Predicting the outcomes of MLB games with a machine learning approach

Baseball is an unpredictable sport where teams are really competitive with each other. Each team can win from the others and it takes 162 games each season to decide which teams will go to the playoffs. In this paper we will investigated if it can be predicted which team will win individual MLB games. This is done by using historical data of games and using different machine learning algorithms such as random forests and XGBoost.When using the XGBoost algorithm it shows the best results with an accuracy of 0.5552. This result can be improved when using more data, more computing power and better feature engineering.