https://orcid.org/0000-0003-1022-5206
Figures
Abstract
The rapid rise in elite women’s football has increased the demand for female specific research to enable more accurate and appropriate assessments of tactical performance. Thereby, this study aimed to describe goal scoring in relation to different attacking styles during a Women’s Super League (WSL) season. Specifically, 1179 attacking sequences leading to shots on target performed by all 12 teams in the 2021/22 WSL season were analysed. The style of attack for each attacking sequence was characterised by research guided key performance indicators and recorded with the outcome of the subsequent shot on target. Descriptive results indicated that most shots (27.23%) were originating from combinative organised attacks, while set plays accounted for the most goals (27.08%), with fast organised attacks demonstrating the best goal conversation rates (53.33%). Outcomes of a chi-square test highlighted a significant (but weak) association between attacking styles and shot outcomes ( = 9.87, P = 0.043) in the considered WSL season, with shots originating from set plays resulting in significantly more goals than expected (AR = 2.45). Overall, the results can be useful for practitioners when formulating tactical game plans and training sessions, while also providing multiple opportunities for future research in tactical analysis of women’s football.
Citation: Craven L, Oxenham P, Ranaweera J (2025) Analysis of attacking styles and goal-scoring in the 2021/22 Women’s Super League. PLoS ONE 20(2): e0318929. https://doi.org/10.1371/journal.pone.0318929
Editor: Anwar P. P. Abdul Majeed, Sunway University, MALAYSIA
Received: August 18, 2024; Accepted: January 23, 2025; Published: February 6, 2025
Copyright: © 2025 Craven et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data relevant to this paper are available from OSF at https://doi.org/10.17605/OSF.IO/RQUVD.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Women’s football has been recognised as the fastest growing sport in the world, with the quality, investment and interest experiencing dramatic exponential growth [1]. Such rising interest and engagement in women’s football have been notable in both domestic and international competitions such as the Women’s Super League (WSL), the Women’s FA Cup, the European Championships and the FIFA World Cup [2, 3]. Increased fan interaction and a subsequent rise in funding has enabled clubs to invest more heavily into supporting coaches and players [4]. This has facilitated greater access to performance analysis in the female game and greater levels of integration of the discipline within women’s teams [5]. Research illustrates that the implementation of performance analysis within day-to-day practice has been directly linked to maximising the growth of women’s football [6]. Subsequently, leading to a comparable increase in performance analysis research within women’s football [7]. However, the extent in which this can be applied to women’s football heavily depends on the social-contextual environment such as the limited female-specific data and financial resources available to clubs [8]. The level of support and overall professionalism varies largely across elite women’s football, with working conditions for coaches, players and support staff being seriously inadequate in some cases [9]. This will negatively impact the extent in which analysis is implemented and the subsequent level of support analysts can provide towards both the coaching process and player development [10]. Consequently, providing research that can be directly applied into the elite women’s game will help support both coaches and analysts.
Analysing tactical aspect of a team’s performance has been noted as fundamental in providing an extensive and accurate assessment and enhancement of performance [11]. Tactical analysis considers a variety of key performance indicators that are likely to result in successful performance and winning games [12]. Previous male-focused research examined individual elements of a team’s tactical performance from attacking and defensive perspectives. Attacking variables considered as vital for successful performance within this research include ball possession [13], passing variables [14] and a variety of shooting variables [15]. From a defensive perspective, variables such as ball recovery [16], defensive pressure [17] and defensive balance [18] have been considered in outlining the tactical trends in a team’s performance. This has allowed analysts to quantify the patterns and trends that constitute successful performances by individuals and teams [19].
At the centre of recent male-focused tactical research has shifted towards analysing playing styles instead of individual performance indicators to produce more effective summaries of tactical performance [20]. Styles of play are defined as the “tactical behaviours” employed by a team that aims to “achieve the attacking and defensive objectives in the game” [21]. It has been noted that assessing a team’s attacking proficiency aids in forming an accurate assessment of that team’s tactical set up and extent of performance success [22]. Interestingly, the low-scoring, invasive nature of football leads to attacking performance success often being centred around the frequency of goals scored and how this impacts wins achieved [23]. While scoring a goal is recognised as the most successful and influential action during a game, it should be considered that only 1% of team possessions lead to a goal being scored [24]. It is therefore vital to transfer the focus of tactical analysis from the simplistic view of goal frequencies to a broadened overview of the construction of goal-scoring attacks [15]. As research suggests, the factors that make up the construction of attacking play determine a team’s tactical style and are directly linked to the frequency and success of goal-scoring [25].
Hughes and Franks [24] categorised a team’s tactical attacking style as “direct” or “possession” and this categorisation has been subsequently integrated in contemporary literature [15, 23, 26]. A "direct" attacking style is characterised as using a limited number of long and or vertical passes to transition the ball quickly to the opposing goal [15]. In contrast, a "possession" attacking style utilises a greater number of passes over a longer duration to progress the ball towards the opposition goal [27]. While this helps to distinguish between the two opposing strategies, research has evolved and attacking styles are now categorised with greater granularity. These studies differentiated between "counterattacks" produced in offensive transitions [28], "direct attacks" characterised by long passes [29], "fast attacks" that use quick and vertical passes [30] and "combinative attacks", characterised by long ball possessions [31]. However, the actual implementation of each team’s style of play not only depends on the team’s preparation but it is strongly influenced by the interaction with the opponent’s tactical behaviour, as well as by the contextual variables.
