Hugging Face's logo

Datasets:
shaked08
/
Football_Results_and_Odds_Data_of_EPL

Modalities:
Tabular
Text
Formats:
csv
Size:
1K - 10K
Libraries:
Datasets
pandas
Polars
Dataset card Data Studio Files
xet
 Community
1
Dataset Viewer
Auto-converted to Parquet Duplicate
HomeTeam
stringclasses
31 values
AwayTeam
stringclasses
31 values
Referee
stringclasses
35 values
FTHG
int64
0
9
FTAG
int64
0
9
FTR
stringclasses
3 values
HTHG
int64
0
5
HTAG
int64
0
5
HTR
stringclasses
3 values
HS
int64
1
33
AS
int64
1
27
HST
int64
0
19
AST
int64
0
20
HC
int64
0
17
AC
int64
0
16
HY
int64
0
7
AY
int64
0
7
HR
int64
0
2
AR
int64
0
2
B365H
float64
1.07
15
⌀
B365D
float64
3
13
B365A
float64
1.14
26
Liverpool
Norwich
M Oliver
4
1
H
4
0
H
15
12
7
5
11
2
0
2
0
0
1.14
10
19
West Ham
Man City
M Dean
0
5
A
0
1
A
5
14
3
9
1
1
2
2
0
0
12
6.5
1.22
Bournemouth
Sheffield United
K Friend
1
1
D
0
0
D
13
8
3
3
3
4
2
1
0
0
1.95
3.6
3.6
Burnley
Southampton
G Scott
3
0
H
0
0
D
10
11
4
3
2
7
0
0
0
0
2.62
3.2
2.75
Crystal Palace
Everton
J Moss
0
0
D
0
0
D
6
10
2
3
6
2
2
1
0
1
3
3.25
2.37
Watford
Brighton
C Pawson
0
3
A
0
1
A
11
5
3
3
5
2
0
1
0
0
1.9
3.4
4
Tottenham
Aston Villa
C Kavanagh
3
1
H
0
1
A
31
7
7
4
14
0
1
0
0
0
1.3
5.25
10
Leicester
Wolves
A Marriner
0
0
D
0
0
D
15
8
1
2
12
3
0
2
0
0
2.2
3.2
3.4
Newcastle
Arsenal
M Atkinson
0
1
A
0
0
D
9
8
2
2
5
3
1
3
0
0
4.5
3.75
1.72
Man United
Chelsea
A Taylor
4
0
H
1
0
H
11
18
5
7
3
5
3
4
0
0
null
3.3
3.5
Arsenal
Burnley
M Dean
2
1
H
1
1
D
16
18
9
5
10
7
2
1
0
0
1.3
5.5
10
Aston Villa
Bournemouth
M Atkinson
1
2
A
0
2
A
22
12
7
4
10
5
0
2
0
0
2.3
3.4
3.1
Brighton
West Ham
A Taylor
1
1
D
0
0
D
16
8
4
3
8
6
0
2
0
0
2.55
3.25
2.87
Everton
Watford
L Mason
1
0
H
1
0
H
12
8
2
2
4
7
2
3
0
0
1.72
3.8
4.75
Norwich
Newcastle
S Attwell
3
1
H
1
0
H
15
10
8
3
7
5
1
3
0
0
2.25
3.3
3.3
Southampton
Liverpool
A Marriner
1
2
A
0
1
A
14
15
3
6
5
9
2
1
0
0
6.5
4.75
1.44
Man City
Tottenham
M Oliver
2
2
D
2
1
H
30
3
10
2
13
2
1
0
0
0
1.36
5.25
8
Sheffield United
Crystal Palace
D Coote
1
0
H
0
0
D
15
6
3
4
8
4
3
1
0
0
2.55
3.1
2.9
Chelsea
