Illustrating changes in landscapes of passing opportunities along a set of competitive football matches

Abstract

This study aims to illustrate the landscape of passing opportunities of a football team across a set of competitive matches. To do so positional data of 5 competitive matches was used to create polygons of pass availability. Passes were divided into three types depending on the hypothetical threat they may pose to the opposing defense (penetrative, support, and backwards passes). These categories were used to create three heatmaps per match. Moreover, the mean time of passing opportunities was calculated and compared across matches and for the three categories of passes. Due to the specificity of player’s interactive behavior, results showed heatmaps with a variety of patterns. Specifically the fifth match was very dissimilar to the other four. However, characterizing a football match in terms of passing opportunities with a single heatmap dismisses the variety of dynamics that occur throughout a match. Therefore, three temporal heatmaps over windows of 10 min were presented highlighting on-going dynamical changes in pass availability. Results also display that penetrative passes were available over shorter periods of time than backward passes that were available shorter than support passes. The results highlight the sensibility of the model to different task constrains that emerge within football matches.

Highlighting Shooting Opportunities in Football

“…Thus, for each defending player, we defined a potential 'sphere' of action as a function of turning ability and speed. The highest turning ability (360 • ) was established for the lowest recordable speed (0 m/s) [1].…”

“…where y p and x p represent the position of a defending player at each moment of time, ∆t = 0.04 s is the time interval, v tp is the total velocity of the defending player at each moment of time, and one of the 200 angles mentioned above is represented by α n . Finally, x n and y n are the coordinates of the point of each of the n points of the coverage area [1]. For velocities below 1.5 m/s, the value of v t was defined to be equivalent to 1.5 m/s, otherwise a defending player moving at a low or null speed will be considered and measured as covering no space.…”

“…These paradoxical purposes create restricted possibilities of action for each team, for each moment of the match, and for each player. These opportunities of action can be conceptualized as affordances that emerge due to space-time constraints, which can be depicted by a landscape [1][2][3].…”

“…The defenders' and goalkeepers' positional data were captured only in the moment of the shot, which is acceptable for the stated purpose of feeding a machine learning algorithm. However, in this paper, we intended to demonstrate that positional data can be used beyond the quantification of the number of players within the line of shot, and we used players' current and estimated positions to specify shooting opportunities that locally emerge due to players' interactive behaviour during the course of a match [1,2].…”

Science and the World Cup: how big data is transforming football

“…Liverpool FC's data team, which includes physicists formerly at CERN and the University of Cambridge, UK, has built a model that can assess whether a player's actions on the pitch make a goal more likely. And in a partnership with the Spanish giants FC Barcelona, sports scientists at the University of Lisbon, Portugal, last year published an analysis of how long opportunities last for different types of pass in a match 2 .…”