To ensure contextual variables are considered, Aranda et al. [32] designed the ‘REOFUT’ framework—a research-guided observational tool that aims to categorise a team’s style of play. The ‘REOFUT’ framework outlines the combination of variables and actions that exist in the start, development, penultimate and final stages of an attack, all of which determine a team’s attacking style. Research has considered the variables that formulate the start of an attacking sequence, including type of start [33], starting zone [34] and initial behaviour from both the attacking and defensive team [35]. Similarly, variables that make up the development of a team’s offensive play have been evaluated in the REOFUT’ framework. These include key performance metrics such as: the number of passes [36], type of passes [37], pass penetration [38], duration [39] and possession width [40]. Finally, both the penultimate and final action have also received attention in the research, with variables such as assist zone [25], assist action [31], type of goal-scoring action [36] and goal-scoring zone [25] being scrutinised. The combination of previously mentioned attacking variables determines the type of attack performed by a team during offensive phases of play and has been linked to variations in the frequency and success in shots on goal taken [25, 34]. A prevalent finding in previous research is that attacking performance is heavily focused on men’s football and fails to consider the variations that exist within female football [41, 42], and as a result could limit the relevance when applying conclusions in analyst practice in the women’s game [42]. Practitioners and analysts working within women’s football are often required to consult and utilise the findings from male-focused research due to the lack of gender-specific research that exists, particularly from a tactical perspective [41]. While research surrounding men’s football has successfully transitioned towards analysing attacking performance from a broadened, stylistic, tactical perspective, female-specific research often reverts to analysing simplistic individualised attacking variables [43]. Given the previously discussed significance of goal scoring and attacking patterns of play, it is crucial for practitioners operating in women’s football to build upon existing evidence to gain a more comprehensive understanding of tactical performance. In this context, expanding and deepening applied research on women’s football is essential in extracting the tactical insights needed by practitioners to execute strategic decision-making processes [44], while also helping to reduce the existing disparity in tactical analysis compared to the men’s game. Therefore, as an initial step to help bridge those gaps in current literature, this study aims to describe goal scoring based on different attacking styles in the Women’s Super League (WSL) tournament. Specifically, the study initially examines literature to identify and characterise different attacking styles and associated variables relating to women’s football. Next, guided by those findings, we provide a descriptive account of goal scoring that had occurred based on the attacking styles in the considered season. Finally, we examine the potential relationship between the style of attack and outcome of the resulting shot on target during the considered WSL season.
2. Methodology
2.1. Operational definitions
The key performance indicators (KPIs) and operational definitions used for this study were established in three stages. Initially, a literature review was conducted (by the lead author LC) to identify KPIs and operational definitions associated with attacking styles of play. Guided by the previously stated ‘REOFUT theoretical framework’, the style of each attacking sequence was determined based on a combination of key performance indicators [32]. The REOFUT framework outlines and details the key action variables associated with five main attack types: combinative organised attack, fast organised attack, direct organised attack, counterattack and set plays and is used in attacking research [15, 23, 31]. Action variables presented in the start and development phases of an attacking sequences, as outlined by the ‘REOFUT theoretical framework’ assist in determining the type of attacking style. A breakdown of each attacking style and associated action variables are presented in Table 1. Importantly, specific evidence from the literature that were used to guide the categorisation of the 5 attack styles from the action variables specified in Table 1 are provided in the supplementary materials (S1 Table). Moreover, since ‘set plays’ are characteristically different to open play attacking styles, characteristics and associated action variables are defined by a limited range of literature (see S1 Table). Moreover, if the attacking sequence and associated variables differed from the guided combination of variables from the ‘REOFUT theoretical framework’ and related research, the attack was labelled as a ‘nonconforming attack’. Thereafter, while each action variable was drawn from the start and development phases of an attack by the ‘REOFUT theoretical framework’, a further literature review was conducted (by the lead author LC) to determine the validity of these variables and provide a more comprehensive overview. Each variable was further divided into categorical groups as guided by research and presented in Table 2. Following the coding of each attacking sequence, the outcome of the resulting shot on target was noted either as ‘Goal’ or ‘No Goal’ as guided by research (definitions presented in Table 3) [45]. Once the KPIs were identified and operational definitions were validated through a review of the existing literature, they were presented to a panel of experts. The panel further validated the definitions, KPIs, action variables, and the categorisation of attack styles, confirming their relevance and accuracy. The expert panel was made up of football coaches (n = 3) who had a combined experience of 24 years working in women’s football. During each individual interview (n = 3), KPIs were presented to each coach firstly by outlining the definition (and visual schematic where relevant) of each KPI and the associated action variables. To further strengthen the validation, selected video footage pertaining to the attack sequences were also used to guide engagement with the coaches. Specifically, each video clip was played twice for a coach (or more if requested) whereby, final agreement was reached.
2.2. Data collection
In terms of the overall study, this investigation utilised a descriptive design to describe the attack styles that led to a shot on target by the twelve teams that competed in an English Women’s Super League (WSL) season. Subsequently, all 132 games played in the 2021/2022 WSL season were represented in the sample with 1179 attacking sequences being obtained from those games. Clips of each attacking sequence of the 132 games played in the 2021/2022 WSL from full game footage was obtained from match analysis software Wyscout (Chiavari, Italy). In relation to postgraduate dissertation ethics approval criteria, the aforementioned data collection process (and overall research project) was approved by the Faculty of Sport, Technology and Health Sciences, St Mary’s University ethics committee. Moreover, written consent was obtained from the video data provider prior to notational analysis. As specified below, guided by the definition provided by Pollard and Reep [46] each clip began at the start of a team possession. All types of attack from Table 1 utilised this definition. “A team possession commences when a player gains ball possession in a controlled and deliberate manner by any means excluding receiving the ball from a player from the same team. This action may be followed by an individual action such as dribbling and carrying the ball or a series of passes between one team. The possession ends when one of the following occurs: 1) a goal is awarded; 2) the ball goes out of play; 3) the ball touches an opposition player to change the sequence of play (a tackle, interception or save). Brief touches that fail to alter the direction of play or sequence are excluded.”
