{"id":71874,"date":"2026-04-01T10:50:08","date_gmt":"2026-04-01T10:50:08","guid":{"rendered":"https:\/\/revista-apunts.com\/?p=71874"},"modified":"2026-06-23T21:41:58","modified_gmt":"2026-06-23T21:41:58","slug":"goalkeeper-competitive-level-and-the-organization-of-spanish-futsal-attacks-an-exploratory-observational-study","status":"publish","type":"post","link":"https:\/\/revista-apunts.com\/en\/goalkeeper-competitive-level-and-the-organization-of-spanish-futsal-attacks-an-exploratory-observational-study\/","title":{"rendered":"Goalkeeper Competitive Level and the Organization of Spanish Futsal Attacks: An Exploratory Observational Study"},"content":{"rendered":"\n

Abstract<\/strong><\/h2>\n\n\n\n

This observational study explored how competitive level relates to the organization of futsal attacking actions involving the goalkeeper in Spain. We analyzed 773 interventions (professional: 529 from the Liga Nacional de F\u00fatbol Sala<\/em>; amateur: 244, from the 2nd and 3rd divisions) from the 2023\u20132024 season. A Random Forest model classified competitive level with 71.1% accuracy, identifying game moment (e.g., M30), pass type (short vs. long), and action outcome (progression\/possession) as key discriminators. Logistic regression indicated that foot receptions and short, precise passes were positively associated with professional status (e.g., M30 coefficient\u00a0=\u00a00.41; progression\u00a0=\u00a00.23; possession\u00a0=\u00a00.19). Principal Component Analysis showed partial separation of profiles, while K\u2011Means yielded two clusters: Cluster 1 contained 66.8% professional players and was characterized by teammate-origin receptions, foot control, and short passes under low pressure; Cluster 0 included 52.9% amateur players, with earlier\u2011phase actions (M10), hand receptions, and bowling passes. Professional goalkeepers exhibited greater adaptability by acting under pressure and facilitating structured build\u2011up, whereas amateur goalkeepers favored conservative, low\u2011risk choices. These findings underscore the goalkeeper\u2019s evolving offensive role and offer practical insights for talent identification, tactical training, and performance assessment across competition levels; interpretations are exploratory and bounded by the observational design and league\u2011specific context.\u00a0<\/p>\n\n\n

Keywords:<\/strong> competitive standard<\/span>, logistic regression<\/span>, match analysis<\/span>, team sports<\/span><\/p> <\/div> \n\n\n\n

Introduction<\/strong><\/h2>\n\n\n\n

Although futsal goalkeepers are traditionally the last line of defense, research shows that 67% of their interventions have offensive intentions (Oszmaniec & Szwarc, 2015). Historically, their main function was to block shots and prevent goals. However, in modern futsal, their role has expanded significantly to include active participation in the team\u2019s offensive phase. The evolution of the game, alongside regulatory changes by FIFA allowing goalkeepers to act as field players during live play (5vs4+GK), has prompted coaches and analysts to reconsider the tactical potential of goalkeepers, not only as defenders but also as auxiliary outfield players who can contribute to attack construction and ball circulation. This shift has been particularly evident in scenarios where the opponent applies high pressure during the build-up phase. In such cases, using the goalkeeper as an additional field player can help to break the press and create numerical superiority (Corr\u00eaa et al., 2014; Vicente-Vila & Lago-Pe\u00f1as, 2016). Their increasing technical proficiency with their feet has facilitated this change, as observed in elite-level futsal (Amatria et al., 2021). In this vein, M\u00e9ndez et al. (2019b) highlighted that although 5v4+GK strategies are more effective at maintaining possession, they do not necessarily translate into more goal-scoring opportunities, indicating that, while the tactic aids in controlling play, its offensive yield may be limited. Beyond these FIFA regulatory changes, evidence from stakeholders in Spain suggests that the post-2006 harmonization of futsal rules\u2014particularly in sideline and corner restart procedures\u2014diminished perceived spectacle and constrained adaptive behaviors of players, coaches, and referees, as shown in a descriptive cross-sectional study combining questionnaires and field diaries (Cachon Zagalaz, et al., 2014).<\/p>\n\n\n\n

