Space-Creating versus Dead Possession: An Off-Ball Possession-Quality Index for Broadcast Football
Introduces the Space-Creation Index (SCI) combined with event-based junk possession index to evaluate spatial utilization quality in football, distinguishing dead possession from space creation.
Key Findings
Methodology
This paper proposes a two-layer framework for assessing possession quality in football: the first layer is an event-side junk-possession index that uses expected threat (xT) grids to evaluate the peak threat gain of each possession segment, combined with game state reconstruction to flag low-threat sequences in tied or losing situations. The second layer is a spatial analysis via the Space-Creation Index (SCI), which projects broadcast video onto pitch coordinates to measure whether possession has created space or merely circulated without threat. This approach integrates event data with spatial control measurements, enabling differentiation between sterile possession and effective space creation, thus providing a more predictive and nuanced understanding of possession quality.
Key Results
- Analysis of 103 matches from the 2026 FIFA World Cup shows that the low-threat possession flag (junk flag) correlates negatively with points (r = -0.37) and expected goal difference (r = -0.51). Even after controlling for VAEP (expected goal value) and field tilt, this flag remains a significant predictor of match outcomes, indicating it captures outcome-relevant information beyond on-ball actions.
- In 35 flagged possession windows, 74% were classified as non-space-creating, 19% as weak progression, and only 6% as space-creating, despite some teams holding over 70% of possession. This demonstrates that high possession does not necessarily imply effective spatial control, emphasizing the importance of spatial analysis.
- The combined two-layer framework effectively distinguishes between 'space-creating but unconverted' and 'sterile' possession, addressing a key limitation of traditional on-ball valuation models. This enhances both tactical insights and predictive accuracy in football analytics.
Significance
This research advances football analytics by moving beyond simple possession percentages to quantify the quality of spatial utilization. It provides a robust, scalable method to evaluate whether teams are genuinely creating dangerous space or merely circulating the ball without threat. Such insights can influence tactical decisions, coaching strategies, and real-time match analysis. The integration of event-based indices with spatial control measures addresses a long-standing challenge in sports analytics: understanding the true effectiveness of possession in creating scoring opportunities. The approach also opens avenues for developing real-time monitoring tools, potentially transforming how teams prepare and adapt during matches.
Technical Contribution
Technically, the paper introduces a novel combination of event-based threat evaluation (using Karun-Singh's xT grid) with spatial control measurement derived from broadcast video via a game-state reconstruction pipeline. The SCI quantifies the net change in pitch control across key zones, capturing whether possession has seized space or merely circulated. The method employs a CPU-only, training-free pipeline for camera calibration, multi-object tracking, and off-screen player imputation, making it practical for large-scale tournament analysis without full tracking data. The integration of these components results in a multi-layered, interpretable metric system that significantly improves the understanding of possession quality.
Novelty
This work is the first to systematically combine event-based threat indices with spatial control analysis to evaluate possession quality in football. Unlike traditional metrics that focus solely on ball actions or territorial dominance, the proposed SCI captures the dynamic spatial impact of possession, addressing a critical gap in sports analytics. The dual-layer framework offers a comprehensive perspective, enabling differentiation between ineffective circulation and genuine space creation, which has been elusive in prior research. This innovation paves the way for more precise tactical assessments and outcome predictions.
Limitations
- The spatial reconstruction relies on broadcast video, which can be affected by camera angles, occlusions, and lighting conditions, leading to potential inaccuracies in pitch control estimation. This affects the reliability of SCI in certain scenarios, especially night matches or low-contrast situations.
- The segmentation of possession sequences is based solely on on-ball event changes, without explicit consideration of pauses, substitutions, or stoppages, which may lead to coarse or imprecise sequence boundaries. The thresholds for space creation classification are empirically set and could benefit from further optimization.
- The computational pipeline, while efficient, may still face challenges in real-time applications, especially during high-intensity phases of play. Future work should focus on integrating multi-modal data sources and refining algorithms for faster, more accurate spatial analysis.
Future Work
Future research will aim to incorporate more advanced tracking data, such as optical or sensor-based systems, to improve spatial accuracy. Developing real-time versions of the SCI for live match analysis could revolutionize tactical decision-making. Additionally, expanding the framework to different leagues and competition levels will test its robustness and generalizability. Further, integrating machine learning models to optimize thresholds and improve off-screen player imputation will enhance the precision of spatial control metrics, ultimately leading to more comprehensive and actionable insights for coaches and analysts.
AI Executive Summary
In the evolving landscape of football analytics, traditional metrics like possession percentage have long been criticized for their inability to accurately reflect team performance or tactical effectiveness. Coaches, analysts, and fans alike have recognized that holding the ball does not necessarily translate into scoring chances or winning outcomes. This disconnect stems from a fundamental limitation: possession quantity does not account for the quality of spatial control or threat creation.