Leicester
O Langford
1
1
D
1
0
H
14
12
5
3
4
5
1
0
0
0
1.7
3.75
5
Wolves
Man United
J Moss
1
1
D
0
1
A
6
9
2
2
4
6
2
2
0
0
3.3
3.3
2.25
Aston Villa
Everton
M Oliver
2
0
H
1
0
H
7
12
3
1
0
6
2
3
0
0
3.25
3.5
2.15
Norwich
Chelsea
M Atkinson
2
3
A
2
2
D
6
23
5
8
1
8
1
1
0
0
4.33
3.75
1.8
Brighton
Southampton
K Friend
0
2
A
0
0
D
12
12
3
4
8
5
1
3
1
0
2.4
3.3
3
Man United
Crystal Palace
P Tierney
1
2
A
0
1
A
22
5
3
3
8
1
2
4
0
0
1.33
5.25
9
Sheffield United
Leicester
A Madley
1
2
A
0
1
A
8
10
3
2
7
4
1
0
0
0
3.3
3.3
2.25
Watford
West Ham
C Kavanagh
1
3
A
1
1
D
23
16
3
10
8
7
1
1
0
0
2.05
3.5
3.5
Liverpool
Arsenal
A Taylor
3
1
H
1
0
H
25
9
5
3
6
4
1
1
0
0
1.5
4.6
6
Bournemouth
Man City
A Marriner
1
3
A
1
2
A
10
19
7
5
4
5
1
3
0
0
15
7.5
1.16
Tottenham
Newcastle
M Dean
0
1
A
0
1
A
17
8
2
3
6
6
2
2
0
0
1.2
6.5
14
Wolves
Burnley
C Pawson
1
1
D
0
1
A
17
13
2
4
4
3
0
2
0
0
1.85
3.4
4.5
Southampton
Man United
M Dean
1
1
D
0
1
A
10
21
2
8
2
3
1
2
1
0
3.8
3.3
2.05
Chelsea
Sheffield United
S Attwell
2
2
D
2
0
H
13
8
5
2
3
4
0
1
0
0
1.36
4.75
9
Crystal Palace
Aston Villa
K Friend
1
0
H
0
0
D
22
10
5
2
13
2
2
4
0
1
2.15
3.3
3.5
Leicester
Bournemouth
P Bankes
3
1
H
2
1
H
13
8
5
2
4
5
1
3
0
0
1.7
4
4.75
Man City
Brighton
J Moss
4
0
H
2
0
H
15
6
6
2
8
1
1
1
0
0
1.08
10
26
Newcastle
Watford
G Scott
1
1
D
1
1
D
13
13
5
3
6
5
2
3
0
0
2.5
3.25
2.87
West Ham
Norwich
P Tierney
2
0
H
1
0
H
18
8
8
3
8
2
2
1
0
0
1.85
3.9
3.9
Burnley
Liverpool
C Kavanagh
0
3
A
0
2
A
7
15
2
7
6
4
0
0
0
0
9.5
5.5
1.3
Everton
Wolves
A Taylor
3
2
H
2
1
H
15
8
6
5
7
7
1
4
0
1
2.2
3.25
3.5
Arsenal
Tottenham
M Atkinson
2
2
D
1
2
A
26
13
8
9
11
6
3
5
0
0
2.37
3.6
2.8
Liverpool
Newcastle
A Marriner
3
1
H
2
1
H
21
8
9
1
10
1
0
0
0
0
1.14
8.5
17
Brighton
Burnley
M Oliver
1
1
D
0
0
D
14
7
5
1
3
6
0
2
0
0
2.15
3.3
3.5
Man United
Leicester
M Atkinson
1
0
H
1
0
H
10
9
5
3
3
9
1
2
0
0
1.95
3.5
4
Sheffield United
Southampton
L Mason
0
1
A
0
0
D
17
11
4
7
12
6
1
1
1
0
2.45
3.2
3
Tottenham
Crystal Palace
C Pawson
4
0
H
4
0
H
13
11
5
6
4
3
4
2
0
0
1.36
5.25
8
Wolves
Chelsea
G Scott
2
5
A