The link between space, numerical balance, and pressure is crucial in understanding the success of offensive strategies in futsal. Similar dynamics have been observed in football, where teams facing low defensive resistance were more successful at creating scoring chances, and ball possession improved when teams were able to manage opponent pressure effectively (Schulze et al., 2019; Forcher et al., 2024). Recent futsal-specific studies provide deeper insight into this phenomenon. Vicente-Vila and Lago-Pe\u00f1as (2016) concluded that the inclusion of the goalkeeper as a fifth field player significantly improves possession effectiveness, especially in short possessions under low defensive pressure. Silva et al. (2021) similarly observed that the primary offensive role of goalkeepers in both professional and amateur games is to support ball retention, with direct contributions to goal scoring remaining sporadic. Furthermore, Szwarc and Oszmaniec (2020; 2021) found that, among top-level teams, most goalkeeper actions during offensive play were aimed at gaining territory and initiating build-up phases. Interestingly, their studies noted that the game\u2019s score (winning, drawing, or losing) had minimal impact on the style and frequency of these actions, suggesting a consistent offensive role regardless of match context. Additionally, the tactical decision to use an outfield goalkeeper alters the physical dynamics of play. According to De Jong et al. (2022), teammates of an outfield goalkeeper covered less distance at high intensity (above 15.4 Km\/h), indicating a more positionally oriented offensive structure during such scenarios. This reinforces the idea that involving the goalkeeper in outfield roles is not merely a reactive tactic, but a deliberate strategy requiring coordination, technical execution, and tactical awareness. Moreover, classic time\u2013motion analysis in elite futsal quantified the spatial\u2013temporal demands on players, demonstrating alternating bouts across five displacement rhythms (walking, jogging, medium speed, high speed, and sprint), with frequent lateral and backward movements and ball-carrying runs\u2014constraints that heighten the need for rapid perception\u2014action coupling, including for goalkeepers (Hern\u00e1ndez, 2001).<\/p>\n\n\n\n

Collectively, the reviewed evidence reinforces the evolving perception that futsal goalkeepers are no longer limited to defensive responsibilities within their own third of the court. Rather, they are emerging as dynamic contributors to the offensive phase, particularly in structured build-up play and in maintaining possession under high pressure. While their direct involvement in goal-scoring opportunities may remain secondary, their participation is increasingly recognized as critical for establishing and sustaining favorable attacking conditions in contemporary futsal. In this vein, we explicitly foreground the contextual variables that most strongly constrain goalkeeper behavior, thus considering the time of play, operationalized in four 10\u2011min segments, the match status (draw, winning, or losing), and strategic situation, in terms of the immediate game context captured through defensive pressure on the ball and numerical configuration, including goalkeeper-as-outfield (5v4+GK). These variables are not ancillary; they are primary constraints shaping the timing, risk profile, and technique of goalkeeper interventions in possession. Empirically, the interaction of match time and scoreline is decisive. Coaches more frequently adopt 5v4+GK under adverse scorelines in late, \u2018critical\u2019 minutes, and goals scored or conceded in this configuration are tightly conditioned by these situational factors, with short and precise attacks being most effective (M\u00e9ndez\u2011Dom\u00ednguez et\u202fal., 2019; 2021; Vicente\u2011Vila & Lago\u2011Pe\u00f1as, 2016). Defensive pressure further moderates goalkeeper effectiveness because ball possession success increases when pressure is low and sequences are brief, conditions under which goalkeepers more profitably act as facilitators in build\u2011up (Vicente\u2011Vila & Lago\u2011Pe\u00f1as, 2016; FIFA Technical Study Group, 2021). Numerical superiority with an outfield goalkeeper also reconfigures physical and positional demands, reducing teammates\u2019 high\u2011intensity running while requiring greater locomotor output from the goalkeeper and supporting a more stable positional structure (De\u202fJong et\u202fal., 2022). Furthermore, match status alone does not always determine an elite goalkeeper\u2019s style across all actions, but situational clusters emerge when scoreline is considered jointly with time and pressure, which justifies considering these contextual factors (Szwarc & Oszmaniec, 2021; M\u00e9ndez\u2011Dom\u00ednguez et\u202fal., 2019). Therefore, to fulfil this expanded role, goalkeepers must possess not only traditional defensive competencies but also technical proficiency with the ball comparable to that of outfield players. However, this dual skill set is relatively rare and likely restricted to players competing at the highest levels of the sport. Accordingly, the aim of the present study was to examine the differential impact of goalkeeper involvement on offensive effectiveness, specifically in terms of goal-scoring opportunities, goals scored, and team ball possession, between professional (Liga Nacional de F\u00fatbol Sala<\/em>) and amateur (2nd and 3rd division) Spanish futsal leagues during the same competitive season. It was hypothesized that goalkeeper participation in attacking phases would have a distinct influence across competition levels, contributing more significantly to ball possession and the creation of scoring opportunities and goals in professional futsal compared to amateur levels.<\/p>\n\n\n\n

