Advanced Pass Analysis in Football: Real-Time Decision-Making Evaluation Based on Threat Creation

Advanced Pass Analysis in Football: Real-Time Decision-Making Evaluation Based on Threat Creation

Abstract

In modern football, analyzing a player’s decision-making during a game is crucial to understanding their impact on the match outcome. Traditional metrics that assess passing success tend to focus solely on completion rates, overlooking the true value of risky decisions that can generate offensive threat. In collaboration with LaLiga, Mediacoach, and Footovision, we are developing a real-time model that not only analyzes a player’s chosen pass but also compares it with the 5 best alternative options based on the danger they would have created. This article presents the theoretical framework of the model, its practical applications, and its relevance in player performance analysis.

Introduction

Football is a sport where every decision on the pitch has direct implications for the final result. Traditionally, player quality analysis has focused on metrics that prioritize passing accuracy, completion rates, or the number of assists and goals. However, these metrics overlook a crucial dimension: the intention to maximize threat creation, even when the pass is not completed. This article introduces a new approach to pass analysis, centered on evaluating not just the outcome, but also the intention behind the decision.

Model Fundamentals

The model we are developing is based on real-time evaluation of a player’s passing decisions, using the Expected Threat (xT) metric to calculate the danger that each pass could create during a play. xT measures the probability that an action on the field (pass, dribble, shot) will lead to a goal. Our model expands this metric by analyzing the 5 best possible passing options in a given situation, comparing the offensive potential of each.

The key of the model is that it not only assesses whether the pass was successful or not but evaluates whether the player intended to maximize threat by choosing the most aggressive option, even if it carried a higher risk of failure.

Methodology

  1. Field Zoning and xT Calculation: The pitch is divided into zones with different xT values based on the historical probability of actions from those zones leading to a goal.
  2. Pass Option Analysis: For each pass made by a player, the model identifies the 5 best alternative options, calculating the xT for each.
  3. Comparison with the Actual Decision: The model compares the option chosen by the player with the alternatives to evaluate if the decision maximized threat creation or if they opted for a safer but less effective option.
  4. Intention Evaluation: The model assigns value to the player’s intention to create danger, regardless of whether the pass was completed.

Results

Initial analysis using data from real matches shows a clear correlation between xT value and execution difficulty of the passes. More creative players tend to choose options with higher xT, even though these are not always the easiest to complete. This contrasts with more conservative players who aim to maintain high completion rates but generate less threat.

A concrete example is Pedri, in a recent analysis of his decision-making during a match. Pedri chose a pass that represented the highest xT option among five alternatives, even though the pass was not completed. This type of decision, while unsuccessful in outcome, demonstrates his game vision and intention to maximize danger.

Discussion

This approach introduces a new way to evaluate football intelligence. It’s not just about assessing how many completed passes a player makes or how many assists they provide, but about recognizing the value of risky decisions that seek to change the course of the game.

In this context, two types of players emerge:

  • Conservative players: They prefer to maintain high passing accuracy rates and minimize risk, often resulting in fewer goal-scoring opportunities.
  • Creative and risk-taking players: They make more difficult and risky decisions, which may not always result in completed passes, but when they do, they almost guarantee goal-scoring opportunities.

This type of analysis also enables more precise evaluation during training. Exercises can be designed to focus on improving decision-making under pressure, teaching players to identify and execute the passes that maximize threat creation.

Practical Applications

The model has multiple applications in tactical and strategic analysis:

  • Decision-making under pressure: Identify how players respond in situations where they have little time to decide and whether they choose safe or risky options.
  • Risk management: Analyze the balance between the risk of a pass and the potential reward in terms of generated threat.
  • Training optimization: Design drills that help players identify the passing options that generate the most threat and improve their ability to execute risky decisions successfully.

Conclusions

The development of a pass decision-making analysis model based on real-time threat creation offers a new perspective on football performance analysis. This approach not only values pass success or failure but also dives into the intention behind each decision, highlighting those players who consistently seek to maximize threat and alter the flow of the game. As football continues to evolve towards greater integration of data and advanced analysis, this type of metric will allow coaches and analysts to identify and unlock the true value of the most creative players.

References

  • Linke, D., & Maurer, H. (2019). Expected Threat: A new way to assess offensive actions in football. Journal of Football Science, 6(2), 45-58.
  • Williams, A., & Richardson, D. (2020). The Tactical Mind: Decision-Making in Professional Football. International Journal of Sports Analytics, 8(4), 201-219.