	Amisco	100%
Kong et al.⁵⁰	240 CSL matches (2019)	TaPA	C	Amisco	94%
Zeng and Zhang⁵¹	1440 CSL matches (2014–2019)	TePA	D, C, P	Champdas	88%
Xu et al.⁵²	1190 CSL matches (2014–2018)	TePA	D, P	Sina Sports	88%
Jimenez et al.⁵³	237 CSL matches (2016)	TePA	D, P	OPTA	94%
Liu et al.⁵⁴	240 CSL matches (2018)	TaPA	D, P	VAEO	88%
Liu et al.⁵⁵	1116 CSL matches (2014–2018)	TePA	D, P	Champdas	94%
Li et al.⁵⁶	960 CSL matches (2016–2019)	TePA, TaPA	C	OPTA	88%
Zhao and Zhang⁵⁷	1200 CSL matches (2014–2018)	TaPA	D, C	Champdas	81%
Zhang et al.⁵⁸	800 CSL matches (2015–2021)	TePA, PPA	D, C	OPTA, Amisco	94%
Jiang et al.⁵⁹	400 CSL matches (2019–2020)	TePA	D, C	OPTA	94%

Note: D: descriptive; P: predictive; C: comparative; TePA: technical performance analysis; TaPA: tactical performance analysis; PPA: physical performance analysis; PT: playing time; VAEO: video analysis by experienced observer; NetEase: NetEase super data live broadcast system.

Methodological quality assessment

The methodological quality assessment was achieved by a scale specifically designed for studies of match performance analysis.¹⁰ Inter-observer reliability of the assessment was tested using the KAPPA value in this study, with the average KAPPA value across the 16 items being 0.96 (with a 90% confidence interval of 0.93–1), indicating very high consistency among the observers in evaluating the quality of the literature. Based on the results of the literature assessment: (1) the average score across all assessed studies was 91%; (2) seven studies achieved a score of 100%; (3) no study scored below 75% (Table 1). Therefore, we consider the included studies to represent the highest quality in terms of literature evaluation.

In the quality assessment of the included studies, the primary areas of point deductions were: (1) failure to obtain appropriate ethical approval or informed consent; and (2) lack of description regarding the limitations of the study.

Discussion

Descriptive analysis

Descriptive analysis involves presenting raw or adjusted data (standardised data) as means or medians, accompanied by measures of variability such as standard deviation, range, or percentiles.⁶⁰ Descriptive analysis is not the primary focus of current research, but it has become a standard paradigm for presenting data in the results section, providing foundational data for subsequent comparative and predictive analyses. Among the included studies, only 11 did not provide descriptive data. In current research, descriptive analysis summarises the overall activity patterns of teams or players during matches,⁹ such as the distribution of various match events and physical performance metrics. However, existing descriptive analysis lacks in-depth exploration of critical match scenarios or specific phases, such as attacking phases, defensive phases, and transitions between attack and defence. While descriptive analysis provides foundational data for research, there remains significant room for improvement in the detailed description and contextual analysis of specific match phases.

In the CSL, the average match duration was 96.68 min, with an average effective playing time of 52.91 min. Fouls and out-of-bounds accounted for the longest periods of stoppage time, with durations of 7.85 and 7.56 min, respectively.⁴¹ Teams in the CSL achieved approximately 9–14 shots per game,^33,38,49 with a shot accuracy rate of around 30%-54%.^33,38 The average number of passes per game ranged from 300 to 450,^14,22,49,51 and the average total distance was approximately 103–109 km.^{14,17,23,33,49} For individual players, the running distance per match was about 8.8–11 km.^28,48,58 However, we were unable to summarise the running characteristics of CSL players across different speed zones due to variations in the speed thresholds defined by different studies, as shown in Table 2. This inconsistency hinders the generalisation of running characteristics and the comparison of research results.

Table 2.

Speed zones for physical performance in the CSL.

Study	Speed zones
Gai et al.²²	Sprint: > 25.1 km/h; HSR: 19.8–25.1 km/h; HIR: > 19.8 km/h
Zhou et al.²³	Sprint: > 23 km/h; HSR: 19.1–23 km/h
Gong et al.²⁴	Sprint: > 25.1 km/h; HSR: 19.7–25.1 km/h; MSR: 14.3–19.7 km/h; LSR: 7.1–14.3 km/h; HIR: > 14.3 km/h
Gai et al.¹⁶	Sprint: > 25.1 km/h; HIR: > 19.8 km/h
Zhou et al.¹³	Sprint: > 23 km/h; HSR: 19.1–23 km/h; HIR: > 19 km/h
Gai et al.²⁸	Sprint: > 25.1 km/h; HIR: ≥ 19.8 km/h
Zhou et al.²⁹	Sprint: > 23 km/h; HSR: 19.1–23 km/h
Li et al.³²	Sprint: > 25 km/h; HSR: 19.7–25 km/h; MSR: 14.3–19.6 km/h; HIR: > 19.8 km/h
Chen et al.³³	Sprint: > 25.1 km/h; HSR: 19.7–25.1 km/h; HIR: > 14.4 km/h
Wei et al.³⁵	Sprint: > 23 km/h; HSR: 17–23 km/h
Zhou et al.¹⁷	Sprint: > 23 km/h; HSR: 19.1–23 km/h
Feng et al.⁴³	Sprint: > 25 km/h; HSR: 19.7–25 km/h; MSR: 14.3–19.6 km/h; LSR: < 14.3 km/h; HIR: > 19.7 km/h
Pan et al.⁴⁶	Sprint: > 23 km/h; HIR: 19.1–23 km/h
Zhou et al.⁴⁸	Sprint: > 23 km/h; HIR: > 19 km/h
Liu et al.⁴⁹	Sprint: > 23 km/h; HIR: 19.1–22.9 km/h
Zhang et al.⁵⁸	Sprint: > 25.1 km/h; HSR: 19.7–25.1 km/h; MSR: 14.3–19.7 km/h; LSR: 7.1–14.3 km/h

Note: HSR: high speed running; MSR: moderate speed running; LSR: low speed running; HIR: high intensity running.