Methods<\/strong><\/h2>\n\n\n\n

Observational Design<\/strong><\/h3>\n\n\n\n

This investigation employed a nomothetic, punctual, and multidimensional observational design, consistent with the canonical framework of systematic observational methodology. This design typology ensures scientific rigor when analyzing naturally occurring behaviors in complex sport settings.<\/p>\n\n\n\n

Following Anguera & Hernandez-Mendo (2014), a nomothetic approach was adopted to capture behavioral variability across a broad set of goalkeepers; a punctual structure was selected, as observations were confined to a single competitive season; and a multidimensional configuration was used to incorporate several interacting behavioral dimensions including contextual, spatial, technical, and outcome\u2011related variables. The study adhered to established observational principles regarding ecological validity, perceptual exhaustiveness, and systematic coding structures recommended for mixed\u2011methods observational sport research.<\/p>\n\n\n\n

Participants<\/strong><\/h3>\n\n\n\n

A total of 26 goalkeepers were included in the analysis. Across the 2023\u20132024 season, these goalkeepers accounted for 529 goalkeeper-outfield participations in the Spanish Liga Nacional de F\u00fatbol Sala<\/em> (LNFS) and 244 participations in Spanish amateur futsal leagues (2nd and 3rd divisions). A formal a priori<\/em> power analysis was not feasible because the study did not prospectively recruit participants from a defined population; instead, it exhaustively included all available observations (convenience census) from the target competitions during the observation period. Consistent with current recommendations on transparent sample size reporting, we explicitly justify this choice and delineate the inferential scope of our analyses (i.e., estimation, pattern detection, and hypothesis\u2011generating insights rather than confirmatory hypothesis testing with prospective power guarantees). As outlined by Lakens (2022), acceptable justifications include (a) collecting data from (almost) the entire available population and (b) explicitly acknowledging when a traditional a priori<\/em> power analysis is not applicable due to design constraints. The league permitted the use of images for research purposes. The Blanquerna Research Committee approved the protocol and procedures with reference number 2425006D and granted that the study complies with the European data protection regulation (General Data Protection Regulation) regarding the processing of publicly available team-sport data. <\/p>\n\n\n\n

The matches were systematically analyzed following the systematic observational methodology (Anguera et al., 2011). LINCE PLUS software (Soto et al., 2021) was used for notational analysis, and the data were transferred to Microsoft Excel (Microsoft Excel 2016, Microsoft Corporation, Redmond, WA, USA) and SPSS (IBM SPSS Statistics Version 30.0, IBM Corp., Armonk, NY, USA) for further analysis. All data were recorded using concurrent time\u2011based (Type IV) observational recording, allowing multiple dimensions to co\u2011occur within the same behavioral event. Content validation followed the criterion of authority within systematic observational methodology. Four experts (national futsal coach of the Real Federaci\u00f3n Espa\u00f1ola de F\u00fatbol<\/em>) independently rated the conceptual adequacy and clarity of each criterion and category (response options: YES\/NO). An item was included if \u2265 3 experts responded YES; otherwise, it was excluded or revised. Following Aixa\u2011Requena et\u202fal. (2025), we computed the percentage of positive coincidences by counting YES\u2013YES agreements across all expert pairs for every item (with six pairs per item) and dividing by the total possible pairs. We then derived a two\u2011sided exact binomial 95% confidence interval for the overall agreement rate. This authoritybased validation procedure conforms to the canonical prescriptions of the systematic observational methodology for developing ad hoc instruments (Anguera & Blanco, 2003; Anguera et\u202fal., 2011).<\/p>\n\n\n\n