The attacking sequence was judged to be complete when one of the following occurred: 1) a goal had been awarded by the referee; 2) the ball travelled out of play (save or block) [47, 48]. During the analysis of each attacking sequence, a coding system was developed to guide the process of recording the notational data. This was done to ensure consistency when coding each clip by the same individual. The consequent data were logged in an observational instrument, designed and formatted in Microsoft Excel Version 16.0.67 (Microsoft Cooperation, United States) allowing for the filtration and processing of data for further analysis. To ensure consistency when coding each attacking sequence, operational definitions and visual schematics were used as guides (see Tables 2 and 3; Figs 1 and 2). Following coding of each attacking sequence, the outcome of the resulting shot was noted either as ‘Goal’ or ‘No Goal’ as guided by research (definitions presented in Table 3) [45].
2.3. Reliability
As the primary data collection instrument, the intra-rater reliability of the lead researcher when performing the notational analysis was done by coding a random sample of 118 attacks (10% of the total) on two separate occasions. Importantly, the ratings rounds were separated by a 1-week washout period [49]. To quantify the intra-rater reliability, Cohen’s Kappa (k) statistic was calculated for: type of attack (k = 0.96), type of start (k = 0.96), opposition defensive balance (k = 0.95), type of passes (k = 0.95), number of passes (k = 1), percentage of penetrative passes (k = 0.95), starting zone of attack (k = 0.96), width of attack (k = 0.97) and duration of attack (k = 0.99), resulting in an overall Kappa value of 0.97. Guided by literature, agreement was interpreted as slight (0.01–0.20), fair (0.21–0.40), moderate (0.41–0.60), substantial (0.61–0.80) and almost perfect (0.81–0.99) [50, 51]. Thereby, the overall notation analysis illustrated almost perfect intra-reliability, indicating that quality of the generated data was suitable for the subsequent analysis.
2.4. Statistical analysis
2.4.1. Descriptive and inferential statistics.
Descriptive analysis of the sample was executed by calculating the relative frequencies and percentages for each style of attack. In terms of the inferential analysis, the potential relationship between style of attack and outcome of the resulting shot on target of the 12 teams that competed in the 2021/22 WSL season was analysed using a chi-square test of independence, with an alpha level of 0.05 considered as the threshold for statistical significance. Importantly, based on the degrees of freedom of the test, effect size of the chi-square test was determined using Cramer’s V and interpreted using the thresholds specified by Cohen [52]. Moreover, as a measure of post-hoc analysis, adjusted standardised residuals (AR) were used to quantify the relationships between each attacking style and shot outcome. Specifically, in accordance with the norms and guided by similar recent research [53], magnitude of the residuals was assessed in relation to the criterion ± 2.0 (1.96), where AR values above or below the thresholds were deemed to signify significantly greater or lower observed counts than the expected frequencies. Finally, all data analysis reported in this study was conducted using R programming language within the RStudio integrated development environment (IDE).
2.4.2. Sample sizes.
To guide sample size requirements, an a-priori sample size calculation was conducted using G*power 3.1.9.2. Subsequently, at a pre specified alpha level 0.05, 80% power and with 4 degrees of freedom, the calculation illustrated that a minimum 133 shots on target/attacking sequences were required to detect a medium effect size of 0.3 [54]. Therefore, since the initial sample of shots on target considered in the study (n = 1179) was much larger than that specified by the priori calculation, the expected power for the analysis was achieved. Thereby, minimising the formulation of potential Type II errors.
3. Results
From the 132 WSL games that were played in the 2021/22 season, 1179 attacking sequences that led to a shot on target were analysed. Open plays accounted for 77.95% of the total shots on target (n = 919), whereas the remaining 22.05% were from ‘set plays’ (n = 260). As presented in Fig 3, considering all shots pertaining to each style of attack, most shots on target (n = 321) were from combinative organised attacks (27.23%), whereas fast organised attacks (n = 15) accounted for the lowest proportion (1.27%) of shots on target. Interestingly, set plays (22.05%) and counterattacks (20.87%) had resulted in approximately similar proportions of shots on target. Moreover, 21.46% of the shots (n = 253) considered in the sample did not meet the criteria of the 5 attack types and were categorised as ‘nonconforming’. However, as reported in the immediate paragraphs, for transparency and consistency, those 253 shots were considered when describing the sample, but were omitted when conducting and reporting outcomes of the inferential analysis.
In terms of goals scored, 32.57% (384) of the total attacking sequences (n = 1179) resulted in a goal. Specifically, as presented in Fig 4, 27.08% (n = 104) of all those goals were scored from set plays, with combinative organised attacks also contributing to 26.04% (n = 100) of goals in the 2021/2022 WSL season. Interestingly, as expected due to the lowest number of shots on target, the least proportion of goals were created from fast organised attacks, contributing only to 2.08% (n = 8) of the total goals scored in the season.