In the analysis of technical-tactical and running metrics, some studies converted these indicators into per minute standardised data.^16,46 Compared to using the data from the entire match, presenting data per minute is a better approach, as it eliminates the impact of match duration variations due to stoppage time. Additionally, when analysing technical-tactical metrics measured in counts, studies such as those by Mao et al.,¹⁴ Zhou et al.,²³ Ruan et al.⁴⁴ have employed mathematical formulas to convert the data into values adjusted for 50% ball possession. This approach effectively removes the influence of ball possession on technical performance indicators.

Comparative analysis

Different team and player characteristics

In comparative analyses of CSL match performance, there has been no direct comparison between different player positions; instead, player position was typically used as a grouping variable. This differs from the findings of Sarmento and colleagues.⁹ The most frequent focus in CSL match performance comparative analyses was on comparing different team and player characteristics (such as team strength, player roles, player types, and age groups), with a total of 12 studies addressing these comparisons (Table 3).

Table 3.

Empirical studies with predominantly comparative analysis based on the team and player characteristics.

Study	Competition level division strategy	Group arrangement	Main Results
Gai et al.²²	Team strength	Top-ranked (1–4), Upper-middle-ranked (5–8), Lower-middle-ranked (9–12), Lower-ranked (13–16)	Sprint: top > upper-middle. Distance covered without possession: upper > lower. Possession in the opponent's half, passes into the final third and penalty area, 50–50 challenges: top > lower.
Kong et al.⁵⁰	Team strength	Strong (ranking 1–5), Medium (ranking 6–11), Weak (ranking 12–16)	Strong teams favoured intensity play, medium and weak teams opted for possession play both home and away. Strong teams exhibited balanced technical-physical playing style, weak teams demonstrated inferior performance in both intensity and possession play.
Zhao and Zhang⁵⁷	Team strength	High group A (top four), upper middle group B (ranking 5–8), lower middle group C (ranking 9–12), and low group D (bottom four)	Density, eigenvalue, passing, clustering coefficient, unweighted clustering coefficient, weighted clustering coefficient: Group A > others groups.
Gong et al.³⁴	Between specific teams	16 CSL teams	Identifiability values: Top-ranked teams > lower-ranked teams, consistency and stable passing patterns: top-ranked teams > others.
Lago-Peñas et al.⁴⁰	Between specific teams	16 CSL teams	Playing style of the teams differed. For example, possession style: Guangzhou Evergrande, Guangzhou RF, set-piece style: Henan Jianye.
Ruan et al.⁴⁴	Between specific teams	16 CSL teams	Defensive performance in the middle and attacking third: High-level teams > others.
Li et al.⁵⁶	Between specific players	CSL players	Performance differences of similar players, the same player across different seasons, and players in the same position were identified.
Gai et al.²⁸	Player nationality	domestic soccer players, foreign soccer players	Passes, shots, aerial and ground challenges: foreign players > domestic players. Total distance: foreign wide midfielders/wide defenders > domestic counterparts, domestic forwards > foreign counterparts.
Gai et al.¹⁵	Player nationality	domestic soccer players, foreign soccer players	Defenders: foreign > domestic in defending actions. Midfielders: foreign > domestic in passing and attack organization; domestic > foreign in defending actions. Forwards: foreign > domestic in passes and attack organization.
Yu et al.³¹	Player nationality	domestic soccer players, foreign soccer players	Foreign midfielders/forwards > domestic in passes in, indegree, stress centrality, partner, and betweenness centrality. Foreign midfielders > domestic in passes out, outdegree, and closeness centrality. Domestic defenders/midfielders > foreign in neighbourhood connectivity. Domestic defenders > foreign in passes in, indegree, and partner.
Gai et al.¹⁶	Player role	Starters, substitutes, full-match players	Physical performance and technical actions: substitutes > starters/full-match players. Efficacy: substitutes < starters/full-match players.
Zhou et al.⁴⁸	Age Group	Group 1 (17–22.9 years), Group 2 (23–28.9 years), Group 3 (29–34.9 years), Group 4 (35–40.9 years)	Physical performance: older < younger. Defenders and forwards: older > younger in passing and forward passes. Midfielders: older > younger in passes and crosses. Midfielders and forwards: older > younger in shooting.

In CSL teams, stronger teams demonstrated superior physical and technical performance.²² These teams tended to adopt a high-intensity, aggressive style of play and exceled in passing-related abilities.^50,57 Although stronger teams adopted an aggressive style of play, they still placed significant emphasis on maintaining possession, featuring a consistent and identifiable playing style, along with stable passing patterns.³⁴ In contrast, mid- and lower-ranked teams were more inclined to adopt a ball possession-based approach; however, weaker teams (those ranked in the bottom five of the league) often struggle with effective ball control.⁵⁰

About comparisons of player types in the CSL, foreign players outperformed domestic players, often holding key positions within the team and playing crucial roles across various positions.^15,28,31 In terms of running performance, foreign players, except for forwards, exhibited a greater total running distance compared to domestic players. Furthermore, foreign wide midfielders surpassed domestic players in both sprinting and high-intensity running.²⁸ In terms of technical performance, foreign players across all positions exhibited stronger outcomes on key metrics. For example, defenders achieved more duels and higher success rates, midfielders exceled in key passes and successful dribbling, and forwards demonstrated superior dribbling success.¹⁵ In terms of tactics, foreign players served more frequently as pivotal links in offensive passing sequences, connecting teammates to build the attacking framework. Domestic players, by contrast, were often positioned more peripherally within the passing network, limiting their involvement in offensive construction and were typically assigned more defensive responsibilities within the team.^15,31

Lago-Peñas et al.⁴⁰ and Ruan et al.⁴⁴ used factor analysis to construct the playing styles and defensive styles of CSL teams, comparing the differences in these styles between teams and assessing the extent to which teams rely on specific playing styles. Li et al.⁵⁶ employed mathematical modelling to compare the performance of different players, as well as the same player across different seasons. For example, they compared the playing styles of Alex Teixeira and Wei Shihao, who share similar styles, and the performance of Wu Lei in the 2017 and 2018 seasons. Combining team and individual performance style assessments can be valuable for coaches in tactical planning, training design, and for club management in player recruitment decisions.