Two experienced observers (8 years of experience in the notational analysis of futsal events using LINCE), participated in the intra- and inter-rater reliability process using 10% of the sample. The observers\u2019 data were compared using Cohen\u2019s Kappa index (\u03ba) (Robinson & O\u2019Donoghue, 2007), obtaining a very good agreement between both independent observers.<\/p>\n\n\n\n

Observational Instrument. Criteria and Categories<\/strong><\/h3>\n\n\n\n

Table 1 shows the criteria, the categories, the codes and a description of the observational tool.\u00a0<\/p>\n\n\n\n

\n
\n
\n
\"\"<\/figure>\n<\/div>\n\n\n\n
\n

Table 1<\/strong><\/p>\n\n\n\n

Observational Tool for the Analyses<\/em><\/p>\n\n\n\n

See Table<\/a><\/p>\n<\/div>\n<\/div>\n<\/div><\/div>\n\n\n\n

\"\"\/
Court Zones (Vicente-Vila & Lago-Pe\u00f1as, 2016)\u00a0<\/em><\/figcaption><\/figure>\n\n\n\n

Procedures<\/strong><\/h3>\n\n\n\n

Video footage of the professional matches was obtained from the official website of the Real Federaci\u00f3n Espa\u00f1ola de F\u00fatbol<\/em>. Amateur matches were registered in agreement with the local teams and the corresponding federative authorities. Goalkeeper interventions were observed frame-by-frame from a side view of the court. Two experienced observers, each with eight years of specialist experience in futsal notational analysis using LINCE and LINCE Pro, independently coded 32 interventions from professional futsal matches (11% of the total sample, randomly selected), and 25 interventions from amateur futsal matches (10% of the total sample, randomly selected) under identical viewing conditions and blinded to each other\u2019s work and to the study hypotheses, following frame\u2011by\u2011frame procedures. To assess inter\u2011rater reliability, both observers coded the same video subset concurrently, and their coding was compared using Cohen\u2019s kappa (\u03ba), which yielded values ranging from .87 to 1, indicating very good to almost perfect agreement according to established benchmarks in performance analysis research. For intra\u2011rater reliability, the principal observer re\u2011coded the same clips after a 10\u201315\u2011day washout interval, consistent with validated temporal stability protocols described in observational methodology, again producing \u03ba values between .89 and 1, demonstrating strong stability of coding decisions over time. Cohen\u2019s kappa was calculated using IBM SPSS Statistics Version 30.0.0. <\/p>\n\n\n\n

Statistical Analysis<\/strong><\/h3>\n\n\n\n

To test the hypothesis that the behavior of amateur and professional futsal goalkeepers differs across multiple observed categories, a multi-step advanced statistical analysis was conducted. Firstly, a Principal Component Analysis (PCA) was applied to one-hot encoded categorical data to explore underlying patterns and visualize potential differentiation in goalkeeper behaviors based on competitive level (Standard 1 for amateur, Standard 2 for professional). Although PCA is not inherently designed for categorical data, it was used here as a surrogate for Multiple Correspondence Analysis (MCA), which could not be implemented in the current computational environment. Then, a supervised classification approach using a Random Forest classifier was employed to evaluate the predictability of goalkeeper level based on game-action variables. Model performance was assessed using classification accuracy, a confusion matrix, and a variable importance ranking. Furthermore, a multinomial logistic regression was used to identify which variables significantly contributed to distinguishing between amateur and professional goalkeepers. Due to convergence issues in full models, a reduced model using the top 10 most important predictors (from the Random Forest model) was fitted to obtain interpretable coefficients. Finally, unsupervised clustering was performed using the k-means algorithm on the encoded dataset to identify natural groupings of goalkeeper behaviors without using level labels. The resulting clusters were cross-tabulated with goalkeeper type to assess alignment with known classifications. All statistical procedures were executed using Python (version 3.11), with libraries including scikit-learn, statsmodels, pandas, and matplotlib.<\/p>\n\n\n\n

