spatio temporal match events in Soccer

0.0(0)
studied byStudied by 0 people
full-widthCall with Kai
GameKnowt Play
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
Card Sorting

1/21

encourage image

There's no tags or description

Looks like no tags are added yet.

Study Analytics
Name
Mastery
Learn
Test
Matching
Spaced

No study sessions yet.

22 Terms

1
New cards

Dataset purpose

This dataset by Wyscout is a large public dataset, with spatiotemporal soccer events that are logged for research

2
New cards

Season coverage

The dataset covers 7 competitions + World Cup 2018 and Euro 2016 

3
New cards

Totals (Matches/events/players)

1941 matches, 3.25M events, 4299 players

4
New cards

Event position coordinates

ises X & Y measured from 0-100% from the attacking team’s perspective.

X represents % distance to the opposition goal

Y represents % distance to right side of field

5
New cards

Main Event Types 

Pass, Foul, Shot, Duel, Free kick, Offside, Touch

6
New cards

Most freqent event

Passes which covered 50% of all events

7
New cards

Spatial Pattern: Shots on Goal

Shows a large cluster around the opponent’s goal 

8
New cards

Spatial Pattern: Defenders vs Forwards

Defenders’ patterns were clustered towards their own half, whereas Forwards’ patterns were clustered on the opposition’s half

9
New cards

Temporal Pattern: Shots

It was found that more shots on goal happened in the second half

10
New cards

Temporal Pattern: Cards

Yellow/Red cards were most common in stoppage time, or closer to end of match

11
New cards

Invasion Index

Represents how close a team player to the opponent’s goal per possession

12
New cards

Invasion Index Calc

The maximum probability of scoring from positions with a possession, averaged across possessions

13
New cards

Acceleration Index

Describes how fast a team reaches its most dangerous position

14
New cards

Acceleration Index Calc

The invasion Index / (time to most dangerous event)²

15
New cards

Passing Networks

Graphical representations of how players interact with one another, Nodes = players, edges = passes

16
New cards

Passing Networks Use

It identifies key players degree, centrality and tactics

17
New cards

Connectivity

Second smallest eigenvalue; with higher values meaning more robust team links and performance

18
New cards

Flow Centrality (Player)

Betweeness centrality in passing network

19
New cards

PlayeRank (Player)

Multidimensioanl/ role-aware ML performance score

20
New cards

Quality Check Pipeline

Auto checks for consistency, ensuring no events were missed + A manual quality control

21
New cards

Method for Data Collection

Step 1: Setting Formations

Step 2: Event Tagging

Step 3: Quality Control

22
New cards