One study compared the match performance differences among players of different roles, finding that substitutes exhibited higher physical output and completed more technical actions per minute than starting players and those who played the full match. However, substitutes were identified to be less efficient.¹⁶ Additionally, another study compared the differences in match performance of players across different age groups, revealing that younger players displayed better physical performance, while older players demonstrated superior technical skills,⁴⁸ which suggests that as the increasing of age, the technical ability and match experience helps compensate for the decline in running capacity in CSL players.

Different match context and conditions

In comparative analyses of CSL match performance, 12 studies focused primarily on different match contexts and conditions, such as match location, match result, ball possession percentages, different leagues, halves of matches, with and without spectators, with or without video assistant referee (VAR), and team formation (Table 4). Among these, 5 studies compared match performance under different match locations. Three of these studies categorised match locations as home and away,^13,26,57 while two studies included comparisons with neutral venues.^37,59 The latter was due to the COVID-19 pandemic, which resulted in CSL matches being concentrated at neutral venues. The findings indicate that a home advantage exists in the CSL, with home teams outperforming away teams in key metrics deceive to winning (e.g., shots, passing accuracy) and underperforming in metrics detrimental to winning (e.g., fouls committed, yellow cards).^13,26,57 When teams shift from home venues to neutral grounds, their performance declines, particularly in offensive metrics (such as goals, shots and shots on target), while fouls and yellow cards increase significantly.^37,59 Notably, after the introduction of the VAR in the 2018 season, CSL players’ fouls were suppressed, referees’ decisions became more objective, and the home advantage was somewhat diminished.²⁷

Table 4.

Empirical studies with predominantly comparative analysis based on the match context and condition.

Study	Competition level division strategy	Group arrangement	Main Results
Liu et al.²⁶	Game location	Home, away	Goals, shots on/off target, penalties, outside-box shots, passes, possession, corner kicks: home > away. Free kicks, fouls, yellow cards: home < away.
Zhou et al.¹³	Game location	Home, away	shots, shots on target, possession (in opponent's half), passes, forward passes, 35 m entries, penalty area entries, crosses, corners: home > away. Yellow cards: home < away.
Han et al.³⁷	Game location	Home, away, neutral	Goals, shots (on target, inside/outside box), corners, key passes, crosses, breakthroughs, tackles: neutral < home. Yellow cards, fouls: neutral > home. Free kicks, pass accuracy: neutral > away. Defensive actions (tackles, clearances, blocks): neutral < away.
Jiang et al.⁵⁹	Game location	Home, neutral	Wining rate, offence (shots, shots on target): neutral > home. Possession (passes, passing accuracy), defence (fouls, yellow cards): neutral < home.
Zhao and Zhang⁵⁷	Game location; Match result	Home, away; Win, draw, loss	Home > away: link, diameter, density, unweighted clustering coefficient (CR1), weighted clustering coefficient (CR2), eigenvalue, pass metrics. Losing > winning: links, density, CR2. Winning > losing/draw: successful passes.
Zhou et al.²³	Match result	Win, draw, loss	Winning > drawing/losing: shots, shots on target, 50–50 challenges won, offsides, sprint distance/effort, sprint distance in possession, high-speed running distance in possession. Winning < drawing/losing: crosses. Losing > winning/drawing: passes, forward passes, high-speed running distance without possession.
Liu et al.⁴⁹	Different ball possession level	High ball possession team (HPBPT), low ball possession team (LPBPT)	HPBPT > LPBPT: high-intensity and sprint running (central defenders, fullbacks); technical and tactical performance (attacking organization). LPBPT > HPBPT: high-intensity and sprint running (wide midfielders, forwards).
Zhao and Liu⁴¹	Different league	CSL, English Premier League (EPL)	EPL > CSL: actual playing time, match density, out-of-bounds. CSL > EPL: general foul time, player minor injury time, off-field time.
Feng et al.⁴³	Match half	First half, second half	Second half > first half: Goals, shots, possession in the attacking third, crosses, yellow cards. Running metrics: trivial changes between halves.
Chen et al.³³	With and without spectators	With spectators, without spectators	With > without spectators: crosses, shots, shot accuracy. Without > with spectators: sprint distance.
Han et al.²⁷	With or without VAR	With VAR, without VAR	Post- < Pre-VAR: fouls, offsides. Post- > Pre-VAR: first half, second half, and total match duration.
Zhang et al.⁵⁸	Team formation	3-4-3, 3-5-2, 4-3-3, 4-4-2, 4-1-4-1, 4-2-3-1	3-5-2 > 4-2-3-1 (defenders): total distance, low-speed distance, speeds. 3-5-2 > 4-4-2 (defenders): speeds. 4-2-3-1 > 4-4-2 (midfielders): sprints efforts and distance. 3-4-3 > 3-5-2 (defenders): possession, passes, and pass success rates. 4-defender vs. 3-5-2 (full-backs, midfielders): more passes and higher pass success rates.

There are 2 studies that compare CSL match performance in matches with different outcomes. Winning teams exhibited significantly higher values in metrics such as shots, shots on target, crosses, 50–50 challenges won, offsides, sprint distance and frequency, as well as sprint and high-intensity running distance both in their own and the opponent's possession, compared to teams that drew or lost.²³ Additionally, winning teams showed higher values in the passing network metrics of eigenvalue, passes and diameter.⁵⁷ These findings suggest that winning teams play more directly and aggressively in the CSL.

