Prediction-based evaluation of back-four defense with spatial control in soccer
Soujanya Dash, Kenjiro Ide, Rikuhei Umemoto, Kai Amino, Keisuke Fujii · 2025
Read Prediction-based evaluation of back-four defense with spatial control in soccer by Soujanya Dash, Kenjiro Ide, Rikuhei Umemoto, Kai Amino, Keisuke Fujii (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-06191.txt.
Introduction
Soujanya Dash and co-authors measure how well a soccer team's back four defends right after losing the ball. Those transition moments leave teams exposed, yet the shared coordination of four defenders is hard to capture in numbers. The study defines plain indicators such as space control, stretch, pressure, and defensive line height.
The team analyzed more than twenty-four hundred defensive sequences from the 2023-24 LaLiga season for Barcelona and Real Madrid. Statistical tests linked relative line height most strongly to defensive success. A gradient boosting model predicted outcomes well and highlighted space control and line height as the dominant signals, with Barcelona showing tighter and more consistent defensive shape.
Catalog record
| CREATOR | Soujanya Dash, Kenjiro Ide, Rikuhei Umemoto, Kai Amino, Keisuke Fujii |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 44 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-06191
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Prediction-based evaluation of back-four defense with spatial control in soccer Soujanya Dash, Kenjiro Ide, Rikuhei Umemoto, Kai Amino, Keisuke Fujii Abstract Defensive organization is critical in soccer, particularly during negative transitions when teams are most vulnerable. The back-four defensive line plays a decisive role in preventing goal-scoring opportunities, yet its collective coordination remains difficult to quantify. This study introduces interpretable spatio-temporal indicators namely, space control, stretch index, pressure index, and defensive line height (absolute and relative) to evaluate the effectiveness of the back-four during defensive transitions. Using synchronized tracking and event data from the 2023-24 LaLiga season, 2,413 defensive sequences were analyzed following possession losses by FC Barcelona and Real Madrid CF. Two-way ANOVA revealed significant effects of team, outcome, and their interaction for key indicators, with relative line height showing the strongest association with defensive success. Predictive modeling using XGBoost achieved the highest discriminative performance (ROC AUC: 0.724 for Barcelona, 0.698 for Real Madrid), identifying space score and relative line height as dominant predictors. Comparative analysis revealed distinct team-specific defensive behaviors: Barcelona’s success was characterized by higher spatial control and compact line coordination, whereas Real Madrid exhibited more adaptive but less consistent defensive structures. These findings demonstrate the tactical and predictive value of interpretable spatial indicators for quantifying collective defensive performance. keywords machine learning, sports, classification, soccer 1 Introduction Defensive organization is a cornerstone of team success in elite soccer, with the defensive line forming the final outfield barrier against opposing attacks. Among various configurations, the back four is a line of four outfield defenders positioned ahead of the goalkeeper which is the most widely adopted system across elite soccer. This configuration is favored by coaches because it provides a clear, stable structure that facilitates rapid reorganization during negative transitions ( Wright \BOthers. \APACyear2011 ; Bradley \BOthers. \APACyear2011 ) where the presence of two central defenders and two fullbacks ensures effective coverage of central and wide areas, reducing the risk of defensive gaps being exploited in moments of instability ( Casal \BOthers. \APACyear2016 ) . Compared to back three or back five systems, the back four enables greater tactical flexibility, improved defensive width, and better coordination for pressing and recovery ( Tierney \BOthers. \APACyear2013 ; Tenga \BOthers. \APACyear2020 ) . Moreover, fullbacks in back four systems are critical for regaining possession and supporting the midfield, further enhancing their team’s ability to quickly re-establish shape after losing the ball ( Gonçalves \BOthers. \APACyear2023 ) . These attributes contribute significantly to the back four’s enduring popularity, especially in the context of elite competition, where the speed and unpredictability of transitions can determine the outcome of matches. While much of soccer analytics has traditionally emphasized attacking strategies or isolated defensive actions, recent studies underscore the importance of analyzing the defensive line as a coordinated unit. The back four’s synchronization, spacing, and denial of key pitch zones have been identified as central determinants of defensive success, particularly during counterattacks and transitional phases ( Bojinov \BBA Bornn \APACyear2016 ; M. Forcher \BOthers. \APACyear2022 ) . Evaluating the