0
3
A
11
15
4
6
7
5
1
2
0
0
2.9
3.3
2.5
Norwich
Man City
K Friend
3
2
H
2
1
H
7
25
3
8
3
16
3
1
0
0
15
9
1.14
Bournemouth
Everton
P Tierney
3
1
H
1
1
D
13
14
6
5
7
7
0
4
0
0
2.9
3.5
2.4
Watford
Arsenal
A Taylor
2
2
D
0
2
A
31
10
7
4
7
1
3
3
0
0
3.6
3.6
2
Aston Villa
West Ham
M Dean
0
0
D
0
0
D
10
13
5
1
2
4
2
1
0
1
2.62
3.5
2.6
Southampton
Bournemouth
C Kavanagh
1
3
A
0
2
A
25
6
6
3
4
5
1
3
0
0
2.05
3.6
3.5
Leicester
Tottenham
P Tierney
2
1
H
0
1
A
16
11
7
4
8
2
1
2
0
0
2.87
3.5
2.37
Burnley
Norwich
D Coote
2
0
H
2
0
H
13
11
6
2
7
3
0
1
0
0
2
3.8
3.5
Everton
Sheffield United
S Hooper
0
2
A
0
1
A
16
2
3
1
12
3
1
3
0
0
1.61
3.75
6
Man City
Watford
M Dean
8
0
H
5
0
H
28
5
11
4
5
4
2
2
0
0
1.1
11
21
Newcastle
Brighton
M Atkinson
0
0
D
0
0
D
11
16
4
3
5
3
2
1
0
0
2.5
3.2
3
Crystal Palace
Wolves
S Attwell
1
1
D
0
0
D
13
14
4
4
6
7
2
1
0
1
2.62
3.1
2.87
West Ham
Man United
A Taylor
2
0
H
1
0
H
8
9
6
3
3
7
2
2
0
0
3.1
3.6
2.2
Arsenal
Aston Villa
J Moss
3
2
H
0
1
A
21
14
6
9
9
4
5
1
1
0
1.4
4.75
8
Chelsea
Liverpool
M Oliver
1
2
A
0
2
A
13
6
2
3
6
4
3
3
0
0
3.5
3.75
2
Sheffield United
Liverpool
A Taylor
0
1
A
0
0
D
12
16
2
4
6
5
1
1
0
0
9
5.25
1.33
Aston Villa
Burnley
L Mason
2
2
D
1
0
H
16
10
3
3
7
3
1
4
0
0
2.3
3.4
3.1
Bournemouth
West Ham
S Attwell
2
2
D
1
1
D
13
17
5
6
6
6
3
1
0
0
2.5
3.5
2.75
Chelsea
Brighton
C Kavanagh
2
0
H
0
0
D
24
8
10
1
5
2
2
3
0
0
1.4
4.75
8
Crystal Palace
Norwich
J Moss
2
0
H
1
0
H
15
10
3
3
5
5
3
1
0
0
1.9
3.75
3.75
Tottenham
Southampton
G Scott
2
1
H
2
1
H
9
14
4
6
6
6
0
2
1
0
1.4
4.75
8
Wolves
Watford
P Tierney
2
0
H
1
0
H
6
14
2
5
1
6
0
1
0
0
1.85
3.4
4.5
Everton
Man City
M Oliver
1
3
A
1
1
D
12
20
8
9
5
2
2
2
0
0
10
5.5
1.3
Leicester
Newcastle
C Pawson
5
0
H
1
0
H
13
3
5
0
9
0
1
1
0
1
1.5
4
7.5
Man United
Arsenal
K Friend
1
1
D
1
0
H
16
10
4
5
8
7
4
2
0
0
2.37
3.5
2.9
Brighton
Tottenham
J Moss
3
0
H
2
0
H
17
8
6
3
4
4
2
1
0
0
null
3.6
1.9
Burnley
Everton
G Scott
1
0
H
0
0
D
9
11
3
2
7
9
2
1
0
1
2.87
3.3
2.5
Liverpool
Leicester
C Kavanagh
2
1
H
1
0
H
18
2
8
1
4
6
1
4
0
0
1.44
4.75
7
Norwich
Aston Villa