When considering conversation rates of the shots on target, as highlighted in Fig 5, although contributing to the lowest percentage of goals scored in the season, 53.33% of the shots on target from fast organised attacks resulted in goals. Similarly, among the attack styles contributing to larger proportions of the total shots, set plays demonstrated the best conversation rates with 40% of those shots leading to goals. However, while most attack sequences had adopted combinative organised attacks to create scoring chances, the results indicate that only 31.15% of those sequences yielded positive outcomes (i.e., goals). Additionally, the results also indicate that direct organised attacks had the least conversation rate (27.38%) of shots on target resulting in goals during the 2021/2022 WSL season.
Regarding inferential statistics, guided by relevant literature [21, 41], as specified previously, the 253 ‘nonconforming’ shots categorised in the sample were excluded for the hypothesis testing, resulting in 926 shots on target being considered for the final analysis. Thereby, outcomes of the chi-square test demonstrated a significant association between the style of attack and outcome of the shot on target, χ2(4) = 9.87, p = 0.043; V = 0.1. However, the small effect size indicated that the observed association between the variables was weak. Moreover, based on the adjusted standardised residuals (AR), the results further suggested that the observed number of goals scored was more than expected for set plays (AR = 2.45). Interestingly, none of the other residuals yielded significant outcomes, suggesting minimal contributions to the association observed between the style of attack and outcome of the shot on target.
4. Discussion
This descriptive study was aimed at exploring the attacking styles utilised in the 2021/22 WSL season and their relationship with the resulting shot on target. Specifically, 1179 attacking sequences (during 132 games) that led to shots on target were analysed, where most shots (27.23%) were originating from combinative organised attacks, while only 1.27% shots were from fast organised attacks. The sample also highlighted that 32.57% of those shots resulted in goals, with most yielding from set plays (27.08%) and combinative organised attacks (26.04%). Interestingly, although only 2.08% of goals in the sample were from fast organised attacks, the results highlighted that 53.33% of the attacks originating from that style resulted in goals, with set plays also demonstrating more than 40% goal conversation rates. Importantly, the inferential analysis highlighted the prevalence of a significant relationship (p = 0.043) between style of attack and outcome of the shot on target in the 2021/22 WSL season, with set plays resulting in significantly (AR = 2.45) more goals than expected. Nevertheless, since the small effect size (V = 0.1) indicated that the observed association was weak and because the study design adopted is primarily descriptive, outcomes from high-quality observational studies may be necessary prior to making strong generalisable judgements from the results.
4.1. Attacking styles and shots on target
As specified in the introduction, since there is limited literature focused on examining the attacking styles in women’s football, the remaining sections of this paper will rely on male focused research (although questions can be raised on the direct relevance) to compare the key outcomes formulated from this study. Thereby, it was firstly important to investigate the potential reasons for most shots on target in the 2021/22 WSL season examined in this study to be manifested from combinative organised attacks. Explicitly, as highlighted in Table 1 (see number of passes KPI), combinative organised attacks characteristically contain longer passing sequences prior to a shot occurring. While early football research concluded that short passing sequences resulted in a greater number of goals scored [55] such judgements have been contested with more contemporary studies finding attacking sequences consisting of three passes or more resulting in greater number of goals being scored [56]. Specifically, high number of consecutive passes has been associated with more effective penetration of space and exploitation of imbalances in the opponent’s defence [57]. The constant movement of possession between players of the attacking team aims to force gaps within the opposition’s defensive structure. Precise passes are then played to the player best positioned to execute a successful attempt on goal [20]. Players that attempt shots in more space and in closer proximity to the goal have a greater chance of the resulting shot being on target and achieving a goal [58]. Research has termed this area ‘the scoring pentagon’ outlining its importance in enhancing a team’s shooting accuracy, conversion rate and overall goal-scoring proficiency [39]. Therefore, combinative organised attacks may characteristically cultivate a larger amount of those high percentage shots and could justify the high frequency of shots on target noted in the present study. It should be considered that the successful execution of longer, low risk passing sequences often heavily relies on the technical abilities of each player in this system [59]. More technically proficient players often play for higher ranked teams allowing for a more successful implementation of combinative organised attacks and could provide reason as to why those teams execute more shots on target, score more goals and win more games [60]. Hence, such judgements highlight that the current study can be extended to the micro-level, where associations between the attacking styles based on team rankings in the WSL can be examined. However, Mitrotasios et al. [39] compared the implementation and effectiveness of different attacking styles, where the results suggested that combinative organised attacks are more frequently employed in male football compared to the women’s game.
Conversely, the results highlighted that fast organised attacks contributed the lowest proportion of shots on target (1.27%) and goals (2.08%) in the considered sample. This result supports previous research, which found that fast organised attacks had resulted in less goals and goal-scoring opportunities [61]. Moreover, González-Rodenas et al. [62] found that despite fast attacks achieving more offensive penetration, they failed to create more goal scoring opportunities compared to combinative attacks. Perhaps, this is because the low number of shorter passes executed across the width of the pitch during a short period is likely to limit to progression of the ball towards the opponent’s goal. Subsequently, enforcing shots to be taken further away from the opponent’s goal, leading to more challenging positions to hit the target successfully [15]. Specifically, it is likely that those shots will be taken outside ‘the scoring pentagon’, resulting in reduced effectiveness of the shot. Thus, reducing the frequency of shots on target achieved and goals scored [39]. Nevertheless, although fast attacks may lead to limited instances of direct shots on target, they have been linked to the creation of successful attacking sequences like set plays (direct free-kicks, penalties and corner kicks) [30].
