An Introduction to Open-Source Football Analytics

Federația Română de Fotbal

Joris Bekkers

Contents

• Introduction


• Football Data & Analytics
• Basics
• Visualizations
• Analytics
• etc.


• Research & Innovation (Tracking Data)
• Formation and Position Label Assignment
• Pressing Intensity
• Graph Neural Networks

Introduction

Joris Bekkers 35 y/o Breda, Netherlands Football Analytics Consultant Research Development Deployment PySport Boardmember Unite Sports Analytics Community Grow and Maintain Open-Source

Introduction to Open-Source Football Analytics

Open-Source Football Analytics Tools!

What is kloppy?

Load, Filter, Transform, Export Standardized Football Data.

Used by

• Universities and Researchers
• Football Data Companies
• Football Clubs / Federations (e.g. Premier League, Bundesliga, Italian Serie A, Eredivisie etc.)

Why kloppy?

Standardization!

WyScout

{"id": 2570396900, "matchId": 5683859, "matchPeriod": "1H", ... "type": {"primary": "pass", "secondary": ["forward_pass", 
"short_or_medium_pass"]}, "location": {"x": 49, "y": 49}, "team": {"id": 9109, "name": "Austria", "formation": "4-2-3-1"}...}

Opta / StatsPerform

<Event event_id="17" type_id="1" period_id="1" min="0" sec="35" player_id="89076" team_id="497" outcome="1" x="41.4" y="16.4">
      <Q id="3925254967" qualifier_id="56" value="Right" /><Q id="3925254973" qualifier_id="212" value="23.1" />
    <Q id="3925254969" qualifier_id="140" value="61.2" /><Q id="3925366067" qualifier_id="178" />
    <Q id="3925254971" qualifier_id="141" value="1.7" /><Q id="3925254975" qualifier_id="213" value="5.8" />
</Event>

StatsBomb

{"period" : 1, "minute" : 0, "second" : 3, "type" : {"id" : 30, "name" : "Pass"},
... "possession" : 2, "location" : [ 36.8, 27.3 ], "pass" : { "recipient" : { "id" : 6613}, "length" : 68.33, "angle" : 0.756}

Kloppy

<PassEvent event_id='2570396900' time='P1T00:01' player='Romano Schmid' result='COMPLETE'>
<CarryEvent event_id='2570396909' time='P1T00:02' player='Christoph Baumgartner' result='COMPLETE'>
<PassEvent event_id='2570396910' time='P1T00:04' player='Christoph Baumgartner' result='COMPLETE'>
<DuelEvent event_id='2570396803' time='P1T00:05' player='Denis Drăguş' result='WON'>
<DuelEvent event_id='2570396912' time='P1T00:05' player='Romano Schmid' result='LOST'>
<CarryEvent event_id='2570396804' time='P1T00:06' player='Nicolae Stanciu' result='COMPLETE'>
<PassEvent event_id='2570396805' time='P1T00:10' player='Nicolae Stanciu' result='INCOMPLETE'>
<InterceptionEvent event_id='2570396918' time='P1T00:10' player='Romano Schmid' result='LOST'>
<DuelEvent event_id='2570396919' time='P1T00:12' player='Romano Schmid' result='LOST'>

Supported Providers

Provider	Event	Tracking	Public Data
Hawkeye (2D)
Metrica
PFF
SecondSpectrum	PR #437
Signality
SkillCorner
Sportec
StatsPerform / Opta
Tracab
DataFactory
StatsBomb
WyScout

Implemented Not yet implemented

Kloppy - Event Data

What is Event Data?

• Shots
• Passes
• Defensive Actions (Interceptions, Tackles, Clearances, Fouls etc.)
• Dribbles
• Substitutions
• Cards
  etc.

