Did My EPL Football Prediction Model Actually Work? A Rigorous Evaluation

2026-07-13 · Source: Machine Learning on Medium · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics, Sports Analytics · Depth: Intermediate, quick

Summary

An evaluation is presented for a football prediction model designed for the 2025/2026 English Premier League season. This model, developed as part of a full-stack MLOps system, predicts various match outcomes including match results, Both Teams to Score (BTTS), Over 2.5 goals, Over 1.5 goals, and Total Goals. The underlying training data encompasses approximately 1,064 matches from all English Premier League seasons up to August 2023, covering three full seasons. The article aims to detail the model's performance against real-world data and explore alternative approaches that could have improved its accuracy. The project's code is available on GitHub under "Soca-Scores".

Key takeaway

For Data Scientists or MLOps Engineers deploying predictive models, rigorously evaluating your model's real-world performance is crucial. This article highlights the importance of assessing predictions against actual outcomes, like those for the 2025/2026 EPL season, to identify performance gaps. You should analyze specific metrics such as BTTS and total goals, and be prepared to explore alternative modeling routes to enhance accuracy and reliability in production.

Key insights

An EPL football prediction model's performance for the 2025/2026 season is rigorously evaluated.

Method

The evaluation process involves assessing a deployed model's predictions for match results, BTTS, and total goals against actual 2025/2026 EPL data, then identifying areas for improvement.

In practice

• Evaluate ML models against real-world data.
• Consider multiple prediction metrics (BTTS, goals).
• Review alternative modeling approaches.

Topics

• EPL Football Prediction
• Machine Learning Models
• Model Evaluation
• MLOps System
• Sports Analytics
• Prediction Metrics

Code references

• Jnyambok/Soca-Scores

Best for: Machine Learning Engineer, MLOps Engineer, Data Scientist

Related on AIssential

• Can Machine Learning Predict the World Cup? — Predicting football outcomes requires robust feature engineering, as complex models offer minimal gains over simpler regression approaches.
• Good Benchmarks — Effective AI benchmarks must mirror real-world problems, verifying outcomes over specific approaches.
• Using FC26 to simulate the world cup ? [D] — Game simulations are unreliable for complex real-world predictions due to inherent randomness and unmodeled variables.
• Goal Analytics: A Fully Built Data Pipeline, Six Models, and a Real Backtest — A robust sports forecasting system integrates diverse models, rigorous point-in-time backtesting, and a meticulously cleaned data pipeline.
• Who Will Win the 2026 Soccer World Cup? — Building transparent, auditable forecasting models, where every assumption is explicit, offers a defensible alternative to black-box ML.
• Mastering Model Evaluation in Machine Learning — Effective model evaluation using appropriate metrics is fundamental to trusting, comparing, and improving machine learning predictions.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine Learning on Medium.