There is a study that compares performance between the first and second half of matches. In the CSL, players’ physical performance did not show significant changes between the halves, while the values for several attacking related technical performance indicators increased.⁴³ Additionally, several other interesting comparisons were made. Zhang et al.⁵⁸ compared the match performance across different formations in the CSL and found significant differences in the physical and technical performance of players in different positions depending on the formation. Chen et al.³³ examined the impact of spectators on CSL players’ match performance, confirming that the presence of spectators did indeed influence both technical and physical performance. Regarding ball possession, teams with higher possession demonstrated superior technical performance compared to those with lower possession. Central defenders and full-backs in high-possession teams engaged in more high-intensity running, while midfielders and forwards in low-possession teams performed more high-intensity running.⁴⁹ This indicates that CSL teams often control possession on the flanks and in the defensive midfield, while forwards and midfielders need to run more actively to regain possession. Another study compared match duration between the CSL and the English Premier League (EPL), finding that the effective match time in the CSL is lower, with interruptions mainly caused by human factors.⁴¹

Longitudinal change

Longitudinal comparative studies provide strong evidence for the development trends in CSL match performance (Table 5). While the total running distance in the CSL has remained relatively stable, the distance covered at high intensity has shown an increasing trend. In terms of technical performance, there has been an increase in the number of crosses, shots on target, and entries into the opponent's penalty area.²⁹ Therefore, we believe that the CSL is evolving towards a style that emphasises offense and high intensity. Research on tactical styles supports this view, suggesting that the CSL is shifting towards a high-intensity, more direct attacking style, with the possession-based style gradually losing prominence.⁴⁵ Studies using passing networks also indicate a decline in possession-oriented play in the CSL.⁵⁷ Throughout a season, physical performance tends to follow a “U-shaped” pattern, peaking at the beginning and end of the season; technical performance remains relatively stable, but towards the end of the season, teams increase their offensive metrics (such as forward passes and time in individual possession) as they strive to earn more points.⁵⁶

Table 5.

Empirical studies with predominantly comparative analysis based on the longitudinal change.

Study	Units	Specific classification measures	Main Results
Zhou et al.²⁹	Year	The six seasons from 2012 to 2017	Upward trends (2012–2017): crosses, shots on target, and opponent penalty area entries. Passing performance (passes, pass accuracy, forward passes, and forward pass accuracy) peaked in 2013–2014.
Li et al.³²	Month	The 2012–2019 seasons are divided into 9 groups by month (from March to November)	Physical performance of CSL teams characterised like a “U” shape, the highest value was observed in the beginning of season (March), decreased gradually, reaching the lowest in August, and rebounded progressively from September to November. Technical performance presented trivial changes through the whole season
Zhou et al.⁴⁵	Year	The six seasons from 2012 to 2017	Factor 1 (physical high intensity), factor 3 (offensive actions):2017 > 2012. Factor 4 (defensive actions), factor 5 (individual challenges): 2017 < 2012. Factor 2 (possession and passing): 2012–2014 increased, 2014–2017 declined.
Zhao and Zhang⁵⁷	Year	The five seasons from 2014 to 2018	Link, density, weighted clustering coefficient: 2018 < 2014–2016. Passes: 2017/2018 < 2016. Eigenvalue: 2015 < 2016.

Predictive analysis

Sarmento et al.⁹ noted that predictive research in football match performance analysis was relatively scarce before the year of 2014. However, this study presents a contrasting finding, with 29 studies involving predictive analysis specifically in CSL match performance research, 24 of which focus primarily on predictive analysis (Table 6). This discrepancy may be attributed to the evolution of match performance analysis over the past decade. Simple comparative analysis is no longer sufficient to explore the relationships between events in football matches. The development of predictive mathematical models and algorithms (e.g., generalised linear models, discriminant analysis, machine learning) has supported match performance analysis, making predictive analysis a mainstream aspect. A common characteristic of these studies is their focus on determining the most effective strategies to improving match performance.⁹

Table 6.

Empirical studies with predominantly predictive analysis.