Event data Documentation

Load Event Data

from kloppy import wyscout

wyscout.load(
    event_data="data/austria-romania.json"
)
home_team, away_team = event_dataset.metadata.teams
print(f"{home_team.name} v {away_team.name}")

>>> Austria v Romania

Event data Documentation

Meta data

Within the metadata we can find a lot of information, for example the player_id of each player.

for p in away_team.players:
    print(f"{p.full_name} - {p.player_id}")

>>> Andrei Rațiu - 504303
>>> Horaţiu Moldovan - 513850
>>> Mihai Popescu - 363609
>>> Răzvan Gabriel Marin - 236820
>>> Daniel Bîrligea - 511261

Player Documentation

Meta data

Pitch Dimensions

event_dataset.metadata.pitch_dimensions

>>> NormalizedPitchDimensions(x_dim=Dimension(min=0, max=1), y_dim=Dimension(min=0, max=1),
standardized=True, unit=<Unit.NORMED: 'normed'>, ..., pitch_length=105, pitch_width=68)

Attacking Left to Right

event_dataset.metadata.orientation

>>> action-executing-team

Metadata Documentation

Filter

player = away_team.get_player_by_id("236820")

player_events = event_dataset.filter(
    lambda event: (
        event.player == player if event.player is not None else False
    )
)

>>> <EventDataset record_count=32>

Filter Documentation

Transform

transformed_event_dataset = (
    event_dataset
    .transform(to_coordinate_system="secondspectrum")
)

event_dataset.metadata.pitch_dimensions

>>> MetricPitchDimensions(x_dim=Dimension(min=-52.5, max=52.5),
 y_dim=Dimension(min=-34.0, max=34.0), 
 standardized=False, unit=<Unit.METERS: 'm'>, 
 ... pitch_length=105.0, pitch_width=68.0)

Transformation Documentation

Transform - Coordinate Systems

Coordinate System Documentation

Export

df = (
  transformed_event_dataset
  .to_df(
    "event_type", "team_id", "player_id", "coordinates_*",
    match_id="match_5678",
    is_pass=lambda e: e.event_type.name in ["PASS", "SHOT"],
    engine="polars"  # < or "pandas"
  )
)

Exporting to Pandas / Polars Documentation

mplsoccer

Andy Rowlinson at PySport X StatsPerform

mplsoccer Examples

mplsoccer Examples

mplsoccer Examples

mplsoccer Examples

mplsoccer!

from mplsoccer.pitch import Pitch
pitch = Pitch(pitch_color='grass', line_color='white', stripe=True)
fig, ax = pitch.draw()

mplsoccer Documentation

from kloppy import wyscout

event_dataset = wyscout.load(
    event_data="data/extract_Austria_Romania_events.json"
)
home_team, away_team = event_dataset.metadata.teams

player = away_team.get_player_by_id("236820")

player_events = (
    event_dataset
    .filter(lambda event: (
        event.player == player if event.player is not None else False
    ))
    .transform(to_coordinate_system="secondspectrum")
)

from mplsoccer import Pitch

def heatmap(xs, ys):
    pitch = Pitch(
        line_zorder=2,
        pitch_type="secondspectrum",
        pitch_length=player_events.metadata.pitch_dimensions.pitch_length,
        pitch_width=player_events.metadata.pitch_dimensions.pitch_width,
    )
    fig, ax = pitch.draw()
    
    ax.set_title(f"{player.last_name} - {player.team.name}")
    pitch.kdeplot(xs, ys, ax=ax, cmap="YlOrRd", fill=True, levels=100)

xs = [event.coordinates.x for event in player_events if event.coordinates is not None]
ys = [event.coordinates.y for event in player_events if event.coordinates is not None]

heatmap(xs, ys)

Shotmap

from kloppy.domain import EventType

team_shots = (
    event_dataset
    .filter(
        lambda event: (
            event.event_type == EventType.SHOT
        )
    ).transform(
        to_orientation="STATIC_AWAY_HOME",
        to_coordinate_system="secondspectrum"
    )
)
home_team, away_team = event_dataset.metadata.teams

pitch = Pitch(
    line_zorder=2,
    pitch_type="secondspectrum",
    half=False,
    pitch_length=events.metadata.pitch_dimensions.pitch_length,
    pitch_width=events.metadata.pitch_dimensions.pitch_width,
)
fig, ax = pitch.draw()

ax.set_title(f"{home_team.name} & {away_team.name} Shots")

for event in events:
    if event.coordinates is not None and event.result_coordinates is not None:
        color = "#DB2129" if event.team.name == "Austria" else "#CE9D00"
        x = event.coordinates.x
        y = event.coordinates.y
        x_end = event.result_coordinates.x
        y_end = event.result_coordinates.y
        
        pitch.scatter(x, y, ax=ax, s=100, color=color, edgecolors="black")
        
        pitch.arrows(x, y, x_end, y_end, ax=ax, width=2, headwidth=5, 
                        headlength=5, color=color, alpha=0.6)