Results<\/strong><\/h2>\n\n\n\n

Figure 2 presents a Principal Component Analysis biplot showing the first two components derived from categorical game data. A partial visual separation was observed between amateur and professional goalkeepers, with some overlap, suggesting underlying behavioral distinctions. The Random Forest classifier achieved an overall classification accuracy of 71.1%, with higher precision for professional goalkeepers (77.7%) compared to amateurs (61.5%). The confusion matrix is shown in Figure 3.\u00a0<\/p>\n\n\n\n

\"\"\/
Principal Component Analysis (PCA) Differentiating Amateur and Professional Futsal Goalkeepers Based on Encoded Contextual, Technical, and Outcome Variables\u00a0<\/em><\/figcaption><\/figure>\n\n\n\n

Note. <\/em>Data were one\u2011hot encoded prior to PCA, producing two principal components that summarize multivariate behavioral patterns across goalkeeper interventions.<\/p>\n\n\n\n

\"\"\/
Confusion Matrix Depicting the Classification Performance of the Random Forest Model Differentiating Amateur and Professional Goalkeepers<\/em><\/figcaption><\/figure>\n\n\n\n

Table 2 lists the top 10 most important features for classification based on Gini importance scores from the Random Forest model. Variables such as \u201cMoment of the match: M30\u201d and \u201cAction outcome: PROG\u201d were prominent discriminators. Due to multicollinearity limitations, the multinomial logistic regression was conducted using only the top predictors. Table 3 displays the estimated coefficients, where positive values indicate a higher likelihood of professional goalkeeper classification.<\/p>\n\n\n\n

\n
\n
\n
\"\"<\/figure>\n<\/div>\n\n\n\n
\n

Table 2<\/strong><\/p>\n\n\n\n

Top 10 Most Important Variables (Random Forest)\u00a0 \u00a0 <\/em><\/p>\n\n\n\n

See Table<\/a><\/p>\n<\/div>\n<\/div>\n<\/div><\/div>\n\n\n\n

\n
\n
\n
\"\"<\/figure>\n<\/div>\n\n\n\n
\n

Table 3<\/strong><\/p>\n\n\n\n

Logistic Regression Coefficients (Professional vs Amateur)<\/em><\/p>\n\n\n\n

See Table<\/a><\/p>\n<\/div>\n<\/div>\n<\/div><\/div>\n\n\n\n

k-means clustering resulted in two groups, with Cluster 1 composed of 66.8% professional goalkeepers and Cluster 0 composed of 52.9% amateur goalkeepers. Figure 4 depicts the cluster composition. Table 4 summarizes the modal characteristics of each cluster, highlighting distinct behavioral profiles for each group.<\/p>\n\n\n\n

\"\"\/
Distribution of Amateur (orange) and Professional (blue) Goalkeepers Within Each K means Cluster: Cluster Composition Reflects Similarity of Encoded Behavioral Profiles<\/em><\/figcaption><\/figure>\n\n\n\n
\n
\n
\n
\"\"<\/figure>\n<\/div>\n\n\n\n
\n

Table 4<\/strong><\/p>\n\n\n\n

Dominant Characteristics (Codes) per Cluster (K-means)<\/em><\/p>\n\n\n\n

See Table<\/a><\/p>\n<\/div>\n<\/div>\n<\/div><\/div>\n\n\n\n

Discussion<\/strong><\/h2>\n\n\n\n

The aim of this study was to investigate the offensive behaviors of futsal goalkeepers at different levels of competition, with the expectation that professional and amateur goalkeepers would demonstrate distinctive tactical profiles. The findings confirm this hypothesis, suggesting clear and systematic differences in behavior based on competitive level. These differences are manifested in the timing, type, and purpose of goalkeeper interventions during offensive phases.<\/p>\n\n\n\n