A Madley
1
5
A
0
2
A
21
22
5
12
10
6
1
3
0
0
2.37
3.6
2.8
Watford
Sheffield United
A Marriner
0
0
D
0
0
D
8
9
2
3
7
7
0
2
0
0
2.05
3.4
3.6
West Ham
Crystal Palace
M Oliver
1
2
A
0
0
D
9
7
4
4
2
2
3
2
0
0
1.95
3.6
3.8
Arsenal
Bournemouth
null
1
0
H
1
0
H
12
10
2
2
14
5
1
2
0
0
1.44
4.75
7
Man City
Wolves
C Pawson
0
2
A
0
0
D
18
7
2
2
9
1
5
2
0
0
1.12
9
21
Southampton
Chelsea
P Tierney
1
4
A
1
3
A
10
13
3
7
3
4
0
1
0
0
4
3.8
1.85
Newcastle
Man United
M Dean
1
0
H
0
0
D
12
12
2
3
4
6
3
3
0
0
4.5
3.5
1.85
Everton
West Ham
P Tierney
2
0
H
1
0
H
19
8
10
4
11
2
2
2
0
0
1.95
3.75
3.75
Aston Villa
Brighton
D Coote
2
1
H
1
1
D
24
20
8
5
7
5
1
2
0
1
2.35
3.4
3
Bournemouth
Norwich
L Mason
0
0
D
0
0
D
11
10
2
1
6
7
1
3
0
0
1.72
4.2
4.2
Chelsea
Newcastle
A Marriner
1
0
H
0
0
D
16
5
8
0
11
0
2
1
0
0
1.3
5.5
10
Leicester
Burnley
J Moss
2
1
H
1
1
D
19
13
3
4
9
4
0
3
0
0
1.57
4
6
Tottenham
Watford
C Kavanagh
1
1
D
0
1
A
12
7
2
2
11
2
4
3
0
0
1.4
5
7.5
Wolves
Southampton
P Bankes
1
1
D
0
0
D
4
14
1
5
1
3
3
2
0
0
1.8
3.6
4.5
Crystal Palace
Man City
A Taylor
0
2
A
0
2
A
7
21
2
10
1
8
1
1
0
0
12
7
1.22
Man United
Liverpool
M Atkinson
1
1
D
1
0
H
7
10
2
4
3
1
0
1
0
0
5.25
3.8
1.65
Sheffield United
Arsenal
M Dean
1
0
H
1
0
H
8
9
2
3
7
12
4
4
0
0
3.9
3.75
1.9
Southampton
Leicester
A Marriner
0
9
A
0
5
A
6
25
3
15
2
7
0
0
1
0
3.1
3.4
2.3
Man City
Aston Villa
G Scott
3
0
H
0
0
D
25
11
9
5
13
7
1
1
1
0
1.08
11
26
Brighton
Everton
A Madley
3
2
H
1
1
D
8
10
4
6
0
5
2
1
0
0
2.9
3.25
2.5
Watford
Bournemouth
M Dean
0
0
D
0
0
D
7
15
3
5
5
10
5
3
0
0
null
3.6
3.25
West Ham
Sheffield United
D Coote
1
1
D
1
0
H
12
10
4
4
10
4
2
2
0
0
2.1
3.5
3.5
Burnley
Chelsea
M Oliver
2
4
A
0
2
A
13
16
5
7
5
6
3
2
0
0
4.5
4
1.72
Newcastle
Wolves
K Friend
1
1
D
1
0
H
13
13
2
5
3
6
2
2
1
0
3
3.1
2.55
Arsenal
Crystal Palace
M Atkinson
2
2
D
2
1
H
15
10
6
4
12
5
2
0
0
0
1.45
4.75
6.5
Liverpool
Tottenham
A Taylor
2
1
H
0
1
A
21
11
13
4
8
3
3
3
0
0
1.5
4.33
6.5
Norwich
Man United
S Attwell
1
3
A
0
2
A
10
21
3
11
1
10
2
2
0
0
4.33
4
1.75
End of preview. Expand in Data Studio

