Pitch-wide space evaluation for soccer transitions

Yohei Ogawa, Rikuhei Umemoto, Keisuke Fujii · 2025

Read Pitch-wide space evaluation for soccer transitions by Yohei Ogawa, Rikuhei Umemoto, 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-2505-14711.txt.

Introduction

Yohei Ogawa, Rikuhei Umemoto, and Keisuke Fujii build a way to grade open space anywhere on a soccer pitch. Space matters most at transitions, when the ball changes hands and a quick counterattack can start far from goal. Older scores rate space by the chance of scoring, which makes them weak far from goal. The head title frames the work as space evaluation at the starting point of soccer transitions.

Their score, called OBPV, rates every spot on the field and leans on real pass patterns to reflect what each position usually does next. Tests on tracking and event data from Spain's La Liga 2023/24 season showed the score lighting up well-used space in counterattacks. It also revealed habits that differ by team after they win or lose the ball, work published in 2025.

Catalog record

CREATORYohei Ogawa, Rikuhei Umemoto, Keisuke Fujii
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE45 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-14711

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Pitch-wide space evaluation for soccer transitions Yohei Ogawa, Rikuhei Umemoto, Keisuke Fujii Abstract Soccer is a sport played on a pitch where effective use of space is crucial. Decision-making during transitions, when possession switches between teams, has been increasingly important, but research on space evaluation in these moments has been limited. Recent space evaluation methods such as OBSO (Off-Ball Scoring Opportunity) use scoring probability, so it is not well-suited for assessing areas far from the goal, where transitions typically occur. In this paper, we propose OBPV (Off-Ball Positioning Value) to evaluate space across the pitch, including the starting points of transitions. OBPV extends OBSO by introducing the field value model, which evaluates the entire pitch, and by employing the transition kernel model, which reflects positional specificity through kernel density estimation of pass distributions. Experiments using La Liga 2023/24 season tracking and event data show that OBPV highlights effective space utilization during counter-attacks and reveals team-specific characteristics in how the teams utilize space after positive and negative transitions. 1 Introduction Advances in soccer science and technology have sharpened player performance and raised the overall pace of play. Transition phases, contested at especially high speed to regain or keep possession, are now viewed as a prime expression of team style. Studies from the German Bundesliga and English Premier League indicate that transition quality strongly influences both attacking and defensive outcomes; better decision-making in these moments therefore increases the likelihood of winning [ 34 , 66 ] . This study aims to identify the requirements for successful counter-attacks that arise during transitions and to characterise a transition-oriented team style. In coaching, knowing which factors enable effective counters is essential, while scouting benefits from understanding an opponent’s transition profile. Earlier work has shown that transitions shape match outcomes [ 48 , 4 ] and pinpointed key components of counter-attacking play [ 23 ] . Other studies have examined shifts from attack to defence [ 2 , 5 ] . However, these investigations seldom considered the spatial configurations created by the full pitch, and transition analyses that emphasise space remain scarce (for details, see Appendix 0.A ). Here, we present a pitch-wide space evaluation method for soccer transitions. Early research visualised pitch control with Voronoi maps [ 59 ] , followed by motion-based models for player movement [ 21 , 7 ] . OBSO (Off-Ball Scoring Opportunity) later provided a mathematical framework that assigns each location a likelihood of producing a goal on first touch [ 57 , 56 ] ; the model has since been transferred to other sports [ 28 , 26 ] . We adopt OBSO for its interpretability, but its goal-oriented design undervalues zones distant from the goal where direct scoring is unlikely. In this paper, we propose the OBPV (Off-Ball Positioning Value) model, which enables OBSO to evaluate the entire pitch. Using this model, we can evaluate spaces such as starting points of counter-attacks based on a mathematical model that is easy to interpret. In addition, it is possible to evaluate the space during transitions, which has not been focused on in previous research [ 57 , 56 , 28 ] . The contributions of this study are as follows. (1) The proposal of a mathematical model that evaluates space with the starting point of transitions and counter-attacks. (2) Among the modules in previous OBSO model [ 56 ] , the field value model was used instead of the score model to enable pitch-wide evaluation in the attack process, and the transition kernel model was used instead of the transition model to reflect the positional specificity of the pitch by estimating the kernel density from the pass distribution in each area. (3) The experiment revealed the importance of effective use of space in counter-attacks and the transition characteristics of each team in the 2023/24 La Liga season. 2 Preliminary for Space Evaluation We first outline the evaluation baseline called OBSO [ 56 ] . OBSO evaluates an off-ball player by considering the joint probability \displaystyle P_{OBSO}(G|D) \displaystyle=\Sigma_{r\in R\times R}P(S_{r}\cap C_{r}\cap T_{r}|D) (1) \displaystyle=\Sigma_{r}P(S_{r}|C_{r},T_{r},D)P(C_{r}|T_{r},D)P(T_{r}|D), (2) where G denotes a goal and D represents a single frame of game data such as player positions and velocities. The details in OBSO are given in Appendix 0.B . P(S_{r}) is the probability of scoring from an arbitrary point r\in R\times R on the pitch, assuming the next on-ball event occurs there. P(C_{r}) is the probability that the passing team will control a ball at point r . P(T_{r}) is the probability that the next on-ball event occurs at point r . For simplicity, P(S_{r}|D),P(T_{r}|D),P(C_{r}|D) are assumed to be independent if the parameter \alpha=0 in the original work implementation [ 56 ] . Then, the joint probability can be decomposed into a series of conditional probabilities as follows: P_{OBSO}(G|D)=\Sigma_{r\in R\times R}P(S_{r}|D)P(C_{r}|D)P(T_{r}|D). (3) P(C_{r}|D) is the probability that the attacking team will control the ball at point r assuming the next on-ball event occurs there, which is called the potential pitch control field (PPCF). P(T_{r}|D) is defined as a fixed two-dimensional Gaussian distribution with the current ball coordinates as the mean. P(S_{r}|D) is simply calculated as a value that decreases with the distance from the goal. We used the grid data and computed P(C_{r}|D) and P(T_{r}|D) according to Appendix 0.B . The overview of the OBSO [ 56 ] is illustrated in Fig. 1 top. In OBSO, the scoring probability was calculated as the output P(S_{r}|D) of the score model as a function of the distance from the goal. However, a limitation of this model is that it is designed to predict scoring opportunities, which results in generally low evaluations for positions far from the goal. To address this issue, we propose an improved model in the next section. Figure 1: Overview OBSO and OBPV. The attacking and defensive players are represented in red and blue, respectively, and the ball is represented in black. The original point (0,0) is the center of the field. One of the attacking players holds the ball near the center and attacks from left to right (x-axis). These models consist of three components: the Score model, PPCF, the Transition model for OBSO, and the field value model, PPCF, the Transition kernel model for OBPV. OBSO shows low evaluations in this pitch. OBPV allows for a more pitch-wide comprehensive assessment. 3 Proposed method In this section, we describe the proposed spatial evaluation model and the dataset used to construct it. As described in Section 2 , since OBSO is a score-prediction-based metric, locations farther from the goal generally receive lower evaluations. To address this issue, we propose a pitch-wide space evaluation model called OBPV (Fig. 1 bottom). OBPV is composed of the field value model, which represents the importance of each location during the attacking phase; the PPCF, which is the same as that used in OBSO; and the transition kernel model, which estimates the next on-ball event location based on actual pass distributions using kernel density estimation. Through these modifications, OBPV enables spatial evaluation during attacking phases that is not limited to scoring opportunities. In the OBPV, the model represents the likelihood that the attacking team can complete a pass to each point on the pitch, as well as how important

Read the complete text