Study	Statistical procedure	Prediction Target	Main Results
Mao et al.¹⁴	Cumulative logistic regressions	Match result	Key performance indicator (KPI) influencing winning: shot on target, shot accuracy, cross accuracy, tackle, and yellow card.
Gai et al.²²	Multinomial logistic regression	Match result	An increase in time of possession raised the likelihood of winning.
Zhou et al.²³	Discriminant analysis	Match result	KPI distinguishing winning teams: shots on target, sprinting distance during ball possession, the quality of opposition, the quantity of passes and forward passes.
Gong et al.²⁴	Multiple linear regression，quantile regression	Ball possession	Technical KPI linked to ball possession: fouls, passes, aerial duels, tackles, shots, and corner kicks. Physical KPI linked to ball possession: HSR distance, sprint distance.
Li et al.²⁵	Linear Support Vector Classifier	Teams ranking, Final Points	The predicted final points and team rankings for all teams are highly correlated with the actual outcomes.
Han et al.²⁷	Generalized linear model	Technical performance	VAR technology somewhat reduced the home team advantage.
Wu and Swartz³⁰	Exploratory data analysis	Speed	We have outlined basic principles from exploratory data analysis to help experimenters obtain more accurate estimates of player speed, which can also be applied to calculating velocities and accelerations.
Chen et al.³³	Univariable logistic regression，Multivariable logistic regression analysis	Match result	The key factors for winning matches remain largely unchanged between those played with and without crowds. Team and opponent quality continue to be crucial in determining match outcomes.
Wei et al.³⁵	Econometric model	Technical and physical performance	Temperature has an inverted U-shaped relationship with athletes’ physical performance, and high temperatures have a significantly adverse impact on their technical performance.
Li et al.¹¹	Regression equation	Technical and physical performance	Air pollution negatively impacts the performance of football players. A 1% increase in air pollution leads to a 0.101% decrease in the number of passes. However, possessing an advantage in physical fitness, technical skills, or playing at home helps mitigate the negative impact of air pollution.
Yao et al.³⁶	Ridge Regression, Logistic Regression	Goal	The in-game outcome prediction (IGSOP) algorithm outperforms existing methods in predicting draws and outcomes in the final moments of a match.
Han et al.³⁷	Generalized linear model	Technical performance	KPI influencing player performance: team strength, quality of opposition, and match location.
Ma et al.³⁸	Two-way fixed effects ordered logit model	Match result	As the league progresses, a rise in goals gradually boosts the chances of better match results, while an increase in red card penalties amplifies the likelihood of worse outcomes.
Zhou et al.¹⁷	Binary logistic regression	Match result	2012–2014: shots on target, possession, total distance in and out of ball possession, and match location had a reduced impact on winning matches, 2014–2017: these factors became more critical for victories. Additionally, HSR and HSR out of ball possession also significantly influenced match outcomes.
Guan and Swartz³⁹	Functional data analysis	Maximum acceleration	Football player's average max acceleration: increases steadily until 24, plateaus from 24–28, declines after 28. Decline rate accelerates in later years.
Ruan et al.⁴²	Logistic regression，multiple regression	Goal	Key strategies to reduce opponent's expected goals: strengthening defence near own goal, high-intensity confrontations, improving goalkeeper defence, minimizing errors, avoiding dangerous situations.
Yuan et al.¹²	Semi-parametric model	Technical and physical performance	30°C vs. < 15°C: ↓ total running distance (−619 m), passes (−1.68 units), fouls (−0.3 units). 50 mm precipitation vs. < 0.1mm: ↓ total running distance (−96.9 m), passes (−2.29 units), fouls (−0.06 units).
Pan et al.⁴⁶	Lined mixed model	Technical and physical performance	Substitute players can enhance physical and technical performance depending on their playing positions. Offensive substitutes: ↑ organizing, scoring. Defensive substitute: ↑ defensive strength.
Zhao and Zhang⁴⁷	Linear mixed model with crossed random effects	Effective playing time	Effective playing time ↓: top 8 matches, halftime ties, high audience participation, red card matches, interruptions (goals, corners, free-kicks, throw-ins), destructive actions (fouls, yellow/red cards). Effective playing time ↑: productive actions (shots on target).
Zeng and Zhang⁵¹	Mixed linear model	Pass	Passing indices were affected by: various game contexts, such as the stage of the season, venue, team level, match type, and match outcome.
Xu et al.⁵²	Ordinary least squares regression mode	Match result	PM2.5: negatively impacts outcomes. PM10, SO₂: negatively impacts home goals, CO: negatively impacts away goals.
Jimenez et al.⁵³	Linear regression analysis	Technical performance	Top teams demonstrated consistent performance regardless of location, opponent, or match outcome. Additionally, the quality of the opponent was identified as the most influential factor on teams’ technical behaviour, while the final result was found to be the least relevant in this context.
Liu et al.⁵⁴	Logistic regression	Goal	Shooting success for all teams: match status, regaining possession position, shooting position. For strong teams: offense category, final passing area. For weak teams: final passing area.
Liu et al.⁵⁵	Multinomial logit regression，Poisson regression model	Match result	Teams scoring first have an 87.01% unbeaten rate and are five times likelier to win than lose. Each minute's delay increases their winning probability by 0.0028.

Currently, predictive research on CSL match performance primarily focuses on predicting match outcomes and performance (technical and physical performance). Metrics such as the number of shots on target, shot accuracy, possession rate, sprint distance, and high-speed running distance were identified as key performance indicators influencing the probability of winning in CSL matches.^14,17,22,23 Teams that scored the first goal are more likely to win, and the probability of winning increases when the first goal is scored later in the match.⁵⁵ Additionally, contextual factors (e.g., opponent quality, match location) and environmental factors (e.g., PM2.5, CO) were also important determinants of match outcomes.^17,23,52

Contextual factors, such as team strength, opponent strength, and match location, significantly impacted CSL players’ technical performance.³⁷ Among these, the strength of the own team is crucial: top teams exhibit consistent performance, outperforming opponents regardless of other contextual factors.⁵³ Substitute players generally outperform starters and full-time players in terms of technical and physical performance, allowing coaches to enhance team performance through strategic substitutions.⁴⁶ The introduction of VAR has also influenced match performance to some extent, primarily by reducing the number of fouls and offside.²⁷ Additionally, technical and physical performance are affected by environmental factors such as air pollution, temperature, and humidity.^11,12,35

Three studies focus on predicting goals. Li et al.¹¹ utilised machine learning algorithms to forecast the probability of goals in real-time based on in-match data. Ruan et al.⁴² examined how predicting the opponent's goal probability reflects the effectiveness of a team's defence. Liu et al.⁵⁴ identified match status, regaining possession position, and shooting position as key variables affecting goal scoring.

Additionally, there are individual predictive studies focused on ball possession, team rankings, end-of-season points, speed, maximum acceleration, match time, and passing. Key factors influencing team's ball possession include fouls, passes, aerial duels, tackles, shots, corner kicks, sprint distance, and high-speed running distance.²⁴ Li et al.²⁵ developed a machine learning-based predictive model that effectively forecasts team rankings and final league points. Wu and Swartz³⁰ constructed a more reliable model for calculating player speed, while Guan and Swartz³⁹ developed a model to predict acceleration and age to assess athletic performance. Zhao and Zhang⁴⁷ found that game interruptions (goals, corner kicks, free kicks) and destructive actions (yellow cards, red cards, and fouls) negatively impact effective match time, whereas productive actions (shots on target) have a positive effect. Zeng and Zhang⁵¹ analysed the factors influencing passing metrics across different player positions, providing support for coaches in targeted training and tactical planning.

