Predicting Football Match With RNN

A winning competition approach

In recent years, predicting football match outcomes has become a fascinating field, especially with the surge in data availability and machine learning techniques. To explore this intersection, we launched a Kaggle competition challenging participants to predict the probabilities of winning, drawing, or losing for over 150,000 football matches from 2019 to 2021. These matches, spanning 860 leagues globally, presented a unique challenge: each entry included a sequence of past performances and ratings for both teams, providing essential time-series data for competitors to analyze.

READ ON MEDIUM