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Strategic Analysis of Home Turf Advantage in Professional Football

Author: Shaked Sabbach | Institution: Reichman University (IDC) | april 2026

click here to watch the video

1. Background & Objectives

In the complex landscape of sports economics, the "Home Advantage" represents a multi-faceted variable that directly influences team valuation, global betting markets, and high-level strategic planning. This research aims to move beyond anecdotal evidence by providing a robust, data-driven quantification of how geographical and environmental factors impact match dynamics. By isolating performance metrics in a controlled dataset, we identify the specific offensive and disciplinary levers that constitute this advantage, providing actionable insights for sports analysts and stakeholders.

2. Research Question

"To what extent does playing at a home stadium provide a statistically significant advantage in performance, discipline, and efficiency, and how does this advantage evolve and compound throughout the 90 minutes of a professional match?"

3. Dataset Overview

The analysis utilizes a high-fidelity dataset of professional football match records, specifically structured to capture granular team-level performance.

  • Source: Professional League Records (Archived for Research).
  • Scope: 1,000 unique match observations.
  • Temporal Focus: Full-match lifecycle tracking (comparing Half-Time vs. Full-Time metrics).

4. Dataset Characteristics & Data Dictionary

The dataset encompasses categorical descriptors and continuous numerical variables, allowing for a comparative longitudinal study:

Variable Description
FTR Full-Time Result (H=Home Win, D=Draw, A=Away Win)
FTHG / FTAG Full-Time Home/Away Goals scored
HTHG / HTAG Half-Time Home/Away Goals scored
HS / AS Home/Away total Shots attempted
HST / AST Home/Away Shots on Target (Primary Efficiency KPI)
HY / AY Home/Away Yellow Cards (Disciplinary stress metric)
HR / AR Home/Away Red Cards

5. Data Preprocessing & Cleaning

To ensure the integrity of the economic modeling and prevent statistical noise, a rigorous preprocessing pipeline was executed:

  • Outlier Mitigation: Matches exhibiting extreme scoring volatility (>8 total goals) were identified as statistical anomalies. These rare events were pruned from the primary sample to ensure that the mean and variance reflect standard competitive environments.
  • Handling Missing Data: * Categorical: Missing Referee data was addressed using Mode Imputation to preserve sample size without introducing selection bias.
  • Numerical: Minor gaps in performance metrics were treated with Median Imputation, insulating the dataset from the influence of market-driven fluctuations.
  • Feature Engineering: Derived variables were generated to calculate "Momentum Shifts," measuring the rate of change in scoring output between match segments.

6. Descriptive Statistics (Post-Cleaning)

The following statistical summary provides a high-level view of the match profiles after data stabilization. This baseline is critical for understanding the typical range of professional match performance.

Metric FTHG (Home Goals) FTAG (Away Goals) HS (Home Shots) AS (Away Shots) HST (Home S.O.T) AST (Away S.O.T)
Count 985.00 985.00 985.00 985.00 985.00 985.00
Mean 1.48 1.14 13.52 10.65 5.92 4.21
Std 1.18 1.02 4.82 3.95 2.14 1.88
Min 0.00 0.00 2.00 1.00 0.00 0.00
25% 1.00 0.00 10.00 8.00 4.00 3.00
50% (Median) 1.00 1.00 13.00 10.00 6.00 4.00
75% 2.00 2.00 17.00 13.00 8.00 5.00
Max 5.00 4.00 28.00 24.00 14.00 11.00

💡 Core Statistical Insights:

  • Efficiency Gap: While the goal difference seems incremental, the Mean scoring for home teams is 29.8% higher than away teams, representing a significant margin in professional sports.
  • Volume vs. Precision: Home teams outperform away teams in both total shots and shots on target, suggesting that home field advantage translates directly into offensive dominance.
  • Controlled Variance: The Standard Deviation remains stable across metrics, confirming that our cleaning process successfully mitigated the "noise" from extreme match results.

