Vincent's Project 1 Plan

Interactive Soccer Data Visualization in VR

Overview

• Soccer match data are now widely collected in every detailed aspects of the game: from every touches of the ball of a player, to teams positioning, to models that calculate expected goals...

• Strongest teams in Europe use these technology to communicate with the coach during the game; weaker teams also use these models to scout players that are less known

• Player review these visualization analysis to better improve their skills and decisions

• However, although lots of these spatial and temporal data are of high dimensions, most of the soccer visualization analysis are 2D, often showing just one aspect of the game

• On the other hand, the field of view of the player is very important to understand a move; it's always obvious to understand a move when overviewing the entire pitch

• The project aims to use VR technology to help:

  • Visualize different aspects of the game interactively at the same time; for example, how positioning & running of an attacker affect the passing of other player so to build up a goal?

  • With the advantage of VR's simulated field of view, we can visualize data of a single player as if we are standing at his position on the pitch, and study his decision

Motivation

• Video Assistant Referee (VAR) in soccer now widely used

• Semi-automated Offside technology & tracking device introduced in World Cup 2022

• Tracking vests for training

• In-game analysis for tactical changes

• Scouting player

Dataset

• StatsBomb

  • Single player / team aggregated

  • Key passes

  • Off-the-ball positioning of a player to find space (xSpace)

  • Expected goals (xG)

  • Pressure

• How to combine some of these in VR to take advantage of the 3D information?

  • Study a team’s performance

  • Study a player’s decision

Tools

• Tableau & statsbombpy - for preprocessing/previsualize data

• D3.js - for interactive visualization

• A-Frame - for building VR experience on web

Schedule

In-Class Activities

• In group of 3 or 4, each person in the group visualize two datasets in one style of visualization; for example, person A visualize passes & off-the-ball positioning of the attacker using color dots and lines, person B visualize those using heatmap…

• Everyone share what the datasets tell them about

• Fill out a Google form

Deliverables & Wiki Contributions

• An interactive visualization web tool to analyze soccer data in VR

• Tutorials on how to use d3.js for VR visualization