The Principal Component Analysis (Figure 2) demonstrated a partial but notable separation between amateur and professional goalkeepers. This spatial divergence suggests an underlying structure of behavioral traits that correspond to player level. As observed in previous studies, elite goalkeepers in futsal are expected to participate more actively in offensive sequences, not only by initiating play but also by adapting their actions to evolving game contexts (Vicente-Vila & Lago-Pe\u00f1as, 2016; M\u00e9ndez et al., 2019b). The separation detected through PCA echoes findings from 11-a-side football, where similar analyses revealed differentiated spatial and temporal positioning of goalkeepers across competition levels (Lamas et al., 2018; Bassek et al., 2025). Supervised classification using a Random Forest model (Table 3) provided further support for this separation, achieving 71.1% accuracy in classifying goalkeeper level based solely on offensive action descriptors. Among the most predictive variables were contextual elements such as the \u201cmoment of the match: M30\u201d and tactical outcomes such as progressive action. These variables correspond to behaviors previously described in the literature as indicative of strategic involvement, where elite goalkeepers operate as facilitators in build-up sequences rather than mere distributors (Paz-Franco et al., 2014; Szwarc & Oszmaniec, 2020). Professional goalkeepers were more likely to receive the ball from teammates and engage in short, precise passes under pressure, actions that demand high technical execution and rapid decision-making (Vilar et al., 2014). These findings are consistent with the work of Paz-Franco et al. (2014), who emphasize that tactical decision-making under pressure is a key differentiator between elite and sub-elite performers. Conversely, amateur goalkeepers tended to rely more on safer options such as hand receptions and bowling passes, particularly in early match stages (M10), reflecting a risk-averse and less versatile behavioral pattern. This rigidity was also observed in the work of Szwarc and Oszmaniec (2020), who argued that amateur goalkeepers generally engage in low-risk actions to retain possession rather than to generate offensive advantage. These results align with M\u00e9ndez et al. (2019a), who found that top-ranked futsal teams adopt highly coordinated attacking profiles that often rely on the goalkeeper as an active component of the attacking structure, contributing to numerical superiority and facilitating dynamic positional rotations. This may explain the increased prevalence of progressive and context-aware actions observed in professional goalkeepers. Moreover, Corr\u00eaa et al. (2014) demonstrated that when goalkeepers assume outfield roles, the opposing team\u2019s defensive organization is directly impacted, often leading to spatial disorganization. The strategic use of the goalkeeper as an additional attacker, therefore, is not just a technical or tactical choice but a systemic adaptation that reshapes team dynamics on both ends of the court. Additionally, M\u00e9ndez-Dom\u00ednguez et al. (2021) demonstrated that the strategic use of the fly goalkeeper, particularly in the final moments of elite futsal matches, is influenced by game status and match context, showing that goals scored using this strategy are highly dependent on situational conditions. This supports our finding that professional goalkeepers are not only technically skilled but tactically adaptive, deploying offensive interventions selectively in response to time-sensitive and score-sensitive match demands. Their study underscores that fly goalkeeper use is not random but governed by shared patterns in critical phases, which may explain the structured yet flexible actions observed in our elite participants.<\/p>\n\n\n\n

The logistic regression analysis (Table 3) further substantiated the Random Forest findings, showing that technical actions such as foot reception, short passing, and receiving from teammates were positively associated with professional level. These elements suggest a higher degree of tactical integration, as supported by De Jong et al. (2022), who described the elite goalkeeper as a positional support in modern offensive schemes. The offensive role of the goalkeeper in futsal is not limited to restarting play but involves real-time problem-solving and the manipulation of space, often under pressure, to preserve or improve positional advantage (Travassos et al., 2012).<\/p>\n\n\n\n

The k-means clustering (Figure 3, Table 4) suggests two clear profiles: one dominated by professional goalkeepers (Cluster 1), and the other more associated with amateurs (Cluster 0). Cluster 1\u2019s dominant behaviors included receiving the ball from a teammate, acting in later phases of the game (M20+), and opting for short or progressive passes. This aligns with previous characterizations of professional behavior as \u2018strategically delayed,\u2019 allowing for better interpretation of space and coordination with teammates (Szwarc & Oszmaniec, 2020; Vilar et al., 2014). By contrast, Cluster 0 behaviors, dominated by early-match actions and simpler passes, reflects more reactive and less structurally informed participation, which may stem from limited tactical training or reduced perceptual capacity (Wilkins et al., 2018). It is particularly noteworthy that amateur goalkeepers show a narrower behavioral repertoire. This might indicate that sub-elite futsal goalkeepers prioritize ball retention over dynamic offensive engagement. The supposed reduced tactical adaptability noted here may also stem from a lack of shared offensive patterns in amateur teams, as emphasized by Travassos et al. (2012), where synchronized decision-making across lines might be less developed.<\/p>\n\n\n\n