Contextual variables

Contextual variables or situational factors refer to the varying competitive conditions under which soccer matches are played, and these factors can influence the behaviour and performance of both teams and individual players.⁶¹ Contextual factors mainly include match outcomes, match status, team strength (both the team and the opponent team), match location, and the halves of the match. Among the 44 articles reviewed, 30 considered contextual variables, indicating that the integration of performance metrics with contextual factors is the prevailing trend in current CSL performance analysis. There are three primary approaches to considering contextual variables in CSL performance analysis: (1) Categorisation of contextual variables: this method involves using contextual variables as categorical variables to group and compare data. For instance, studies like Gai et al., Zhou et al., Liu et al.^22,23,26 (2) Exploring the impact of contextual variables: this approach investigates how contextual variables influence match performance. Examples include studies by Gong et al., Zhou et al., Jimenez et al.^17,24,53 (3) Modelling contextual variables as fixed effects: this method involves incorporating contextual variables as fixed effects in mathematical models, enabling the calculation of dependent variables while accounting for the influence of these contextual factors. Studies by Li et al.,³² Feng et al.⁴³ exemplify this approach.

In studies analysing CSL match performance, match outcomes are categorised into several types: win, draw, and loss^23,32,47,57; win and non-win²⁵; loss and non-loss⁴⁴; and goal difference.^51,53 Among these, winning teams generally demonstrate superior performance in technical, tactical, and physical metrics.

Home advantage is also a prominent topic in the analysis of match performance. In the CSL, home advantage significantly influences match outcomes and performance metrics.^{17,37,51,53,57} The likelihood of victory increases for the home team if they score the first goal.⁵⁵ However, after the introduction of VAR, the degree of home advantage in the CSL has somewhat declined.²⁷ Additionally, temperature and humidity posed greater challenges to the away team,¹² while playing at home can mitigate the adverse effects of poor air quality.¹¹ Nevertheless, one study presents a differing view, arguing that the home team is more affected by air pollution.⁵² This discrepancy may result from differences in the selection of performance metrics, the covariates considered, and the statistical methods employed.

Team strength is another crucial factor influencing match performance.^17,51 It generally encompasses both the team's own strength and that of the opponent, typically categorised by the final rankings of the season. Teams with greater inherent strength were less affected by other variables, thereby achieving superior performance in matches.^11,53 Stronger teams in the CSL were more likely to adopt fast-paced, high-intensity styles of play,⁵⁰ and matches between top teams tended to have reduced effective playing time.⁴⁷

Additionally, the match period (1^st or 2^nd half) and the phases of the season have also been shown to influence CSL match performance to some extent.^43,51 Notably, environmental factors were significant determinants of match performance as well.^11,12,35,52 According to Zhou et al.,¹³ technical performance was more affected by situational factors, while physical performance was more influenced by environmental conditions.

Limitations and recommendations for future research

First, the operational definitions of relevant metrics have not been standardised, and some studies have not standardised the data. For example, in-consistencies in speed zones make cross-study comparisons challenging, and data on distance and counts have not been converted to per-minute values, making them susceptible to errors due to variations in match duration. Future research should adopt standardised definitions of metrics, speed zones and ensure that data processing uses standardised methods (e.g., in accordance with FIFA's “Football Language”).

Second, contextual and environmental factors have been under considered. Football matches are complex systems influenced by multiple factors. Currently, the consideration of contextual factors is relatively consistent, mainly focusing on match location, team strength, and match outcomes. Research on environmental factors is limited to air quality, temperature, and humidity. Future studies should incorporate more contextual factors (e.g., match progression) and environmental factors (e.g., weather, altitude) into the analysis of CSL match performance.

Third, there is a lack of research on match phases from a micro perspective. Most current CSL match performance analyses adopt a macro perspective, treating the entire match data as a single sample. However, football matches consist of five phases: established attack, established defence, transition from attack to defence, transition from defence to attack, and set pieces.⁶² Future research should focus on a micro perspective, using match phases as samples for analysis. This approach would provide more practical guidance for coaches in training design and match preparation.

Fourth, research on the match performance analysis of tactics, goalkeepers, and set-pieces remains limited. Current studies on CSL tactics primarily rely on technical and physical performance metrics, utilising factor analysis to construct playing styles. However, these studies tend to overlook critical dimensions such as time, space, and the number of players involved. Additionally, there is an absence of studies specifically examining the match performance of goalkeepers and set-pieces in the CSL. Future research should aim to enrich the analysis of CSL match performance by incorporating these additional dimensions.

Conclusion

The CSL is receiving increasing attention in the research area, with prediction and comparative studies emerging as the predominant research types. Presenting descriptive data has become the fundamental paradigm in research. Research topics mainly focus on technical and physical performance analysis, with fewer studies on tactical performance. To ensure the reliability of results, it is crucial to use databases with validated reliability; if manual statistics are employed, data should be subject to reliability checks. Standardised data is expected to be a major trend in future research, as it effectively mitigates the impact of differences in match duration and possession rates.

The CSL is evolving towards a style that emphasises attacking play and high intensity. Foreign players play a crucial role in matches, with many of them primarily responsible for organising the attack, while domestic players are more often tasked with defensive duties. Additionally, substitutes have proven to be effective contributors during matches. Contextual factors are significant in influencing CSL match results and performance, with current research mainly focusing on factors such as team strength, opponent strength, match venue, and match outcomes. Stronger teams can mitigate the impact of various other factors. Additionally, research on environmental factors affecting match performance is in its early stages, with studies confirming that air quality, temperature, and humidity influence CSL players’ performance.