Advanced Event Data Analytics

Socceraction

from socceraction.spadl.kloppy import convert_to_actions
import socceraction.atomic.spadl as atomicspadl

event_dataset = wyscout.load(
    event_data="data/extract_Austria_Romania_events.json"
)

spadl_actions = convert_to_actions(event_dataset)
atomic = atomicspadl.convert_to_atomic(spadl_actions)

Socceraction / Atomic VAEP (KU Leuven)
Example Notebooks (KU Leuven)

Advanced Football Analytics

Positional Tracking Data

The unravelsports package aims to aid researchers, analysts and enthusiasts by providing intermediary steps in the complex process of converting raw sports data into meaningful information and actionable insights.

• Polars DataFrame Conversion
• Formation and Position Identification (EFPI) [Bekkers (2025)]
• Pressing Intensity [Bekkers (2024)]
• Graph Neural Network Training, Conversion and Prediction
  [Bekkers & Sahasrabudhe (2023)]

1. Load Kloppy & Convert

from kloppy import sportec

from unravel.soccer import KloppyPolarsDataset

kloppy_dataset = sportec.load_open_tracking_data()

kloppy_polars_dataset = KloppyPolarsDataset(
    kloppy_dataset=kloppy_dataset
)

KloppyPolarsDataset

Advanced Football Analytics

Elastic Formation and Position Identification (EFPI)

from unravel.soccer import EFPI

model = EFPI(dataset=kloppy_polars_dataset)
model.fit(
    # Default 65 formations , or specify a subset (e.g. ["442" , "433"])
    formations=None,
    # specific time intervals (e.g. 1m, 1m14s, 2m30s etc.), or specify "possession", "period" or "frame".
    every="5m",
    substitutions="drop",
    change_threshold=0.1,
    change_after_possession=True,
)

EFPI Example

Advanced Football Analytics

Pressing Intensity

Pressing Intensity

Requirements

Use positional tracking data Multipurpose / Extendable No ML / prediction model Not too complex Not too simple Measurable at every moment Summable Measure pressing in the whole "system”

Total Pressure = 1 - [(1 – 86%) * (1 – 19%)] = 89%

What is the chance 🏃‍♂️^defense will reach 🏃‍♀️^attack or ⚽️ within
some amount of time given their current direction and speed?"

UnravelSports Blog

2. Pressing Intensity!

from unravel.Football import PressingIntensity

import polars as pl

model = PressingIntensity(
    dataset=kloppy_polars_dataset
)
model.fit(
    start_time=pl.duration(minutes=1, seconds=53),
    end_time=pl.duration(minutes=2, seconds=32),
    period_id=1,
    method="teams",
    ball_method="max",
    orient="home_away",
    speed_threshold=2.0,
) 

>>> PressingIntensity(n_frames=143)

Probability to Intercept

[[0.07 0.   0.   0.   0.   0.03 0.01 0.   0.   0.   0.23]
 [0.28 0.   0.   0.   0.   0.02 0.   0.   0.   0.   0.  ]
 [0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.  ]
 [0.09 0.   0.   0.   0.   0.1  0.06 0.   0.   0.   0.05]
 [0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.  ]
 [0.   0.   0.   0.   0.   0.09 0.6  0.   0.   0.   0.  ]
 [0.   0.   0.   0.   0.   0.83 0.08 0.   0.   0.   0.  ]
 [0.02 0.   0.   0.   0.   0.02 0.   0.   0.   0.01 0.  ]
 [0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.  ]
 [0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.71]
 [0.3  0.   0.   0.   0.   0.06 0.   0.   0.   0.   0.03]]

3. Video

Pressing Intensity Video Example

Advanced Football Analytics

Graph Neural Networks

Graph Neural Network

Attack Outcome Predictions

from unravel.soccer import SoccerGraphConverter

converter = SoccerGraphConverter(dataset=kloppy_polars_dataset)

dataset = GraphDataset(graphs=converter.to_spektral_graphs())

PyData London 2025

Thanks! Questions? :)