From a methodological perspective, this study indicates the value of combining observational data with advanced mixed methods to investigate tactical behaviors in futsal (Camerino, Casta\u00f1er & Anguera, 2012). Thus, stable tactical behaviors are best identified via repeated observations, context\u2011sensitive analyses, and complementary analytical approaches rather than causal inference. This perspective supports interpreting our multivariate results (Random Forest, logistic regression, clustering) as convergent evidence of robust, recurring patterns in goalkeeper\u2011in\u2011possession behaviors across contexts, strengthening coherence between design, analysis, and claims while avoiding over\u2011attribution of effects (Pompa et al., 2024). The analytical process follows key principles in performance analysis, notably the use of contextualized notational systems, multidimensional coding schemes, and multi-method triangulation as outlined by O\u2019Donoghue (2010) and Hughes et al. (2019). The consistent use of categorical variables rooted in competition-relevant game situations strengthens the ecological validity of the dataset, a criterion emphasized by Anguera et al. (2011) in observational methodology. The integration of dimensionality reduction (PCA), supervised classification (Random Forest, logistic regression), and unsupervised clustering (k-means) offers a comprehensive framework that aligns with recent methodological trends in sports science. These techniques are particularly suitable for exploring complex interaction patterns without imposing restrictive a priori<\/em> assumptions, a necessity in team sports where behaviors are emergent and nonlinear (Weiwei, 2021). In addition, methodological rigor is enhanced through reliability criteria consistent with the standards proposed by Anguera et al. (2017), such as the definition of exhaustive and mutually exclusive categories and the use of expert consensus during the design of the observational tool. The study also reflects the observational principles proposed by Lapresa et al. (2013), with a clear distinction between structural patterns and contextual dimensions, a feature necessary to properly account for tactical variability. These criteria are essential to ensure internal validity and the interpretative power of the conclusions drawn from coded game behavior. As demonstrated by Wilkins et al. (2018), combining qualitative and quantitative perspectives in sports analysis maximizes explanatory depth, especially when analyzing player-environment interactions such as those involving the goalkeeper. The current study adheres to these guidelines by using statistical models not only to classify but also to explain performance differences rooted in game context, tactical function, and temporal distribution of actions. This mixed strategy reflects best practices in contemporary performance analysis research, where the interaction between technical-tactical actions, game context, and player decision-making is analyzed as a dynamic system rather than a sequence of isolated events (Travassos et al., 2013; McLean et al., 2017).<\/p>\n\n\n\n

Limitations <\/strong><\/h2>\n\n\n\n

This study is observational and relies on a convenience census of goalkeeper offensive actions drawn from Spanish professional and amateur leagues within a single season; as such, causal inferences cannot be made, and the findings should be interpreted as exploratory associations rather than effects. The sampling frame (specific competitions, 2023\u201324 season) and contextual constraints (e.g., league styles, tactical norms, and scheduling) may limit generalizability to other countries, competition formats, or future seasons. In addition, although we implemented rigorous coding procedures with very high inter\u2011 and intra\u2011rater agreement and used multivariate models to detect patterns, model outputs (e.g., variable importance, clustering structure) remain contingent on the selected categories, the one\u2011season window, and the ecological variability of match contexts; unmeasured factors (e.g., team tactics, coaching instructions, player fatigue), could partly account for the observed profiles. Together, these limitations recommend caution in interpretation and underscore the need for multi\u2011season, multi\u2011league replications and confirmatory designs before deriving prescriptive conclusions beyond settings similar to those analyzed.<\/p>\n\n\n\n

Conclusions<\/strong><\/h2>\n\n\n\n

To conclude, the data confirm that professional futsal goalkeepers not only possess superior technical abilities but are also tactically integrated actors who influence the game\u2019s offensive flow. Their decision-making is more contextually tuned, their actions more temporally distributed, and their role better aligned with positional play principles. These attributes are consistent with the increasing complexity and multifunctionality required at elite levels and should inform both scouting and training practices moving forward.<\/p>\n\n\n\n