4.2. Goals from attacking styles
The present study highlights that set plays led to the most goals (27.08%), while also resulting in effective conversion rates (40%) in the 2021/22 WSL season. The goal-scoring effectiveness of set plays has been heavily acknowledged within research that had examined multiple competitions (in alignment with current results), where they appear to account for approximately 25–33% of goals scored [30, 63]. Importantly, the goal-scoring effectiveness from set plays have been examined within women’s football [36, 43, 64, 65]. Specifically, Beare and Stone [64] summarised that corner kicks occur more frequently in women’s football and were more efficient in producing goals compared to male football. However, research further indicates that less goals were produced from set plays, such as corners during international tournaments in women’s football [65, 66]. The contrasting nature of the prior judgements may perhaps indicate the impact of physical differences of players in WSL teams to goals scored from set plays. Specifically, we hypothesise that WSL teams with greater representation of international players may pose a physical dominance during set plays. This possibly rationalises the large proportion of goals from set plays observed in the WSL season considered in this study. Conversely, this effect may be minimised during international tournaments due to uniform physical representation of players across national teams. Nevertheless, further research that examines the relationships between attacking styles, shot outcomes and the physical characterises of players in WSL teams (e.g., to examine set play performance of WSL teams on goal scoring opportunities in relation to team quality—i.e., composition of international versus non-international players) may be necessary to evaluate such prior hypothesis. Furthermore, research also indicates that set plays characteristically allow a greater number of attacking players to take a shot at goal, direct passes into ‘the scoring pentagon’, shots on a proximity to the goal and enforces less time pressure on the attacker to make an effective decision regarding a strategic attacking play—all of which may contribute to high goal conversion rates [35].
In terms of conversation rates, Hughes and Franks [24] noted that shorter, more direct attacks had higher shot conversion rates than longer sequences. Perhaps this justifies the outcome that fast organised attacks had the highest conversation (53.33%) rate among the attack styles. Fast organised attacks aim to use quick, vertical passes to exploit gaps and imbalances within the opposition’s defensive structure [23]. Explicitly, high percentage of penetrative passes, noted as extremely effective within research, enables the attacking team to move beyond the defensive unit and retain ball possession to have an attempt at the goal [28, 67]. Although this may occur infrequently during quick, wide attacks that contain shorter passes, if an opportunity arises, shots can be made within ‘the scoring pentagon’ [68]. Thereby, potentially maximising the conversion rate of the shots on target taken during those fast organised attacks. However, such judgements should be undertaken with care due to the low proportion of goals (2.08%) originating from fast organised attacks in the overall sample of shots considered in this study.
4.3. Relationship between attacking style and shot on target
From an inferential analysis perspective and as specified previously, the results indicated a significant association (p = 0.043) between the style of attack and the shot outcome within the 2021/22 WSL season. Interestingly, perhaps aligning to the above result and although the outcome of the shots were not considered, a study by González-Rodenas et al. [62] found significant differences between the attacking styles utilised to create goal scoring opportunities in male football. Specifically, in comparison to combinative attacks, counterattacks were more effective at creating scoring opportunities, while direct attacks were less likely to create chances and no differences were observed in relation to fast attacks. Thereby, when comparing such outcomes to the findings of the current study, from a purely statistical perspective and as discussed previously, because only shots originating from set plays led to significantly more goals than expected (AR = 2.45), further research outcomes based on high-quality observational designs is necessary (mainly because designing a randomised control trial for this if type of tactical research may not be practically viable) to examine justifications for the non-significant contributions of combinative organised (AR = -1.29), counterattack (AR = -0.69), direct organised (AR = -1.33) and fast organised (AR = 1.6) attacks to shot outcomes. Such judgements are further justified due to the small effect size (V = 0.1) of the chi-square test, where replication studies may be necessary to examine if the observed effect is consistent when examining attacking sequences in other WSL seasons. Furthermore, research heavily acknowledges the influence of contextual factors on the goal-scoring proficiency of different attack styles [34]. Hence, contextual dimensions like match location, opposition level, team level, match status, time of game, etc. associated with a team’s tactical style, can influence performance trends, patterns and overall statistical significance, all of which are important to consider when generating specific practical conclusions [31, 32].
4.4. Study limitations
Beyond statistical considerations pertaining to small effect sizes of the chi-square test discussed previously, one of the key limitations of the current study was the categorisation and the subsequent exclusion of ‘nonconforming attacks’ from the final analysis. Specifically, ‘nonconforming attacks’ accounted for 21.46% of all attacking sequences and 18.23% of goals considered in the sample. this represents a considerable proportion of attacks and goals scored within the considered WSL season and therefore limits the formulation of strong generalisable conclusions from the results. Importantly, we didn’t consider resolving this issue in this study since our objective was to provide an initial descriptive account of the attacking styles and their potential relationships to shot outcomes. However, we believe this situation creates an additional opportunity for future research to explore how different data analysis approaches like machine learning models can be used to categorise those attacks not meeting the criteria of the five styles considered in the current study. Specifically, if the objective is to categorise the nonconforming attacks to the existing five styles, perhaps a supervised classification approach can be used [69]. Alternatively, an unsupervised classification technique like clustering may also be used to generate fresh clusters of attack styles based on the KPI and action variables specified in Table 1 [70]. However, such machine learning approaches may need to utilise techniques like ‘one-hot encoding’ to transform the categorical data of the KPIs conditions into numeric representations suitable for classification or clustering. Subsequently, we have provided the raw dataset of the current analysis to support such future research.