Addressing this challenge, the present study introduces a novel two-layer framework that combines an event-based junk possession index with a spatial control measure called the Space-Creation Index (SCI). The first layer, the junk-possession index, leverages expected threat (xT) grids to evaluate the maximum threat generated during each possession segment. By reconstructing game states and filtering out low-threat circulation in tied or losing situations, this index flags possessions that are unlikely to contribute positively to the outcome. Empirical analysis of 103 matches from the 2026 FIFA World Cup demonstrates that teams engaging in high proportions of low-threat possession—despite controlling the ball—tend to perform worse, with significant negative correlations to points and expected goal difference.
The second layer, the SCI, employs a sophisticated projection of broadcast video onto the pitch coordinate system. This allows the measurement of pitch control changes over short windows, revealing whether possession has genuinely created space or merely circulated without threat. The analysis of 35 flagged windows across nine matches shows that most low-threat possessions do not result in spatial advantage, highlighting the importance of spatial control over mere possession metrics.
Together, these two layers provide a comprehensive picture of possession quality, bridging the gap between event data and spatial dynamics. The framework effectively distinguishes between 'space-creating but unconverted' and 'sterile' possession, offering deeper tactical insights and improving outcome prediction. Despite some limitations related to video-based spatial reconstruction and sequence segmentation, the approach marks a significant step forward in sports analytics.
This research opens new avenues for real-time tactical monitoring, strategic planning, and automated decision support in football. By focusing on the quality of space utilization rather than just possession quantity, teams can better understand their strengths and weaknesses, ultimately leading to more effective gameplay and coaching. The integration of event and spatial analysis represents a paradigm shift, promising to enhance both the scientific understanding and practical management of football matches in the years to come.
Deep Dive
Abstract
Ball possession is the most-cited and most-misleading number in football: 60% recycled in one's own half is not 60% spent pinning the opponent back. Existing event-based possession-value frameworks (expected threat, VAEP, on-ball value) price on-ball actions but ignore the off-ball question a sterile possession poses: did holding the ball create space, or was the circulation dead? We answer this in two layers. First, an event-side junk-possession index prices each possession sequence by its peak threat gain under an expected-threat grid and -- after reconstructing the live scoreline to exclude lead-protecting circulation -- flags low-threat sequences in tied-or-losing states. On the 2026 FIFA World Cup (103 matches, 206 team-matches) the flag correlates negatively with points (r=-0.37) and xG difference (r=-0.51, partly index-coupled). It is not a repackaging of on-ball value: with team offensive VAEP and field tilt held fixed, the junk flag stays strongly negatively associated with points (p<0.0001, also match-clustered) while VAEP is not significant -- in this same-match (descriptive) regression it adds information beyond this on-ball action-value model. Second, for a flagged window we resolve whether it was spatially dead or space-creating by projecting broadcast video to pitch coordinates and measuring a Space-Creation Index (SCI): a net pitch-control change capturing whether the possession seized space or pushed the opponent's block back. Across 31 of 35 flagged windows from nine World Cup matches (a purposive sample), 74% are spatially non-space-creating, 19% weak progression, and 6% space-creating windows the event flag alone would score as failure -- including a side with 73% of the ball that exited on penalties (two non-creating windows). The two layers separate space-creating-but-unconverted from sterile possession, a distinction event-only on-ball value cannot make.
References (12)
Inferring Player Location in Sports Matches: Multi-Agent Spatial Imputation from Limited Observations
Gregory Everett, Ryan Beal, Tim Matthews et al.
A framework for the fine-grained evaluation of the instantaneous expected value of soccer possessions
Javier Fernández, L. Bornn, D. Cervone
Not All Passes Are Created Equal: Objectively Measuring the Risk and Reward of Passes in Soccer from Tracking Data
P. Power, Héctor Ruiz, Xinyu Wei et al.
Actions Speak Louder than Goals: Valuing Player Actions in Soccer
Tom Decroos, L. Bransen, Jan Van Haaren et al.
TranSPORTmer: A Holistic Approach to Trajectory Understanding in Multi-Agent Sports
Guillem Capellera, Luis Ferraz, Antonio Rubio et al.
Enhancing Soccer Camera Calibration Through Keypoint Exploitation
Nikolay S Falaleev, Ruilong Chen
PnLCalib: Sports field registration via points and lines optimization
Marc Guti'errez-P'erez, Antonio Agudo
Physics-Based Modeling of Pass Probabilities in Soccer
W. Spearman, A. Basye, G. Dick et al.
Real Time Quantification of Dangerousity in Football Using Spatiotemporal Tracking Data
D. Link, Steffen Lang, Philipp Seidenschwarz
Multiagent off-screen behavior prediction in football
Shayegan Omidshafiei, Daniel Hennes, M. Garnelo et al.
SoccerNet Game State Reconstruction: End-to-End Athlete Tracking and Identification on a Minimap
Vladimir Somers, Victor Joos, A. Cioppa et al.
BoT-SORT: Robust Associations Multi-Pedestrian Tracking
Nir Aharon, Roy Orfaig, B. Bobrovsky