⚽ Machine Learning × Football

ELO-Insights

Predicting Premier League Match Outcomes Using Advanced ML Ensemble Models

0% Model Accuracy
0 Seasons Analyzed
0+ Matches Trained
View Predictions →
Scroll to explore
About The Project

Machine Learning for Football

📊

Data-Driven Analysis

Analyzing 10 seasons of Premier League data from 2014-2024, including ELO ratings, expected goals, and team form metrics.

🤖

Ensemble Learning

Combining Random Forest, XGBoost, and Gradient Boosting models using majority voting for robust predictions.

Match Outcome Prediction

Predicting home wins, away wins, and draws for every fixture in the 2024-2025 Premier League season.

Model Performance

Ensemble Model Accuracy

🌲

Random Forest

72.2%

1000 estimators with optimized depth

XGBoost

75.0%

2000 estimators with tuned hyperparameters

Class-wise Prediction Accuracy

🏠 Home Win 79.8%
🤝 Draw 72.0%
✈️ Away Win 62.1%
Season 2024-2025

Predicted Final Standings

Pos Team Pts Zone
1 🏆 Tottenham 78 Champions League
2 Chelsea 74 Champions League
3 Arsenal 70 Champions League
4 Liverpool 70 Champions League
5 Manchester Utd 66 Europa League
6 Manchester City 66 Europa League
7 Newcastle Utd 65 Conference League
8 Aston Villa 59
9 Crystal Palace 48
10 Leicester City 45
11 Fulham 40
12 Everton 34
13 Nottingham Forest 34
14 Ipswich Town 32
15 West Ham 31
16 Southampton 31
17 Brighton 29
18 Wolves 26 Relegation
19 Brentford 19 Relegation
20 Bournemouth 17 Relegation
Key Insights

Feature Importance

1

Away XG

Expected goals for away team

12.8%
2

Home XGA

Expected goals against home team

12.7%
3

Away Team ELO

ELO rating of the away team

8.6%
4

Home Team ELO

ELO rating of the home team

8.0%
5

Win Percentage

Historical win rates of both teams

7.5%
6

Team Form

Recent match performance metrics

3.8%