5. Conclusion
Overall, the outcomes of this study indicate a significant association between attacking styles and shot outcomes in the 2021/22 Women’s Super League (WSL) season. From a coaching perspective, due to the highest prevalence of shots in the sample, if the offensive tactics in the WSL are focused on creating more shots on target, perhaps there is an opportunity to formulate attacking sequences based on combinative organised attacks. Nevertheless, despite generating the least shots on target and goals in this study, the high conversion rates of shots may indicate possibilities to adopt fast organised attacks in tactical scenarios aimed at maximising shooting opportunities. Specifically, while it may not lead to direct shooting chances, attacks originating from fast organised attacks may create opportunities for set plays. Thereby, potentially resulting in positive outcomes since the study outcomes indicate that significantly more goals than expected were scored from set plays in the 2021/22 WSL season. Finally, owing to the small effect size of the statistical test and because the adopted study design is mainly descriptive, future research should focus on implementing replication studies (perhaps considering different WSL seasons), alongside high-quality observational designs to further validate the current findings (e.g., observed associations). Additionally, there are also opportunities to investigate micro-level relationships between attacking strategies, shot outcomes and the physical characterises of players in WSL teams. Finally, future research can also explore the implementation of machine learning models to classify or cluster attacking strategies based on the data generated from the current study.
Supporting information
S1 Table. Supplementary information for this manuscript.
https://doi.org/10.1371/journal.pone.0318929.s001
(DOCX)
References
- 1.
Petty K, Pope S. A New Age for Media Coverage of Women’s Sport? An Analysis of English Media Coverage of the 2015 FIFA Women’s World Cup. Women’s Football in a Global, Professional Era: Emerald Publishing Limited; 2023. p. 205–19.
- 2.
Okicic M. Motivations for attending the FA Women’s Super League (WSL) games. Birkbeck Sport Business Centre; 2023.
- 3.
Cootes I. 4 ways to ensure WSL attendance records continue next season: Optus Sport; 2024 [https://sport.optus.com.au/news/womens/os58315/wsl-attendance-up-by-173-percent-in-202223-4-things-needed-to-keep-crowds-growing#:~:text=Heading%20into%20the%20final%20matchday,of%201931%20fans%20per%20match.
- 4.
Culvin A. Football as work: The new realities of professional women footballers in England: University of Central Lancashire; 2019.
- 5. Pielichaty H. Pleasure and the Sanctuary Paradox: Experiences of girls and women playing soccer. International Review for the Sociology of Sport. 2019;55(6):788–806.
- 6. Başkaya G. Performance Analysis of 2022 Women’s European Football Champion England National Team. Turkish Journal of Sport and Exercise. 2022;24(3):319–28.
- 7. Kubayi A. The Creation of Goal-Scoring Opportunities at the 2019 FIFA Women’s World Cup. Journal of Human Kinetics. 2022;82:165–72. pmid:36196349
- 8. Clarkson BG, Culvin A, Pope S, Parry KD. Covid-19: Reflections on threat and uncertainty for the future of elite women’s football in England. Managing Sport and Leisure. 2022;27(1–2):50–61.
- 9. Culvin A. Football as work: the lived realities of professional women footballers in England. Managing Sport and Leisure. 2023;28(6):684–97.
- 10. Serpell BG, Harrison D, Dower R, Cook CJ. The under representation of women coaches in high-performance sport. International Journal of Sports Science & Coaching. 2023;18(4):1320–32.
- 11. Hughes M, Lovell T. Transition to attack in elite soccer. Journal of Human Sport and Exercise. 2019;14(1):236–53.
- 12. Perl J, Memmert D. A Pilot Study on Offensive Success in Soccer Based on Space and Ball Control—Key Performance Indicators and Key to Understand Game Dynamics. International Journal of Computer Science in Sport. 2017;16(1):65–75.
- 13. Merlin M, Cunha SA, Moura FA, Torres RdS, Gonçalves B, Sampaio J. Exploring the determinants of success in different clusters of ball possession sequences in soccer. Research in Sports Medicine. 2020;28(3):339–50. pmid:31973582
- 14. Rein R, Raabe D, Memmert D. “Which pass is better?” Novel approaches to assess passing effectiveness in elite soccer. Human Movement Science. 2017;55:172–81. pmid:28837900
- 15. Gonzalez-Rodenas J, Mitrotasios M, Aranda R, Armatas V. Combined effects of tactical, technical and contextual factors on shooting effectiveness in European professional soccer. International Journal of Performance Analysis in Sport. 2020;20(2):280–93.
- 16. Barreira D, Garganta J, Guimarães P, Machado J, Anguera MT. Ball recovery patterns as a performance indicator in elite soccer. Proceedings of the Institution of Mechanical Engineers, Part P: Journal of Sports Engineering and Technology. 2013;228(1):61–72.
- 17. Low B, Rein R, Raabe D, Schwab S, Memmert D. The porous high-press? An experimental approach investigating tactical behaviours from two pressing strategies in football. Journal of Sports Sciences. 2021;39(19):2199–210. pmid:33982645
- 18. Tenga A, Mortensholm A, O’Donoghue P. Opposition interaction in creating penetration during match play in elite soccer: evidence from UEFA champions league matches. International Journal of Performance Analysis in Sport. 2017;17(5):802–12.
- 19. Herold M, Kempe M, Bauer P, Meyer T. Attacking key performance indicators in soccer: current practice and perceptions from the elite to youth academy level. Journal of sports science & medicine. 2021;20(1):158. pmid:33707999
- 20. Lopez-Valenciano A, Garcia-Gómez JA, López-Del Campo R, Resta R, Moreno-Perez V, Blanco-Pita H, et al. Association between offensive and defensive playing style variables and ranking position in a national football league. Journal of Sports Sciences. 2022;40(1):50–8. pmid:34498545
- 21. Fernandez-Navarro J, Fradua L, Zubillaga A, McRobert AP. Evaluating the effectiveness of styles of play in elite soccer. International Journal of Sports Science & Coaching. 2019;14(4):514–27.
