This document describes predicting NHL players' contributions over the life of their contracts using statistical data from a private hockey analytics site. It explores various machine learning models like random forest, elastic net, and support vector regression applied to features including goals, assists, advanced stats, and quality of competition metrics to predict the following season's output. The best models were elastic net and random forest, achieving predictions within .359-.364 accuracy. Improvements could include more data, different target variables, and additional features like age and injury status.