Longevity Predictor Soccer

Abstract

Longevity Predictor Soccer is an ongoing football analytics project built around a question clubs, fans, and players keep revisiting: when do elite footballers really hit their peak, and how quickly do they decline after turning thirty? In practice, clubs often offer only short-term contracts to older players because they assume the best years are already behind them, even when performance evidence is more nuanced.

The project starts from that tension between perception and reality. Transfer markets heavily reward youth and future upside, yet the last decade of Ballon d'Or winners shows that world-class output often continues well into a player's thirties. That mismatch suggests there may be undervalued players whose contract risk is overstated simply because age is being used as a blunt heuristic.

To study this, the approach uses historical player performance data with special attention to age and playing position. Past ratings are converted into percentile ranks among comparable peers, producing a more context-aware signal than raw ratings alone. That metric, along with the player's current age, is then fed into a machine learning model to estimate overall rating for the current year and support more grounded scouting or contract decisions.

What I Built

• Age and position needed to be normalized through percentile-based peer ranking before player performance could be compared fairly.
• The project challenges the blunt assumption that players over thirty are always on irreversible decline.

Impact

• Creates a scouting-oriented analytics tool for spotting undervalued veterans and more rational contract decisions.
• Applied percentile-based normalization to turn age and position context into a more defensible player-evaluation workflow.