In future research, on one hand, the operational definitions of indicators and speed zones should be standardised according to FIFA guidelines, and data should be processed with standardisation techniques. On the other hand, it is important to broaden the scope and depth of research by exploring a wider range of contextual and environmental factors, enriching the analysis of tactical, goalkeeper, and set-piece performance, and focusing on match phase-related studies.

Footnotes

Acknowledgments

None.

Declaration of conflicting interests

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

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the Graduate Education Innovation Plan Project of Guangdong Province (2024SFKC_027) and the Sports Performance Analysis and Scientific Training Innovation Team (2024WCXTD030).

ORCID iDs

Zhaoyang Wang

Zunqi Niu

Hongyou Liu

References

Vilar

Araújo

Davids

, et al. The role of ecological dynamics in analysing performance in team sports. Sports Med 2012; 42: 1–10.

Laird

Waters

. Eyewitness recollection of sport coaches. Int J Perform Anal Sport 2008; 8: 76–84.

Liu

Gómez

MÁ

Lago-Peñas

, et al. Match statistics related to winning in the group stage of 2014 Brazil FIFA world cup. J Sports Sci 2015; 33: 1205–1213.

Liu

Giménez

, et al. Performance profiles of football teams in the UEFA champions league considering situational efficiency. Int J Perform Anal Sport 2015; 15: 371–390.

Mackenzie

Cushion

. Performance analysis in football: a critical review and implications for future research. J Sports Sci 2013; 31: 639–676.

Wright

Atkins

Jones

, et al. The role of performance analysts within the coaching process: performance analysts survey ‘the role of performance analysts in elite football club settings’. Int J Perform Anal Sport 2013; 13: 240–261.

Souza

López-Del Campob

Blanco-Pita

, et al. A new paradigm to understand success in professional football: analysis of match statistics in LaLiga for 8 complete seasons. Int J Perform Anal Sport 2019; 19: 543–555.

Cerda

Sanchez-Sanchez

Viejo-Romero

, et al. Characterisation of goal scoring patterns during open play related to zone pitch division and number of players involved in the 2018 FIFA world cup. Sensors (Basel) 2021; 21: 5601.

Sarmento

Marcelino

Anguera

, et al. Match analysis in football: a systematic review. J Sports Sci 2014; 32: 1831–1843.

10.

Sarmento

Clemente

Araújo

, et al. What performance analysts need to know about research trends in association football (2012–2016): a systematic review. Sports Med 2017; 48: 799–836.

11.

Sun

. Immediate impacts of air pollution on the performance of football players. J Sports Econom 2024; 25: 753–776.

12.

Yuan

Chang

Huang

, et al. Weather shocks and athlete performance: evidence from the Chinese soccer super league. J Clean Prod 2024; 451: 142080.

13.

Zhou

Hopkins

Mao

, et al. Match performance of soccer teams in the Chinese super league-effects of situational and environmental factors. Int J Environ Res Public Health 2019; 16: 4238.

14.

Mao

Peng

Liu

, et al. Identifying keys to win in the Chinese professional soccer league. Int J Perform Anal Sport 2016; 16: 935–947.

15.

Gai

Volossovitch

Lago-Peñas

, et al. Technical and tactical performance differences according to player's nationality and playing position in the Chinese football super league. Int J Perform Anal Sport 2019; 19: 632–645.

16.

Gai

Volossovitch

Leicht

, et al. Technical and physical performances of Chinese super league soccer players differ according to their playing status and position. Int J Perform Anal Sport 2019; 19: 878–892.

17.

Zhou

Calvo

Robertson

, et al. Long-term influence of technical, physical performance indicators and situational variables on match outcome in male professional Chinese soccer. J Sports Sci 2020; 39: 598–608.

18.

Palucci Vieira

Carling

Barbieri

, et al. Match running performance in young soccer players: a systematic review. Sports Med 2019; 49: 289–318.

19.

Liu

Hopkins

Gómez

MÁ

, et al. Inter-operator reliability of live football match statistics from OPTA sportsdata. Int J Perform Anal Sport 2013; 13: 803–821.

20.

Castellano

Alvarez-Pastor

Bradley

. Evaluation of research using computerised tracking systems (Amisco® and Prozone®) to analyse physical performance in elite soccer: a systematic review. Sports Med 2014; 44: 701–712.

21.

Gong

Cui

Gai

, et al. The validity and reliability of live football match statistics from champdas master match analysis system. Front Psychol 2019; 10: 1339.

22.

Gai

Leicht

Lago-Peñas

, et al. Key team physical and technical performance indicators indicative of team quality in the soccer Chinese super league. Res Sports Med 2018; 26: 158–167.

23.

Zhou

Zhang

Lorenzo Calvo

, et al. Chinese Soccer association super league, 2012–2017: key performance indicators in balance games. Int J Perform Anal Sport 2018; 18: 645–656.

24.

Gong

Cui

Zhang

, et al. Impact of technical and physical key performance indicators on ball possession in the Chinese super league. Int J Perform Anal Sport 2021; 21: 909–921.

25.

Goncalves

, et al. Data-driven team ranking and match performance analysis in Chinese football super league. Chaos Solitons Fractals 2020; 141: 110330.

26.

Liu

García -De-Alcaraz

Zhang

, et al. Exploring home advantage and quality of opposition interactions in the Chinese football super league. Int J Perform Anal Sport 2019; 19: 289–301.

27.

Han

Chen

Lago-Peñas

, et al. The influence of the video assistant referee on the Chinese super league. Int J Sports Sci Coach 2020; 15: 662–668.

28.

Gai

Leicht

Lago-Peñas

, et al. Physical and technical differences between domestic and foreign soccer players according to playing positions in the China super league. Res Sports Med 2018; 27: 314–325.

29.

Zhou

Gómez

MÁ

Lorenzo

. The evolution of physical and technical performance parameters in the Chinese soccer super league. Biol Sport 2020; 37: 139–145.

30.