The findings of this study offer actionable guidance for coaches and practitioners seeking to optimize the offensive contribution of futsal goalkeepers across competitive levels. Professional teams can enhance their attacking structure by further integrating the goalkeeper into controlled build-up play, emphasizing foot-based receptions, short passing under pressure, and coordinated positional rotations that exploit numerical superiority and facilitate progression. In contrast, amateur teams should prioritize foundational technical work, particularly first-touch quality, body orientation, and simple short-passing connections, to reduce reliance on low-risk hand distributions and encourage tactical involvement beyond the early phases of play. Across levels, designing training tasks that incorporate contextual constraints such as match moment, defensive pressure, and ball origin can foster more adaptive, context-sensitive behaviors. By incorporating goalkeeper-inclusive positional play circuits, pressure-resistance drills, and structured support patterns, coaches can cultivate decision-making, technical precision, and tactical synergy, enabling goalkeepers to act not only as defenders but also as meaningful contributors to their team\u2019s offensive organization.<\/p>\n\n\n\n

Funding<\/strong><\/h2>\n\n\n\n

No funding was received.<\/p>\n\n\n\n

Acknowledgements<\/strong><\/h2>\n\n\n\n

The Python code for the data analysis was partially developed with the support of Artificial Intelligence tools (ChatGPT 4.0).<\/p>\n","protected":false},"excerpt":{"rendered":"

Abstract This observational study explored how competitive level relates to the organization of futsal attacking actions involving the goalkeeper in Spain. We analyzed 773 interventions (professional: 529 from the Liga Nacional de F\u00fatbol Sala; amateur: 244, from the 2nd and 3rd divisions) from the 2023\u20132024 season. A Random Forest model classified competitive level with 71.1% […]<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_editorskit_title_hidden":false,"_editorskit_reading_time":0,"_editorskit_is_block_options_detached":false,"_editorskit_block_options_position":"{}","inline_featured_image":false,"advgb_blocks_editor_width":"","advgb_blocks_columns_visual_guide":"","footnotes":""},"categories":[49,49],"tags":[14689,8232,7065,381],"author_meta":{"display_name":"finderwilber","author_link":"https:\/\/revista-apunts.com\/en\/author\/finderwilber\/"},"featured_img":null,"coauthors":[],"tax_additional":{"categories":{"linked":["Sport Training<\/a>","Sport Training<\/a>"],"unlinked":["Sport Training<\/span>","Sport Training<\/span>"]},"tags":{"linked":["competitive standard<\/a>","logistic regression<\/a>","match analysis<\/a>","team sports<\/a>"],"unlinked":["competitive standard<\/span>","logistic regression<\/span>","match analysis<\/span>","team sports<\/span>"]}},"comment_count":"0","relative_dates":{"created":"Posted 3 months ago","modified":"Updated 11 hours ago"},"absolute_dates":{"created":"Posted on 1 April 2026","modified":"Updated on 23 June 2026"},"absolute_dates_time":{"created":"Posted on 1 April 2026 10:50","modified":"Updated on 23 June 2026 21:41"},"featured_img_caption":"","series_order":"","_links":{"self":[{"href":"https:\/\/revista-apunts.com\/en\/wp-json\/wp\/v2\/posts\/71874\/"}],"collection":[{"href":"https:\/\/revista-apunts.com\/en\/wp-json\/wp\/v2\/posts\/"}],"about":[{"href":"https:\/\/revista-apunts.com\/en\/wp-json\/wp\/v2\/types\/post\/"}],"author":[{"embeddable":true,"href":"https:\/\/revista-apunts.com\/en\/wp-json\/wp\/v2\/users\/2\/"}],"replies":[{"embeddable":true,"href":"https:\/\/revista-apunts.com\/en\/wp-json\/wp\/v2\/comments\/?post=71874"}],"version-history":[{"count":2,"href":"https:\/\/revista-apunts.com\/en\/wp-json\/wp\/v2\/posts\/71874\/revisions\/"}],"predecessor-version":[{"id":72677,"href":"https:\/\/revista-apunts.com\/en\/wp-json\/wp\/v2\/posts\/71874\/revisions\/72677\/"}],"wp:attachment":[{"href":"https:\/\/revista-apunts.com\/en\/wp-json\/wp\/v2\/media\/?parent=71874"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/revista-apunts.com\/en\/wp-json\/wp\/v2\/categories\/?post=71874"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/revista-apunts.com\/en\/wp-json\/wp\/v2\/tags\/?post=71874"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}