- 22. Yang G, Leicht AS, Lago C, Gómez M-Á. Key team physical and technical performance indicators indicative of team quality in the soccer Chinese super league. Research in Sports Medicine. 2018;26(2):158–67. pmid:29382229
- 23. González-Ródenas J, Aranda R, Aranda-Malaves R. The effect of contextual variables on the attacking style of play in professional soccer. Journal of Human Sport and Exercise. 2020;16(2).
- 24. Hughes M, Franks I. Analysis of passing sequences, shots and goals in soccer. Journal of Sports Sciences. 2005;23(5):509–14. pmid:16194998
- 25. Papadopoulos S, Papadimitriou K, Konstantinidou X, Matsouka O, Pafis G, Papadopoulos D. Factors Leading to Goal Scoring in the Spanish and Italian Soccer Leagues. Sport Mont. 2021;19(1):13–8.
- 26. Liu T, Zhou C, Shuai X, Zhang L, Zhou J, Yang L. Influence of different playing styles among the top three teams on action zones in the World Cup in 2018 using a Markov state transition matrix. Frontiers in Psychology. 2022;13. pmid:36467172
- 27. Forcher L, Forcher L, Wäsche H, Jekauc D, Woll A, Gross T, et al. Is ball-possession style more physically demanding than counter-attacking? The influence of playing style on match performance in professional soccer. Frontiers in Psychology. 2023;14. pmid:37484109
- 28. Gonzalez-Rodenas J, Lopez-Bondia I, Calabuig F, Pérez-Turpin JA, Aranda R. The effects of playing tactics on creating scoring opportunities in random matches from US Major League Soccer. International Journal of Performance Analysis in Sport. 2015;15(3):851–72.
- 29. Lago-Ballesteros J, Lago-Peñas C, Rey E. The effect of playing tactics and situational variables on achieving score-box possessions in a professional soccer team. Journal of Sports Sciences. 2012;30(14):1455–61. pmid:22856388
- 30. Sarmento H, Clemente FM, Harper LD, Costa ITd, Owen A, Figueiredo AJ. Small sided games in soccer—a systematic review. International Journal of Performance Analysis in Sport. 2018;18(5):693–749.
- 31. Mitrotasios M, Gonzalez-Rodenas J, Armatas V, Aranda R. The creation of goal scoring opportunities in professional soccer. Tactical differences between Spanish La Liga, English Premier League, German Bundesliga and Italian Serie A. International Journal of Performance Analysis in Sport. 2019;19(3):452–65.
- 32. Aranda R, González-Ródenas J, López-Bondia I, Aranda-Malavés R, Tudela-Desantes A, Anguera MT. “REOFUT” as an Observation Tool for Tactical Analysis on Offensive Performance in Soccer: Mixed Method Perspective. Frontiers in Psychology. 2019;10.
- 33. Bondia IL, González-Rodenas J, Moreno FC, Pérez-Turpin JA, Malavés RA. Creating goal scoring opportunities in elite soccer. Tactical differences between Real Madrid CF and FC Barcelona. RETOS Nuevas Tendencias en Educación Física, Deporte y Recreación. 2017;16(32):233–7.
- 34. Fernandez-Navarro J, Fradua L, Zubillaga A, McRobert AP. Influence of contextual variables on styles of play in soccer. International Journal of Performance Analysis in Sport. 2018;18(3):423–36.
- 35. Gonzalez-Rodenas J, Lopez-Bondia I, Calabuig F, Pérez-Turpin JA, Aranda R. Creation of goal scoring opportunities by means of different types of offensive actions in US major league soccer. Human Movement Special Issues. 2017;2017(5):106–16.
- 36. Wang SH, Qin Y. Analysis of shooting and goal scoring patterns in the 2019 France Women’s World Cup. Journal of Physical Education and Sport. 2020;20(6):3080–9.
- 37. Michailidis Y, Michailidis C, Primpa E. Analysis of goals scored in European Championship 2012. 2013;8(2):367–75.
- 38. Gonzalez-Rodenas J, Lopez-Bondia I, Calabuig F, Pérez-Turpin JA, Aranda R. Association between playing tactics and creating scoring opportunities in counterattacks from United States Major League Soccer games. International Journal of Performance Analysis in Sport. 2016;16(2):737–52.
- 39. Mitrotasios M, Ródenas JG, Armatas V, Malavés RA. Creating goal scoring opportunities in men and women UEFA Champions League soccer matches: Tactical similarities and differences. Retos: Nuevas tendencias en educación física, deporte y recreación. 2022(43):154–61.
- 40. Oliva-Lozano J M., Martínez-Puertas H, Fortes V, López- Del Campo R, Resta R, Muyor J.M. Is there any relationship between match running, technical-tactical performance, and team success in professional soccer? A longitudinal study in the first and second divisions of LaLiga. Biology of Sport. 2023;40(2):587–94. pmid:37077785
- 41. Casal C, Losada J, Dios R, Arda A. Gender differences in technical-tactical behaviour of La Liga Spanish football teams. Journal of Human Sport and Exercise. 2021;16:37–52.
- 42. de Jong LMS, Gastin PB, Angelova M, Bruce L, Dwyer DB. Technical determinants of success in professional women’s soccer: A wider range of variables reveals new insights. PLOS ONE. 2020;15(10):e0240992. pmid:33091064
- 43. Kubayi A, Larkin P. Technical performance of soccer teams according to match outcome at the 2019 FIFA Women’s World Cup. International Journal of Performance Analysis in Sport. 2020;20(5):908–16.