Swartz

. The calculation of player speed from tracking data. Int J Sports Sci Coach 2023; 18: 516–522.

31.

Gai

Gong

, et al. Using passing network measures to determine the performance difference between foreign and domestic outfielder players in Chinese football super league. Int J Sports Sci Coach 2020; 15: 398–404.

32.

Zhong

Chmura

, et al. Match performance of football teams in different competition phases: analysis on the data of eight consecutive seasons in the Chinese super league. Front Psychol 2023; 13: 1069082.

33.

Chen

Zhai

, et al. Impact of absent crowds on technical and physical performances in the Chinese soccer super league. Front Psychol 2022; 13: 959213.

34.

Gong

Zhou

Gómez

MÁ

, et al. Identifiability of Chinese football teams: a complex networks approach. Chaos Solitons Fractals 2023; 166: 112922.

35.

Wei

Liu

, et al. Temperature and outdoor productivity: evidence from professional soccer players. J Asian Econ 2023; 87: 101622.

36.

Yao

Wang

Zhu

, et al. Goal or miss? A Bernoulli distribution for in-game outcome prediction in soccer. Entropy (Basel) 2022; 24: 971.

37.

Han

Yang

Pan

, et al. The influence of removing home advantage on the Chinese football super league. BMC Sports Sci Med Rehabil 2022; 14: 208.

38.

Quan

, et al. The influence of technical performance indicators on the results of the Chinese Football Super League at different stages of the season-based on evidence from the 2010–2019 seasons. Proc Inst Mech Eng P J Sport Eng Technol. 30 April 2023. Epub ahead of print. DOI: https://doi.org/10.1177/17543371231170065.

39.

Guan

Swartz

. Acceleration and age in soccer. Int J Sports Sci Coach 2024; 19: 1035–1041.

40.

Lago-Penas

Gomez-Ruano

Yang

. Styles of play in professional soccer: an approach of the Chinese soccer super league. Int J Perform Anal Sport 2018; 17: 1073–1084.

41.

Zhao

Liu

. Factors that influence actual playing time: evidence from the Chinese super league and English premier league. Front Psychol 2022; 13: 907336.

42.

Ruan

Shen

, et al. Quantifying the effectiveness of defensive playing styles in the Chinese football super league. Front Psychol 2022; 13: 899199.

43.

Feng

Chmura

Liu

. Are there meaningful changes in the performance of professional soccer teams from the first half of the match to the second? Evidence from the data of eight consecutive seasons of the Chinese super league. Int J Perform Anal Sport 2024; 24: 626–640.

44.

Ruan

Gómez

MÁ

, et al. Analysis of defensive playing styles in the professional Chinese football super league. Sci Med Footb 2022; 7: 279–287.

45.

Zhou

Lago-Peñas

Lorenzo

, et al. Long-Term trend analysis of playing styles in the Chinese soccer super league. J Hum Kinet 2021; 79: 237–247.

46.

Pan

Han

, et al. Exploring the impact of professional soccer substitute players on physical and technical performance. BMC Sports Sci Med Rehabil 2023; 15: 143.

47.

Zhao

Zhang

. Effective playing time in the Chinese super league. Int J Sports Sci Coach 2021; 16: 398–406.

48.

Zhou

Lorenzo

Gómez

MÁ

, et al. Players’ match demands according to age and playing position in professional male soccer players. Int J Perform Anal Sport 2020; 20: 389–405.

49.

Liu

Yang

Chen

, et al. Impact of possession and player position on physical and technical-tactical performance indicators in the Chinese football super league. Front Psychol 2021; 12: 722200.

50.

Kong

Zhang

Zhou

, et al. The evaluation of playing styles integrating with contextual variables in professional soccer. Front Psychol 2022; 13: 1002566.

51.

Zeng

Zhang

. Analysis of influencing factors of passes in the Chinese super league. BMC Sports Sci Med Rehabil 2022; 14: 180.

52.

Wang

Yang

. The impact of air pollution on game outcome. J Sports Econom 2024; 25: 557–582.

53.

Jimenez

Bustamante-Sanchez

Kelly

, et al. Influence of the situational variables on the performance of the teams competing in the Chinese super league. Int J Perform Anal Sport 2022; 22: 455–466.

54.

Liu

Chen

Micha

, et al. Tactical factors that influence shooting success in the Chinese Football Association Super League. Proc Inst Mech Eng P J Sport Eng Technol. 16 February 2024. Epub ahead of print. DOI: https://doi.org/10.1177/17543371231219730.

55.

Liu

García-De-Alcaraz

Wang

, et al. Impact of scoring first on match outcome in the Chinese football super league. Front Psychol 2021; 12: 662708.

56.

Zong

Shen

, et al. Characterizing player's playing styles based on player vectors for each playing position in the Chinese football super league. J Sports Sci 2022; 40: 1629–1640.

57.

Zhao

Zhang

. Eigenvalues make the difference-A network analysis of the Chinese super league. Int J Sports Sci Coach 2020; 15: 184–194.

58.

Zhang

Gong

Tao

, et al. The influence of tactical formation on physical and technical performance across playing positions in the Chinese super league. Sci Rep 2024; 14: 2538.

59.

Jiang

Cui

. Effect of tournament format change on team performance of Chinese football super league during COVID-19 pandemic. Rev Int Cienc Deporte 2021; 18: 72–85.

60.

O’Donoghue

Holmes

Data analysis in sport. New York: Routledge, 2015, p. 4.

61.

Gómez

MÁ

Lago-Peñas

Pollard

. Situational variables. In: Gómez

M-Á

Lago-Peñas

Pollard

(eds) Routledge handbook of sports performance analysis. London: Routledge, 2013, pp.259–269.

62.

Hewitt

Greenham

Norton

. Game style in soccer: what is it and can we quantify it? Int J Perform Anal Sport 2016; 16: 355–372.