7. Exploratory Data Analysis (EDA)

📈 Q1: The Probability of Victory

Question: What is the fundamental probability distribution of match results based on location?

image

Insight: The analysis reveals a stark Home Win dominance of 46%, leaving Draws and Away Wins trailing significantly. This "Home Bias" indicates that visiting teams face a structural disadvantage that goes beyond mere tactical execution. From an economic perspective, this win-rate disparity suggests that market models failing to heavily weight stadium location will consistently yield inaccurate predictions. The statistical gap of nearly 17% over Away Wins confirms that the stadium environment—inclusive of crowd psychology and travel fatigue—is a primary predictor of success.

📈 Q2: The Compound Momentum Effect

Question: Does the performance disparity widen or contract as the match progresses?

image

Insight: By comparing Half-Time (HT) and Full-Time (FT) scoring, we observe a fascinating "Compounding Effect." Home teams increase their scoring rate from 0.64 to 1.48 goals on average. The gap between teams nearly doubles in the second half. This suggests that "Home Advantage" is not static; it is a dynamic force that leverages visiting team fatigue and psychological pressure, reaching its peak in the final 30 minutes of play. As the match reaches its climax, home teams exhibit superior stamina or psychological resilience, effectively translating environmental support into tangible scoreboard results.

📈 Q3: Disciplinary Stress & Tactical Desperation

Question: Do away teams receive more yellow and red cards compared to home teams?

image

Insight: Our analysis using a Point Plot shows a clear trend: Away teams receive significantly more Yellow Cards. Interestingly, the gap in Red Cards is much smaller. This suggests that while referees might be influenced by the home crowd or the away team's defensive pressure on "minor" fouls (Yellow), they remain more objective and consistent when it comes to "major" disciplinary actions (Red).

📈 Q4: The Engine of Victory: Shots on Target

Question: What specific offensive KPI serves as the primary differentiator for winning teams?

image

Average Shots on Target Insight: This visualization establishes a clear performance benchmark: Winners average 7.85 shots on target, a near 45% increase over losing teams. This confirms that victory is not a result of "shot volume" (luck-based shooting), but rather a result of clinical efficiency and high-quality chance creation. In our study, the home team's ability to reach this threshold more frequently is the engine behind their higher win probability. This KPI underscores that offensive quality—the ability to penetrate the defensive third and test the goalkeeper—is the most reliable predictor of match outcome.

📈 Q5: The Distribution of Offensive Quality

Question: How does the "Floor" and "Ceiling" of offensive performance vary between Home and Away teams?

image

Insight: The Box Plot analysis highlights a superior "Offensive Floor" for Home teams. Not only is their median performance higher, but their entire distribution (25th to 75th percentile) is shifted upward. This means that even an "average" performance by a Home team is often equivalent to a "top-tier" performance by an Away team. The consistency of the home team's offensive pressure is what ultimately erodes the visiting team's defense over time. While both teams exhibit outliers, the home team's interquartile range indicates a far more stable and predictable offensive output.

8. Final Conclusion

The research concludes that Home Advantage is a multi-dimensional feedback loop driven by three core pillars:

  1. Structural Efficiency: Home teams maintain a higher "Offensive Floor," generating high-quality shots more consistently and establishing early dominance.
  2. Disciplinary Leverage: Away teams suffer from heightened "Tactical Stress," leading to defensive fragmentation and higher card counts as they struggle to maintain defensive shape.
  3. Compounding Momentum: The advantage is not immediate; it builds throughout the match, peaking in the second half when environmental pressure and fatigue intersect.

[From an entrepreneurial and economic perspective, these insights prove that Stadium Location is a critical risk factor. Predictive models and strategic frameworks must treat "Home Turf" as a weighted variable to account for the systematic shift in performance efficiency identified in this study.

💻 Technical Stack & Environment

The analysis was conducted in a Google Colab environment, leveraging Python's powerful data science ecosystem.

Libraries Used:

  • Data Manipulation: pandas, numpy
  • Visualization: matplotlib.pyplot, seaborn
  • Statistical Formatting: matplotlib.ticker
  • Environment Utilities: google.colab.files (for seamless data export)

💻 Project Implementation

To view the full data cleaning process, statistical analysis, and visualization code, you can access the interactive notebook here:
Downloads last month
378
Number of rows:
1,000
Total file size:
1.52 MB