collective mechanisms of the defensive line thus provides both theoretical insights into tactical organization and practical applications for coaching interventions. Despite this importance, the empirical study of defensive lines has been limited. The complexity of integrating high-resolution tracking and event data, combined with the fluid and reactive nature of defensive behavior, has constrained prior analyses. As the previous work ( M. Forcher \BOthers. \APACyear2022 ) highlighted in their systematic review, most quantitative research has either focused on broader notions of team compactness or on individual defensive duels, leaving the back four’s coordinated behaviors underexplored. Similarly, another work ( Casal-Sanjurjo \BOthers. \APACyear2021 ) noted that defensive evaluation often relies on outcome-based metrics or manual annotation, which fail to capture dynamic, context-dependent group interactions. In recent years, efforts have been made to formalize collective defensive performance through interpretable indicators such as space control , pressure , and the stretch index . Space control quantifies the ability of defenders to occupy and deny high-value zones, while pressure captures the intensity and proximity of defensive pressing ( Taki \BOthers. \APACyear1996 ; Spearman \APACyear2018 ; Teranishi \BOthers. \APACyear2022 ; Toda \BOthers. \APACyear2022 ; Umemoto \BOthers. \APACyear2022 ; Ide \BOthers. \APACyear2025 ; Umemoto \BBA Fujii \APACyear2023 ; Yeung \BBA Fujii \APACyear2024 ) . The stretch index, by contrast, reflects compactness through average player-to-centroid distances, thereby measuring the dispersion of defensive units ( Clemente \BOthers. \APACyear2015 ; Bojinov \BBA Bornn \APACyear2016 ; M. Forcher \BOthers. \APACyear2022 ) . However, each metric has limitations when applied in isolation: pressure measures often overlook attacker interactions, while space control and stretch indices are too static to capture the fluid dynamics of transitions ( Ogawa \BOthers. \APACyear2025 ) . This has motivated calls for context-sensitive, risk-weighted models that integrate spatial and temporal factors into defensive evaluation. Parallel strands of research have also emphasized defensive positioning in specific game contexts, particularly crossing situations. The previous study ( Pafis \BOthers. \APACyear2025 ) demonstrated that the height of the defensive line and the origin of the cross significantly affect defensive success across elite teams in LaLiga, the Premier League, and the Bundesliga. Their findings highlight vulnerabilities when crosses originate from half-spaces, reinforcing the broader tactical emphasis on coordinated defensive positioning and subunit synchronization. These results align with those of L. Forcher \BOthers. \APACyear2024 , who showed that local compactness near the ball rather than overall team compactness was the most robust predictor of defensive effectiveness. Despite recent advances in spatial metrics such as pitch control and expected possession value, these approaches primarily emphasize offensive progression. Existing defensive measures, by contrast, remain limited to isolated actions (e.g., tackles, interceptions) and fail to capture the coordinated role of the back four during transitions. There is still no framework that jointly evaluates the coordinated actions of the back-four line during negative transitions using synchronized tracking and event data. This gap limits both tactical understanding and the development of quantitative tools for coaches and analysts. Our study addresses this need by introducing interpretable spatio-temporal indicators of defensive line behavior, testing their predictive power, and demonstrating their tactical relevance through comparative analysis. In doing so, we provide a framework that connects defensive coordination to measurable outcomes, a necessary step for advancing both research and applied performance analysis. Building on the tactical importance of defensive line coordination, this study aims to quantitatively evaluate how interpretable indicators of back-four organization relate to defensive outcomes during negative transitions. Specifically, we compare two elite teams namely FC Barcelona and Real Madrid CF, representing contrasting defensive philosophies. We hypothesize that (1) Defensive success can be explained by measurable differences in spatial compactness, pressure intensity, and line positioning; (2) These relationships differ between teams due to distinct tactical behaviors. Materials and methods Dataset In this study, we used two synchronized datasets: ( 1 ) tracking data from SkillCorner at 25 Hz, providing positional and velocity information for all 22 players and the ball, and ( 2 ) event data from StatsBomb, which includes detailed annotations of passes, duels, tackles, and pressures etc. ( Hughes \BOthers. \APACyear2019 ) for Laliga 2023/24 season. Despite being vision-based, SkillCorner has been validated in prior elite-level research for modeling defensive organization ( Andrienko \BOthers. \APACyear2022 ) .