- 44. Harkness-Armstrong A, Till K, Datson N, Myhill N, Emmonds S. A systematic review of match-play characteristics in women’s soccer. PLOS ONE. 2022;17(6):e0268334. pmid:35771861
- 45. Pulling C. Long corner kicks in the English Premier League: Deliveries into the goal area and critical area. Kinesiology. 2015;47(2.):193–201.
- 46. Pollard R, Reep C. Measuring the effectiveness of playing strategies at soccer. Journal of the Royal Statistical Society Series D: The Statistician. 1997;46(4):541–50.
- 47.
Luxbacher JA. Soccer: Steps to success. 4 ed: Human Kinetics Publishers; 2013.
- 48.
Wesson J. The science of soccer: Crc Press; 2019.
- 49.
Altman DG. Practical statistics for medical research: Chapman and Hall/CRC; 1990.
- 50. Viera AJ, Garrett JM. Understanding interobserver agreement: the kappa statistic. Fam med. 2005;37(5):360–3. pmid:15883903
- 51. Pouwer A-FW, Bult P, Otte I, van der Brand M, van der Horst J, Harterink LJV, et al. Measuring the depth of invasion in vulvar squamous cell carcinoma: interobserver agreement and pitfalls. Histopathology. 2019;75(3):413–20. pmid:31021414
- 52.
Cohen J. Statistical power analysis for the behavioral sciences: Routledge; 2013.
- 53. Almeida CH, Volossovitch A. Multifactorial analysis of football penalty kicks in the Portuguese First League: A replication study. International Journal of Sports Science & Coaching. 2022;18(1):160–75.
- 54. Cohen J. Statistical power analysis. Current directions in psychological science. 1992;1(3):98–101.
- 55.
Bate R. Football chance: tactics and strategy. Science and Football (Routledge Revivals): Routledge; 2013. p. 293–301.
- 56. Garratt K, Murphy A, Bower R. Passing and goal scoring characteristics in Australian A-League football. International Journal of Performance Analysis in Sport. 2017;17(1–2):77–85.
- 57. Tenga A, Holme I, Ronglan LT, Bahr R. Effect of playing tactics on goal scoring in Norwegian professional soccer. Journal of Sports Sciences. 2010;28(3):237–44. pmid:20391095
- 58. Hewitt JH, Karakuş O. A machine learning approach for player and position adjusted expected goals in football (soccer). Franklin Open. 2023;4:100034.
- 59. Lago-Peñas C, Gómez-Ruano M, Yang G. Styles of play in professional soccer: an approach of the Chinese Soccer Super League. International Journal of Performance Analysis in Sport. 2017;17(6):1073–84.
- 60. Redwood-Brown AJ, O’Donoghue PG, Nevill AM, Saward C, Sunderland C. Effects of playing position, pitch location, opposition ability and team ability on the technical performance of elite soccer players in different score line states. PLOS ONE. 2019;14(2):e0211707. pmid:30721265
- 61. Gouveia V, Duarte JP, Nóbrega A, Sarmento H, Pimenta E, Domingos F, et al. Notational Analysis on Goal Scoring and Comparison in Two of the Most Important Soccer Leagues: Spanish La Liga and English Premier League. Applied Sciences [Internet]. 2023; 13(12).
- 62. González-Rodenas J, Aranda-Malaves R, Tudela-Desantes A, Nieto F, Usó F, Aranda R. Playing tactics, contextual variables and offensive effectiveness in English Premier League soccer matches. A multilevel analysis. PLOS ONE. 2020;15(2):e0226978. pmid:32069336
- 63. Stafylidis A, Michailidis Y, Mandroukas A, Gissis I, Metaxas T. Analysis of goal scoring and performance indicators in the 2020–2021 Greek soccer league. Journal of Physical Education and Sport. 2022;22(1):91–9.
- 64. Beare H, Stone JA. Analysis of attacking corner kick strategies in the FA women’s super league 2017/2018. International Journal of Performance Analysis in Sport. 2019;19(6):893–903.
- 65. Lee J, Mills S. Analysis of corner kicks at the FIFA Women’s World Cup 2019 in relation to match status and team quality. International Journal of Performance Analysis in Sport. 2021;21(5):679–99.
- 66. Kubayi A, Larkin P. Analysis of teams’ corner kicks defensive strategies at the FIFA World Cup 2018. International Journal of Performance Analysis in Sport. 2019;19(5):809–19.
- 67. Hewitt A, Greenham G, Norton K. Game style in soccer: what is it and can we quantify it? International Journal of Performance Analysis in Sport. 2016;16(1):355–72.
- 68. González Rodenas J, Lopez Bondia I, Aranda Malavés R, Tudela-Desantes A, Sanz-Ramírez E, Aranda R. Technical, tactical and spatial indicators related to goal scoring in European elite soccer. Journal of Human Sport and Exercise. 2019;15.
- 69. Adeyemo VE, Palczewska A, Jones B, Weaving D, Whitehead S. Optimising classification in sport: a replication study using physical and technical-tactical performance indicators to classify competitive levels in rugby league match-play. Science and Medicine in Football. 2024;8(1):68–75. pmid:36373953
- 70. Dalton-Barron N, Palczewska A, Weaving D, Rennie G, Beggs C, Roe G, et al. Clustering of match running and performance indicators to assess between- and within-playing position similarity in professional rugby league. Journal of Sports Sciences. 2022;40(15):1712–21. pmid:35938184